528: My Favorite Slap in the Face


00:00:00   Mmm, good cookies.

00:00:01   - Oh, what are you eating?

00:00:03   - Girl Scout S'mores.

00:00:04   - You know, I am not the world's biggest fan

00:00:08   of all of the Girl Scout cookies that the world loves.

00:00:11   Like, the peanut buttery ones are good.

00:00:14   - Which ones, the chocolate covered ones

00:00:16   or the sandwich ones?

00:00:17   - The former, the chocolate covered ones.

00:00:19   - Yeah, those are good.

00:00:20   Tagalogs.

00:00:21   - I mean, they're not amazing in my personal estimation,

00:00:24   but they're good.

00:00:25   Thin mints aren't really for me,

00:00:28   But the key for me, and we've probably talked about this

00:00:30   at some point, is the shortbread, the trefoils,

00:00:34   trefoils, however you pronounce it.

00:00:35   - Yeah, those are good.

00:00:35   - I could go through a box of those in a sitting.

00:00:37   It wouldn't even blink an eye.

00:00:38   - Are you in the Samoa zone, Casey?

00:00:42   - I honestly don't know.

00:00:44   Is there any way to find out?

00:00:45   - I do not know.

00:00:46   That is the best Girl Scout cookie

00:00:47   and it's the only one you should even bother getting.

00:00:50   - It's the coconut circle with caramel.

00:00:51   - Yeah.

00:00:52   - That is two things that I am convinced

00:00:54   that I don't enjoy, but Aaron will regularly point out to me

00:00:57   that any time I consume caramel or coconut,

00:01:00   I always say, you know, I really don't like blank,

00:01:03   but this is actually pretty good.

00:01:05   (laughing)

00:01:06   - Which suggests that you probably actually like it.

00:01:08   - Which suggests that I probably do.

00:01:10   - I was asking, is there the two different bakers

00:01:12   who do the Girl Scout cookies, and one of them,

00:01:14   the bad version of that is called Caramel Delights,

00:01:17   and I am unfortunately in the Caramel Delight zone,

00:01:19   so I have to get illicit imported Samoas.

00:01:21   - We have not one, but two new anniversaries

00:01:27   to celebrate, I am not even kidding.

00:01:29   - What now?

00:01:30   - Oh, I saw that in the email.

00:01:32   Now I've decoded the notes here.

00:01:34   (laughing)

00:01:36   I think I replied to that and shamed him

00:01:38   for having another anniversary.

00:01:40   - Of course you did.

00:01:41   All right, well actually, so the first one is for a Mr.,

00:01:44   let me check my notes, John Siracusa,

00:01:46   who a year ago, tomorrow as we record this,

00:01:49   but by the time you hear this, a year ago today,

00:01:51   in all likelihood, had announced

00:01:54   that he was going independent.

00:01:55   So John, how's the last year been?

00:01:57   - I could swear I put that in my calendar

00:02:00   to remind myself to talk about it.

00:02:02   Are you off by the day or something?

00:02:04   - No, it's hypercritical.co/2022/03/30/independence-day.

00:02:09   - Maybe I just missed it.

00:02:13   - We don't need to belabor this,

00:02:14   but how's the last year been, bud?

00:02:16   - That's fine.

00:02:18   I mean, I did put, like I said,

00:02:21   I did put, I thought I put something in my calendar

00:02:23   to say, hey, it's been a year or whatever,

00:02:25   but really realistically setting aside that thing

00:02:27   I probably put in my calendar like a year ago,

00:02:30   the main thing I'm thinking about is that 2023

00:02:33   will be the first full year.

00:02:35   And this is in front of my mind

00:02:36   because we're doing like tax stuff now,

00:02:38   will be my first full year without jobby job income.

00:02:41   So like, I feel like it just doesn't count

00:02:43   'cause last year was like a jobby job

00:02:45   and then I didn't this year.

00:02:46   And you know, if everything works out,

00:02:48   there'll be no jobby job income.

00:02:49   And so this will be the first full year.

00:02:51   So if you're excused to have another anniversary,

00:02:53   which is basically next year around tax time,

00:02:55   you can ask me, what was it like having an entire year

00:02:59   without a jobby job?

00:03:00   And I'm gonna say, well, with the podcast ad market

00:03:03   the way it is, it was pretty rough, but anyway.

00:03:05   - I was gonna say, you picked a hell of a year.

00:03:08   It's like the worst ads we've seen in years.

00:03:11   - It's kinda like when my son was born,

00:03:15   my first child, my oldest was born,

00:03:16   and then I got laid off like a week later.

00:03:18   So that's just kinda the way I roll.

00:03:20   - Did I know that?

00:03:21   I don't feel like I knew that.

00:03:22   It was like it was one of the ebook company got bought out and they said, "Hey everybody,

00:03:26   you could either pick up your family and move to North Carolina or whatever or you're all

00:03:30   out of a job tomorrow.

00:03:31   Let us know.

00:03:32   You get 24 hours."

00:03:33   Nice.

00:03:34   And we all know you wouldn't set foot anywhere south of Long Island.

00:03:39   If it was a job that I love for people that I thought were the people who bought the company,

00:03:44   if I didn't think they were a bunch of tools, I might have considered moving.

00:03:48   But no, they were terrible people and none of us liked them and we all just left.

00:03:53   But yeah, Alex, I remember bringing Alex to the office in his little infant carrier thingy

00:03:58   or whatever.

00:03:59   So anyway, here I am.

00:04:01   I'm still hanging in there, atp.fm/join.

00:04:03   Yes, please.

00:04:04   We're going to talk about that more in a minute.

00:04:06   But I am glad that you're still hanging in and congratulations, all snark aside, congratulations

00:04:10   on a year of independence.

00:04:12   You've made at least one, so that's a good start.

00:04:14   And then we alluded to this moments ago, but it was brought to our attention via email

00:04:18   that we have another anniversary to celebrate, which is the 10 year anniversary of friend

00:04:22   of the show, Jonathan Mann's ATP ending theme.

00:04:27   I almost said theme song, but it was actually entitled on YouTube, ATP ending theme.

00:04:31   And so it was released 10 years ago, a couple of days back, was it the 26th of March, 2023.

00:04:40   And we started using it every day.

00:04:42   I don't know, Marco, what was it, like the 27th, basically?

00:04:45   It was like immediate.

00:04:46   - Yeah, it was pretty soon afterwards, yeah.

00:04:48   - So thank you to Jonathan Mann.

00:04:50   Please check out all of his work.

00:04:52   - That song, I remember we were talking,

00:04:55   at the time, we were talking to Merlin Mann,

00:04:58   just casually like, "Hey, are you interested

00:05:00   "in maybe writing us a theme song?"

00:05:01   - Yeah, I forgot about that.

00:05:03   - And then Jonathan wrote this song,

00:05:04   you know, unbeknownst to us,

00:05:06   he kinda came out of nowhere with this song.

00:05:08   I forget exactly the timeline of it,

00:05:10   but after some brief amount of time,

00:05:12   this song was still in our heads.

00:05:14   And it was so much in our heads,

00:05:16   and I'm like, first of all, I like this song a lot,

00:05:19   and it stuck in my head.

00:05:20   Second of all, if I like it and it stuck in my head,

00:05:23   other people would probably like it as well,

00:05:25   and it would probably stick in their heads,

00:05:27   and that's probably good for our podcast.

00:05:29   So we basically went to Merlin, like, you know what,

00:05:31   if since you haven't started yet, nevermind,

00:05:33   we got something else, thanks, sorry.

00:05:34   And we started putting this song in every show,

00:05:38   and it's fantastic.

00:05:40   I love the song so much.

00:05:41   Just so thankful to Jonathan for writing it

00:05:43   and letting us use it all this time.

00:05:45   - Yep, and then Merlin did do the theme song

00:05:46   for Reconcilable Differences,

00:05:48   so he got to do a podcast theme.

00:05:49   - Yeah. - Oh, that's true.

00:05:50   There you go.

00:05:51   So yeah, we'll have some links in the show notes

00:05:53   to Jonathan's stuff, to the actual YouTube video,

00:05:56   to John's announcement post,

00:05:59   and the ATP wherein we discussed all this.

00:06:02   And it was one of my favorite, bar none,

00:06:04   one of my favorite moments of ATP

00:06:05   when John just dropped this on us.

00:06:07   And I certainly had no idea,

00:06:09   And I think Marco, you also had no idea, is that correct?

00:06:12   - What were your guesses?

00:06:12   Like I made you guess, what were your guesses?

00:06:14   - Oh, it was terrible.

00:06:15   It was truly terrible.

00:06:16   I know what you're thinking of

00:06:17   and I can't remember what we guessed.

00:06:18   I think Marco was closer,

00:06:19   but both of us were pretty far off.

00:06:22   Well, anyway, so congratulations.

00:06:24   More anniversaries to come.

00:06:25   I just gotta figure out what excuse I can come up with

00:06:27   to figure some out.

00:06:29   But maybe we'll let the occasional inflation rest for now.

00:06:33   All right, we also need to talk about,

00:06:35   speaking of ATP members and ATP.fm/join,

00:06:39   We have recorded another ATP member special.

00:06:42   Jon, would you like to tell us about it?

00:06:44   - We talked a lot about it on the actual episode,

00:06:46   so I won't go too far into it,

00:06:47   but it's another ATP Movie Club episode.

00:06:51   This time I picked the movie,

00:06:53   but I picked it kind of at the request of Marco and Casey

00:06:57   because they expected me to make them watch

00:07:02   a Studio Ghibli movie or something,

00:07:04   or The Godfather or something like that,

00:07:06   and I didn't, I made them watch Edge of Tomorrow.

00:07:09   And so they said, "Well, what if we do that?

00:07:11   What if you pick out a Studio Ghibli movie for us to watch?"

00:07:14   And then I had a big debate about which one I was gonna pick,

00:07:16   which I talk about on the show.

00:07:18   I ended up picking "My Neighbor Totoro,"

00:07:20   which is kind of straight up the middle

00:07:21   for a first Studio Ghibli movie.

00:07:23   But it's not straight up the middle

00:07:25   when the people watching it for the first time

00:07:26   are Marco and Casey.

00:07:27   So if you are an ATP member,

00:07:31   please check out this episode.

00:07:33   It may not be exactly what you expect,

00:07:35   but I think we delve into the movie in a way

00:07:39   that you probably haven't heard before

00:07:41   because the other two people on the show are first timers.

00:07:45   - I think it was quite interesting.

00:07:46   (laughing)

00:07:47   - It was, it was something.

00:07:49   So yeah, so please feel free to check that out.

00:07:51   ATP.fm/join, you can join for a month,

00:07:54   you can join for a year.

00:07:55   You can join for a month just to get this one episode

00:07:57   and then, oh, forget to cancel it, whoopsie-doopsies,

00:07:59   no problem there.

00:08:01   So do what you gotta do, but we encourage you,

00:08:04   especially in this genuinely trying time

00:08:06   when it comes to podcast advertising,

00:08:09   it would be lovely if you had a few bucks to send our way.

00:08:12   We would appreciate it.

00:08:13   And we try to do good stuff for you.

00:08:16   I don't wanna make any promises

00:08:17   about future HP member episodes.

00:08:19   We do intend to do them maybe once every month or two

00:08:24   is our goal.

00:08:25   It doesn't mean we'll succeed, but that's the goal.

00:08:26   We've been kicking around some other ideas

00:08:28   that are not movie club.

00:08:29   I'm sure we will return to the movie club well,

00:08:32   But we are trying to be inventive and creative

00:08:36   even though that's not necessarily in our wheelhouse.

00:08:38   So we will actually be talking a little bit more

00:08:41   about this probably in the post show.

00:08:42   But anyways, atp.fm/join, you can join us in watching

00:08:47   or discussing at least My Neighbor Totoro.

00:08:50   - And keep in mind that if you join up,

00:08:52   even if you're just joining up to hear this one episode,

00:08:53   like if it's a monthly membership at minimum,

00:08:56   you can listen to all of the member special episodes

00:08:59   in that time, there's only like what,

00:09:00   four or five of them at this point.

00:09:01   So you're not that far behind and you'd be getting

00:09:04   your money's worth even if you just pay for one month.

00:09:06   - Yeah, you have access to everything

00:09:07   that we've ever done as a member, anything.

00:09:09   The full fee, it isn't just from the time

00:09:12   you sign up forward, you have access to everything

00:09:14   that we've done in the past.

00:09:15   So yeah, feel free, help yourself,

00:09:17   and then just, yeah, kinda forget to cancel.

00:09:19   You'd be shocked how bad the podcast ad market

00:09:22   has been so far this year.

00:09:24   So we're fine, don't worry about us,

00:09:27   but if you've been on the fence about becoming a member,

00:09:30   this is the great time to do it.

00:09:32   - Yeah, selfishly it would be lovely.

00:09:35   There are six episodes, there are five movie club episodes,

00:09:37   the original trilogy, if you will,

00:09:39   and then the oopsie doopsie,

00:09:41   we never watched "Hunt for October"

00:09:42   and oopsie doopsie, we never watched a Ghibli movie,

00:09:45   and then the frozen dinner fiasco,

00:09:47   which we may never live down.

00:09:50   So there's plenty of good stuff for you to check out

00:09:53   if you're interested.

00:09:53   Moving on, let's do some follow up,

00:09:55   and John, there's maybe good news,

00:09:59   maybe not about HDMI quick media switching in action

00:10:02   on an Apple TV 4K.

00:10:04   Tell me about this, please.

00:10:04   - I think we talked about it when the new Apple TV 4K

00:10:07   came out and had support for this.

00:10:09   And we discussed the technology involved a little bit.

00:10:12   And then the capper was unfortunately

00:10:15   no television support this feature.

00:10:17   So you've got it on your Apple TV box, but it's pointless.

00:10:21   Well, now it's 2023,

00:10:22   the 2023 crop of televisions have come out

00:10:25   and several of them do support quick media switching.

00:10:29   What this is supposed to do for you is make it so that when the television switches something

00:10:34   about the picture it displays, it switches from 1080 to 4K, it switches from 60 frames

00:10:40   per second to 24, it switches from SDR to HDR, if it does any of those things, most

00:10:45   modern televisions will black out the screen, kind of like it's an old CRT or something,

00:10:50   make it go all black, and then it waits a couple seconds, maybe a couple three, couple

00:10:55   four seconds and then it will come back on with whatever the new settings are.

00:10:59   This is relevant because if you use my recommended settings on the Apple TV

00:11:04   that tell you to match frame rate and match what is it called match dynamic

00:11:09   range or whatever that means any time you switch between one mode and another

00:11:14   you will get that black screen and that happens surprisingly often because

00:11:19   probably when you're using the Apple TV initially you're just on the menu and

00:11:22   and you're going through the menus

00:11:23   and you can pick what you want the menus to be displayed in,

00:11:26   you know, like the little grid of all the icons,

00:11:27   all your different apps on your Apple TV.

00:11:29   You can pick if you want that to be in 1080, in 4K,

00:11:31   in SDR, in HDR, right?

00:11:33   But it doesn't matter what you pick

00:11:34   because chances are good that when you launch an app

00:11:37   and start watching something in the app,

00:11:39   it will be different than what you picked

00:11:41   because I think, actually, I think your only choice,

00:11:43   well, I guess it's not your only choice,

00:11:44   but the default choice that most people do, for example,

00:11:47   is 60 frames per second for the thing

00:11:49   where you see the apps and stuff.

00:11:51   Like why would you put that in 24 frames per second?

00:11:53   That would be weird.

00:11:54   But again, no matter what you pick,

00:11:56   if you watch a TV show,

00:11:57   that's probably not gonna be 24 frames per second.

00:11:58   If you watch a movie,

00:11:59   it probably will be 24 frames per second.

00:12:01   Is the show you're watching HDR?

00:12:02   Is the show you're watching SDR?

00:12:04   There's gonna be a mode switch somewhere in your future.

00:12:06   And that induces a black screen

00:12:09   and a couple seconds of wait, and that is annoying.

00:12:13   And the reason I recommend having that feature on,

00:12:15   even though that black screen is annoying,

00:12:18   is because to do otherwise would be to pick a mode

00:12:22   and dynamic range thing and just watch everything in that.

00:12:25   So you say 60 frames per second HDR,

00:12:27   everything's gonna be like that.

00:12:28   Well, everything's not 60 frames per second HDR.

00:12:31   And if you ask your television or your Apple TV

00:12:34   to convert everything to 60 frames per second HDR,

00:12:37   it'll do it and it will look awful.

00:12:39   No matter what you pick, pick 24 SDR,

00:12:42   it's going to be incorrect for some show.

00:12:45   So you do want it to switch, but you don't wanna wait.

00:12:48   Quick media switching was in theory some part of the HDMI spec that was supposed to help

00:12:52   with this, but as we said last time we discussed it, the only thing it helps with is when you

00:12:57   change frame rate.

00:13:00   That's it.

00:13:01   So if you change any other aspect of the thing, if you change from 1080 to 4K or back, if

00:13:05   you change from SDR to HDR, you still get a black screen.

00:13:09   So I was kind of disappointed in that standard, but it was like, well, let's wait to see when

00:13:13   it comes out on TV what it's like.

00:13:14   So now it's out on the TV, we'll put a link in the show notes to a YouTube video so you

00:13:18   You can see it in action with an Apple TV 4K on a new 2023 LG Television.

00:13:26   And when you switch only resolution, no black screen.

00:13:30   Thumbs up!

00:13:31   You know, it still does like a crossfade or whatever, but at least the screen doesn't

00:13:34   go completely black and you have to wait.

00:13:37   When you do anything else, black screen.

00:13:40   But – and I don't understand this, but you can see the results in the video for yourself

00:13:43   If you have quick media switching enabled on this particular LG television and it goes to the black screen to switch modes

00:13:50   The black screen is up for less time than if you don't have quick media switching on that makes no sense to me because it's like

00:13:57   if you change something other than

00:14:00   What do you call it frame rate?

00:14:02   It's quickly the switching shouldn't be involved and yet it is so you can see the results

00:14:06   It also puts up a big this LG television puts up a big gray banner that says quick media switching or something

00:14:11   It's like, it kind of defeats the purpose.

00:14:13   I don't need to see a banner.

00:14:14   Just do it.

00:14:15   Don't, anyway.

00:14:17   So the struggle continues for actual timely, fast switching.

00:14:22   And I guess we kind of just got that with the ARM based Macs.

00:14:26   Remember we were talking about when the M1 Macs

00:14:28   first came out, how quickly they changed resolution

00:14:31   and how we were used to the idea that on the Intel Macs,

00:14:33   of course your screen's gonna blank out for a second.

00:14:35   And when they didn't do that and it was like instant,

00:14:37   like, wow, this is great.

00:14:39   We're still waiting for the day that TVs do that.

00:14:41   But in the meantime, it seems like quick media switching

00:14:44   is a slight, very slight upgrade from not having it.

00:14:48   So if you happen to have a fancy new Apple TV 4K,

00:14:51   you happen to be in the market to buy a new television 2023,

00:14:55   I guess look for one with quick media switching

00:14:57   and then just wait with the rest of us

00:14:59   for five to 10 years when HDMI standards catch up

00:15:02   with what we want them to do.

00:15:04   - Yeah, we'll see what happens.

00:15:05   But I was, like you said, I was super disappointed

00:15:07   that it was only for, what was it, frame rate you said?

00:15:11   Or whatever it was.

00:15:12   And everything else, it was the exact same thing that we--

00:15:16   - But not the exact same thing.

00:15:17   It's faster black screen.

00:15:18   That's what's so weird.

00:15:19   All right, and the other thing,

00:15:22   the reason it's only frame rate

00:15:23   is because it's built on VRR, variable refresh rate,

00:15:27   which is a feature of HDMI where,

00:15:29   it's mostly for gaming, like when you're playing a game,

00:15:32   instead of having the game,

00:15:33   demanding that the game produce 60 frames every second,

00:15:37   and refreshing the screen at 60 frames per second.

00:15:40   Instead, you can, the television or the screen

00:15:43   or whatever says to the game,

00:15:44   just give me a frame when it's ready, right?

00:15:46   You know, if you don't have a frame ready

00:15:48   when you're supposed to, don't worry,

00:15:49   I won't refresh the screen,

00:15:50   I'll just wait for you to give me the frame.

00:15:51   That's variable refresh rate.

00:15:52   So since they already had that feature implemented,

00:15:55   it's many, many years old feature,

00:15:57   they already had that implemented,

00:15:59   you can think of changing from 24 frames per second to 60

00:16:02   as a weird kind of variable refresh rate

00:16:04   where it's like 24, 20, 20, 24,

00:16:06   and then 60 and so the tech,

00:16:08   the sort of the hardware and software to do that

00:16:11   was already kind of built in

00:16:12   so they could build this on top of it,

00:16:14   which basically makes quick media switching.

00:16:16   It's not like, it's like they didn't really do any work.

00:16:18   They're like, well, we've already got VRR,

00:16:19   can we just do something with that?

00:16:21   Sure, we'll call it quick media switching.

00:16:22   It's basically VRR, but for your television.

00:16:24   Anyway, disappointing, but you know,

00:16:27   it's HDMI, what do you expect?

00:16:29   - And then John, you have all sorts of new figurative

00:16:33   and potentially literal tools in your tool chest

00:16:35   with regard to destroying and recreating

00:16:38   your own custom furniture.

00:16:40   - Yeah, my little story about cutting some threaded rod

00:16:43   to shrink some furniture that I bought

00:16:45   resonated with a lot of people.

00:16:48   The most common suggestion I got was a better place

00:16:51   to buy threaded rod and other things.

00:16:53   I bought mine from like grainger.com or something.

00:16:56   Everybody recommended it.

00:16:58   McMaster-car, it's mcmaster.com, M-C-M-A-S-T-E-R.com.

00:17:03   They sell stuff like that, threaded rod fasteners

00:17:05   or whatever, very popular company, everybody who wrote in about them loves them, says this

00:17:09   is where you should get this stuff.

00:17:10   I did look and they did have Threaded Rod, the price was similar to what I paid, they

00:17:15   didn't have stainless steel though, which is kind of what I preferred as opposed to

00:17:18   like zinc coated whatever, but anyway, that's a website people recommended.

00:17:23   Sean Cameron was one of many people to recommend tool lending services, everybody said this

00:17:31   as well expressed, "As a fellow technologist who is tool inclined and having a lot of the

00:17:36   same predilections as John about how things should be around the house, I also find myself

00:17:39   holding back on buying all sorts of tools, including vices."

00:17:43   As in things that squish things.

00:17:44   "I wanted to mention that many communities have either workshop spaces or tool lending

00:17:49   libraries through which you can get access to specific tools for a specific job.

00:17:53   This is a much better approach than spending money on your own version of a tool that you

00:17:56   only need once or twice."

00:17:58   And then, what's this?

00:18:00   CC Helberg tooted to say, "I couldn't find one near you, but tool-ending libraries are

00:18:06   a thing for both woodworking and more, and also for things like specialized-shaped baking

00:18:09   dishes or even fly-fishing kits.

00:18:11   I've seen everything from power tools, drill bits, ladders, garden tools, and even a cement

00:18:14   mixer in them."

00:18:16   And the website, which we'll link in the show notes, is localtools.org/find.

00:18:20   Many people sent me this URL while also noting that they could not find any near me.

00:18:24   So I guess it's not everywhere, but check around you and maybe nearby.

00:18:29   That would be a good idea if I cared a lot more about cutting some threaded rod.

00:18:35   Again I got the job done.

00:18:36   It would have taken me more than a day to find this place, drive to it and see if they

00:18:39   had something for me and then return it and blah blah blah.

00:18:41   So probably not appropriate for what I was doing but very handy for other things.

00:18:45   And it does save you from having to buy a tool.

00:18:48   And finally lots of people suggested something that I already knew about but did not employ

00:18:55   which is when cutting a threaded rod and you don't want to script the threads with your

00:19:00   hacksaw or whatever, one trick is to thread one or two nuts onto the threaded rod, either

00:19:04   having two nuts tightened against each other or two nuts with a gap between them so that

00:19:09   you're basically protecting the threaded rod with a metal thing that is fixed in place.

00:19:13   And also, after you finish cutting through it with your hacksaw pressed against one of

00:19:19   the nuts, then you back the nut off of the threaded rod and that will smooth out any

00:19:24   burrs that are on the threaded rod, you know, to get it so the thing goes on and off.

00:19:28   All good ideas.

00:19:29   I did not have any nuts to put on the threaded rod.

00:19:31   And they're like, "Oh, you can just buy one of those at McMaster."

00:19:34   Yeah, I could have, but now I'm ordering another thing, waiting for it to come, or I'm going

00:19:37   to another home store and looking for an M6 1.0 nut and buying a bag of them for $5 that

00:19:42   I'm never going to use again, and yada, yada, yada.

00:19:44   Anyway, I cut it by hand and I survived, so thank you for the suggestions, but if you

00:19:48   had to cut more than one piece of threaded rod, get a vice.

00:19:52   Get a vise, get some nuts to fit on it,

00:19:54   get a new hacksaw blade, I guess don't do what I did.

00:19:58   - Only on this show do we have a threaded rod follow up.

00:20:02   - Yeah. - Indeed.

00:20:02   - And everybody wanted me to buy a hacksaw blade,

00:20:04   like, "I'll buy a hacksaw blade, they're $2."

00:20:06   They're $6, you know, they're not $2, but yeah.

00:20:09   (laughing)

00:20:11   - It's a tough ad market,

00:20:14   we can't be buying hacksaw blades every week.

00:20:15   - Right, right?

00:20:17   Yeah, no, they're not, yeah, you can buy new one all day.

00:20:19   I should probably get a new one, but you know, again,

00:20:20   I was just the home store for thread rod.

00:20:23   I didn't think to look for hacksaw blades.

00:20:24   Then you gotta take the old one off and blah, blah, blah.

00:20:27   Next time I need to use my hacksaw, I will definitely do it.

00:20:30   But truthfully, I thought the hacksaw blade

00:20:32   was not as in rough shape as it was.

00:20:34   Because like, how often do I use it?

00:20:36   But I had forgotten that this, I think this hacksaw

00:20:38   was like from my father-in-law's tool collection.

00:20:42   So it had seen a lot of use before it even got to me.

00:20:45   And I was under the impression that it was actually

00:20:47   a newish hacksaw with a newish blade and I was wrong.

00:20:50   [LAUGHTER]

00:20:52   Ay yai yai.

00:20:53   Well, I'm glad we have that resolved.

00:20:55   Josh Calvetti writes with regard to camera apps,

00:21:00   as in for physical cameras and geotagging and things.

00:21:03   Josh writes, "When I was shooting Sony about a year ago,

00:21:05   I was using Camrot--" I'm assuming that's a camera mode

00:21:08   or something.

00:21:09   Anyway, "for geotagging and control of my A6600.

00:21:12   Much more user friendly and stable than the official Sony

00:21:15   app, and it doesn't require you to sign up for anything."

00:21:17   and that's C-A-M-R-O-T-E dot app on the web.

00:21:20   We'll put a link in the show notes.

00:21:22   - Yeah, there's a lot of third party apps.

00:21:23   As we were complaining about first party apps,

00:21:24   like the apps from Fuji or Sony

00:21:26   or whatever, how terrible they are.

00:21:27   There are actually a lot of third party apps

00:21:29   that will work with various brands of camera

00:21:30   and as you would imagine, they're all, you know,

00:21:32   much, much better than the first party ones.

00:21:34   I actually did try this camrote thing.

00:21:37   I don't have occasion to use my big camera in geotagging.

00:21:40   Like I'm just using it around the house.

00:21:41   I guess I could still geotag those,

00:21:42   but I can geotag those manually too

00:21:44   'cause I'm literally in my house.

00:21:46   But next time I go, you know, on my next outing,

00:21:48   I'm definitely gonna use that because I said before,

00:21:50   I can't actually sync both of my Sony cameras with my phone

00:21:53   because the stupid Sony app only lets you sync

00:21:54   one camera at a time.

00:21:55   So I will try the third party one and we'll see how it goes.

00:21:58   - And then speaking of cameras,

00:22:00   but this time ones within phones,

00:22:03   apparently the Samsung fun hasn't ended yet.

00:22:06   You can opt into, so this is different

00:22:09   than the moon discussion we had a couple of weeks back.

00:22:11   The moon thing was happening just kind of automagically,

00:22:14   but you can also opt into quote-unquote "remastering" photos.

00:22:19   And according to one person, and the Verge picked this up,

00:22:23   and they said that they weren't able to recreate it themselves.

00:22:26   But nevertheless, according to this one person,

00:22:29   they took pictures of their baby

00:22:31   and then tried this remaster thing.

00:22:33   And this very gummy baby suddenly had teeth

00:22:36   because the remaster thing kind of just thought,

00:22:40   oh, that human should have teeth there.

00:22:42   And it's a little creepy.

00:22:44   I think it's not as a dramatic problem or difference,

00:22:48   perhaps is a better word for it, as the whole moon thing,

00:22:51   but it's definitely a little weird and kind of funny.

00:22:54   So there's a link in the show notes to the Verge

00:22:56   that covers all this.

00:22:57   It has some GIFs that you can look at.

00:22:59   It's something else.

00:23:00   - I think the moon feature is on by default.

00:23:02   I'm not sure if this one is,

00:23:03   but this is definitely the type of thing

00:23:05   where if you didn't notice it

00:23:07   and then went back years later,

00:23:08   especially if it burns it into your picture,

00:23:10   you know, like if you can't go back to the original

00:23:12   as captured by the camera or whatever,

00:23:14   You're like, how does this three-week-old baby have teeth?

00:23:18   It's like the magic of Samsung.

00:23:19   [LAUGHTER]

00:23:22   We are brought to you this week by Collide.

00:23:25   And they have some big news.

00:23:26   If you're an Okta user, they can get your entire fleet

00:23:29   to 100% compliance.

00:23:31   How?

00:23:32   If a device isn't compliant, the user

00:23:34   can't just log into your cloud apps.

00:23:35   They just can't do it till you've fixed the problem.

00:23:38   It is that simple.

00:23:39   Collide patches one of the major holes in zero trust

00:23:42   architecture, device compliance.

00:23:45   Without Collide, IT struggles to solve basic problems,

00:23:48   like keeping everyone's OS and browser up to date.

00:23:50   Insecure devices can be logging into your company's apps

00:23:53   because there's nothing there to stop them.

00:23:55   Collide is the only device trust solution

00:23:58   that enforces compliance as part of authentication,

00:24:01   and it's built to work seamlessly with Okta.

00:24:04   The moment Collide's agent detects a problem,

00:24:06   it alerts the user and gives them instructions to fix it.

00:24:09   If they don't fix the problem within the time you set,

00:24:12   They're blocked.

00:24:13   Collide's method means fewer support tickets,

00:24:15   less frustration, and most importantly,

00:24:17   100% fleet compliance.

00:24:20   Visit collide.com/ATP to learn more or book a demo.

00:24:25   That's Collide spelled K-O-L-I-D-E.

00:24:27   Collide, K-O-L-I-D-E.com/ATP.

00:24:32   Thank you so much to Collide for sponsoring our show.

00:24:35   - All right, and then we have some information.

00:24:41   Did we cover the true Siri experience?

00:24:43   I didn't think we covered this on the show, did we?

00:24:45   - Yeah, that was the title of the last episode

00:24:46   and we covered this exact thing.

00:24:48   I couldn't remember the details,

00:24:49   but then I revisited it or whatever.

00:24:50   It was the person asking Siri for the weather.

00:24:53   They just got a HomePod and they said,

00:24:55   "I just tried to ask it the weather

00:24:56   "and all it said was done."

00:24:57   - Okay, my mistake.

00:24:59   So anyway, so with regard to that,

00:25:01   Benjamin Mayo wrote and had a really good point.

00:25:04   Benjamin wrote, "Siri having just said done

00:25:07   "implies that it's running a shortcut.

00:25:09   "Do you happen to have a Siri shortcut in your library?"

00:25:11   Benjamin was talking to the person who originally posted this,

00:25:13   "Do you happen to have a shortcut in your library

00:25:15   that has the name somewhere to weather?

00:25:16   It might be getting confused and running the shortcut

00:25:18   instead of actually looking up the weather."

00:25:19   And the original poster, Nairobi, wrote back and said,

00:25:21   "I sure did.

00:25:22   I had two that seemed to have no real purpose.

00:25:25   I deleted them, and Siri seems to be able to answer me now,

00:25:27   consistently even, which is good news."

00:25:29   So as much as I love crapping on Siri,

00:25:31   turns out this one was a legitimate oops.

00:25:34   -Yeah, but this is a situation where you're like,

00:25:36   "Yeah, there is a problem,

00:25:37   but how would anyone figure that out?"

00:25:39   You know, like you just, you, you ask, you know, that you're able to ask the home pod.

00:25:43   What the weather is.

00:25:44   You do that and it does something different.

00:25:46   Like what's your next debugging step.

00:25:48   Are you going to know that you have some random shortcut that happens to have, you

00:25:51   know, it's called check the weather or something.

00:25:53   Cause you clicked on some link that you forgot about a year ago.

00:25:56   It's the debug ability of voice and systems.

00:25:58   The discoverability we talked about before of like, what can I actually say to you is not great.

00:26:03   And then when something goes wrong, like what do you do to figure out what the problem is?

00:26:09   because the interface is talking,

00:26:11   and if it doesn't understand what's the weather,

00:26:13   the idea that it's going to understand you conversing

00:26:16   to it about what went wrong is probably not particularly

00:26:20   likely, although this next item has more on that topic.

00:26:24   - Indeed, so somebody put together, basically,

00:26:27   chat GPT inside an iOS shortcut,

00:26:30   which is kind of bananas.

00:26:32   And there's a really interesting,

00:26:33   unfortunately it's on Medium,

00:26:34   but there's a really interesting post about this,

00:26:36   and there's a video included as well.

00:26:39   It is fascinating and from the way the video is edited and I think there was a little bit of

00:26:44   complementary or

00:26:46   Aggressive editing on this video I think but nevertheless the way the video is edited it is darn impressive

00:26:52   So the the person whose name I don't have in front of me wrote I explained everything in plain English

00:26:58   Oh, I'm sorry. So this is with regard to how did they put hang all this together? And so

00:27:03   Apparently what this person wanted to do was have ChatGPT basically interpret their verbal

00:27:11   commands and turn it into a JSON payload that could be sent on to other things.

00:27:16   And so how did they convince ChatGPT to do this?

00:27:20   Well, they literally explained it to them.

00:27:22   So now quoting from this post, "I explained everything in plain English.

00:27:25   I described the types of requests, the exact structure of the response, and asked it to

00:27:28   behave like a sentient AI, giving advice even for personal questions.

00:27:31   I also provided a few details about time location in the devices in rooms in the house. From this

00:27:36   we will receive a perfectly structured message, and that's all there's no, that's all there is to programming it. So there was no

00:27:42   you know, direct specification of here's an example JSON.

00:27:46   It was just build a JSON object that has the keys and values and so on and so forth. It's nuts.

00:27:51   So there's an example here, an example command quote, "I sent my son to bed to read for another 20 minutes.

00:27:59   "Can you switch off the lights in his room

00:28:00   when it's time to sleep?"

00:28:02   And sure enough, this works.

00:28:03   And so in this case, GPT-3 understood

00:28:05   that it is probably the bedroom that needs switching off,

00:28:07   and it added the correct timestamp,

00:28:09   which is 20 minutes after the time we passed the request.

00:28:11   And you can see a little sample JSON there.

00:28:13   All of this done without actually writing any code,

00:28:16   at least on the chat GPT side.

00:28:18   And then I guess the other end of this

00:28:20   was like some bananas, just bananas complex shortcut,

00:28:25   Iowa shortcut that processes this JSON

00:28:27   and takes action on it.

00:28:28   I think it's like sending it to Home Assistant or something.

00:28:32   So the reason I put this in here,

00:28:33   and the reason everyone was sending this to me,

00:28:35   is because it's exactly what I described in the last episode

00:28:38   when I talked about how useful is ChatGPT

00:28:41   to making Siri better, essentially.

00:28:44   And I said, well, one thing you could do with it

00:28:47   is have it interpret what the person is saying

00:28:51   and then translate it to the very, very limited

00:28:54   and rigid structured vocabulary of Siri,

00:28:57   Because that seems to be the stumbling point.

00:28:58   Like, you have to phrase things in a certain way for them

00:29:00   to work with Siri.

00:29:01   We talked about this with adding new words,

00:29:03   taking six weeks to rebuild the database and everything.

00:29:06   Siri can understand all sorts of things that you say,

00:29:08   but every single one of those had

00:29:09   to be thought of and explicitly put into Siri by a person.

00:29:13   The number of variations are not infinite.

00:29:15   You could never say something to Siri like, hey, dingus,

00:29:18   I sent my son to bed to read for 20 minutes.

00:29:20   Can you switch off the lights in his room

00:29:22   when it's time to sleep?

00:29:23   Siri will not make heads and tails about it.

00:29:25   be like, "That's not one of the forms that I know how to parse, I have no idea what you're

00:29:29   talking about."

00:29:30   Or it'll make a bad guess, or terrible things will go wrong, right?

00:29:32   So the idea that you have one of these large language models sitting in front of Siri,

00:29:38   listening to what you say, figuring out, in this case, once it figures out what you want,

00:29:42   also formulating a JSON message that it then sends to Home Assistant or some other thing,

00:29:47   or translates into the form that Siri understands it.

00:29:51   Because the chat GPT thing, the large language model figured out, you probably mean the key

00:29:55   kids bedroom, 20 minutes from now means that's when it should happen and you

00:29:58   want the lights to go off. Like it figured all that out and then it can

00:30:01   issue a command "hey dingus" in 20 minutes turn the lights off in the

00:30:05   bedroom and that Siri can understand that. I don't necessarily think layering

00:30:09   things in that way is the best approach for Apple but at the very least it is a

00:30:13   way to take two things that we have now. Siri, that can do a bunch of things as

00:30:18   long as you phrase it in one of a very long list of ways, and large language

00:30:23   models that can take a bunch of input text and figure out the most likely output text

00:30:30   for it that fits the prompt and everything and sew those two things together.

00:30:35   The video is a little bit janky because obviously having to run a shortcut is not the same as

00:30:40   being able to say "hey dingus" and shortcuts take time to run.

00:30:43   You can see the edit points in the thing where it's not as seamless as you could imagine.

00:30:46   But of course if Apple implemented this they wouldn't make you run a shortcut to do it.

00:30:49   They would fuse this into Siri and they say,

00:30:51   "Oh, it's Siri 2, now powered by quote unquote AI."

00:30:55   And honestly, it would be better, right?

00:30:57   'Cause we don't care what's happening.

00:30:58   We don't care that under the covers

00:31:00   it's a language model translating it into

00:31:02   a very text adventure style thing.

00:31:04   All we know is that now we are able to say more things

00:31:08   and actually get what we want from them.

00:31:10   - Yeah, I mean, it was certainly impressive.

00:31:13   - Honestly, I would even take just simpler things.

00:31:15   Make Siri work every time.

00:31:17   - Imagine that.

00:31:18   It's just, it fails in such weird, stupid ways so often.

00:31:23   Like, honestly, this is, you know,

00:31:26   I go on a rollercoaster up and down

00:31:28   with what I currently think about my HomePods

00:31:31   as I have for the entire lifetime of these products now.

00:31:35   And I'm currently at a bit of a down phase

00:31:39   in terms of the functionality of Siri and Apple Music.

00:31:44   Like, it's, oh, it's so buggy.

00:31:47   It's so unreliable.

00:31:49   These are brand new products.

00:31:51   This is like, they're so buggy.

00:31:54   Like now they're faster and they're bugs.

00:31:57   Like they behave buggy faster.

00:32:01   But they're still, and it's, oh.

00:32:05   Can somebody like, somehow make sure

00:32:10   that Tim Cook listens to music every single day

00:32:12   using a stereo pair of HomePods and operating them by a Siri?

00:32:16   Like somehow, someone make that happen.

00:32:18   Like somehow in some non-creepy way,

00:32:21   replace all the music playing equipment

00:32:24   in Tim Cook's home and office,

00:32:26   again in a non-creepy way,

00:32:28   with stereo pairs of home pods.

00:32:30   And just make sure that he has to operate them

00:32:33   every single day.

00:32:34   And let's see if maybe this product

00:32:36   can't get a little bit better.

00:32:37   - That or use Apple Music in any platform,

00:32:40   for any reason.

00:32:42   - Well 'cause that's using Apple Music.

00:32:43   I seriously doubt he's a Spotify user.

00:32:46   So he's gonna be using Apple Music via Siri,

00:32:49   this'll be good.

00:32:50   - I'm not sure if his main motivational driver

00:32:54   is how frustrating he finds the products.

00:32:57   Steve Jobs, sure, if something went wrong for him,

00:33:01   make it his mission in life to make sure that that gets

00:33:04   fixed 'cause he is embarrassed to be shipping a bad product.

00:33:07   - He would burn the world down until it was fixed.

00:33:10   - Or at least he would try or until he gets bored

00:33:11   and moves on to something else.

00:33:12   But anyway, I feel like Tim Cook's reaction

00:33:15   would be to look at how the HomePod is selling and if sales seem like in line with projections

00:33:21   then I guess everything's fine.

00:33:22   Yeah, but see, and again, the HomePod's one part of this. Like this is why, you know,

00:33:27   we keep harping on, hey, Siri has to be better and not only in these, you know, cleverness

00:33:34   ways, like as we are seemingly in full swing now of the AI revolution here, you know, the

00:33:40   expectations people have are going to keep going up for how smart they expect it to be.

00:33:45   But also, it still doesn't get the basics right.

00:33:49   It still is unreliable and slow

00:33:51   and does stupid things with basic requests very frequently.

00:33:54   And inconsistent is another big problem that it has.

00:33:57   It's very inconsistent.

00:33:58   Apple is not only gonna fall behind

00:34:01   in competitive expectations of assistants

00:34:04   as they all move more into AI stuff,

00:34:07   but also, Apple is about to launch a brand new product

00:34:13   that seems like it's gonna be a pretty big bet

00:34:16   the company is making, that also by all accounts

00:34:20   seems like it might be pretty heavily relying on Siri

00:34:23   for certain functionality.

00:34:24   How are they gonna do that if Siri continues

00:34:29   to have the reputation of seeming to work a lot better

00:34:33   in Apple executives' homes than in any other house

00:34:36   in the world?

00:34:37   I don't know anybody for whom Siri works

00:34:40   as well as Apple seems to think it works.

00:34:43   and this is gonna hold them back.

00:34:45   It's gonna keep holding them back.

00:34:47   Imagine the products that they envision.

00:34:51   I mean, look, obviously I'm talking about

00:34:52   the VR headset thing, but also look at things like AirPods,

00:34:56   or the Apple Watch, or the phone, or the HomePod.

00:34:59   All of their products now involve Siri in some way,

00:35:02   to varying extents, some more reliant on it than others.

00:35:06   If they try to launch a product that depends heavily on Siri,

00:35:10   they're gonna present it one way,

00:35:11   and that'll be nice and it'll seem like everything

00:35:14   is awesome and works, but then when we actually get

00:35:17   the product, it's gonna have all these weird inconsistencies

00:35:20   and shortcomings and that's gonna make the product

00:35:23   itself look and work badly.

00:35:26   Siri is such a fundamental technology

00:35:30   to Apple's modern product line and they keep only leaning

00:35:33   more into that over time and for the amount

00:35:37   that they are relying on Siri for the operation

00:35:41   and success of their products,

00:35:43   they seem to be allowing it to be

00:35:45   a very poor performer in quality.

00:35:48   They care so much about so many of the details

00:35:52   and the fundamental technologies their products depend on.

00:35:56   And then Siri is just miserable.

00:35:59   And I don't understand why they don't seem

00:36:02   to put a higher priority on making

00:36:04   that fundamental technology as good as it can be.

00:36:09   I almost wonder if it's because when you're on the inside,

00:36:12   you see how the sausage is made or not,

00:36:14   depending on how you want to look at it.

00:36:16   And, you know, maybe they all know it's trash,

00:36:19   but they can explain it away.

00:36:21   Well, it's garbage because blah, blah, blah.

00:36:23   Well, it's garbage because politics.

00:36:25   It's garbage because, you know, servers.

00:36:28   It's garbage because any number of reasons.

00:36:30   And I mean, I've been told from anyone I know

00:36:34   that works or has worked at Apple

00:36:36   that they are their own biggest critic,

00:36:39   which I would believe, but golly, from an outsider's point of view, and we're going to be talking about this a lot later,

00:36:45   from an outsider's point of view, we sure can't tell, because Siri sure ain't getting better.

00:36:50   My keyboard on my phone still wants to change W-E-L-L to W-E-apostrophe-L-L and vice versa,

00:36:57   no matter what I do, it's always choosing the wrong one.

00:37:00   It sure doesn't look like anyone cares from the outside.

00:37:04   And at some point, everyone has a different line,

00:37:08   but at some point people are gonna stop being like,

00:37:10   well, it's okay.

00:37:12   At some point it's just gonna be so frustrating

00:37:14   that people are gonna stop using these products.

00:37:16   I mean, I don't have HomePods

00:37:18   or any other voice cylinder in the house

00:37:22   because the Amazon one got way too chatty

00:37:26   and all it wants to do is have me talk to it

00:37:28   and advertise things to me and so on and so forth.

00:37:31   - By the way. - Yeah, exactly.

00:37:33   Then I never wanted a home pod originally because they were too expensive and then later

00:37:38   I just didn't feel like it was solving a need I have and now I

00:37:41   Certainly don't want any because I'm such a Sonos fanboy. Nobody can nobody even wants to hear me talk about it anymore

00:37:46   So it's it's if the home pod to your point Marco if the home pod in Siri were amazing

00:37:52   If they were really and truly great, I would probably have one in the house. I haven't tried the Google stuff

00:37:58   And again, I don't feel like this is a need

00:38:00   I need to fill and maybe the Google stuff is great

00:38:03   I'm waiting for John to pipe in as soon as I stopped talking, but if if Siri was amazing

00:38:08   I would probably have a home pod by now

00:38:10   But why would I spend a pile of money for a home pod even leaving aside the Sonos stuff?

00:38:14   Why would I spend a pile of money on a home pod when one of the marquee features?

00:38:18   Never friggin works from everything I've ever heard like why would I do that? It just seems bananas

00:38:24   I know John tell me I'm tell me I'm a dummy and that I should get some Google stuff

00:38:27   We talked about this last week, like with the difficulty of Siri.

00:38:30   One of the downsides of being early in the market, like Apple was pretty early with a

00:38:36   voice assistant heavily integrated into its product family that does voice assistanty

00:38:42   things, right?

00:38:43   Siri was 2011 or whatever, as we discussed, and people who come later with using different,

00:38:49   entirely different approaches to solving this problem, like the large language models, which,

00:38:53   give you capabilities that Siri does not have,

00:38:56   although there is, like we said, there are still gaps

00:38:58   that a large language model can't do that Siri can do.

00:39:01   But anyway, using more modern technology,

00:39:03   yeah, you're later to the market

00:39:05   and you missed out on all those years

00:39:06   of having products with these features,

00:39:08   but you get to start with a newer, better technology

00:39:10   that's on a faster trajectory.

00:39:13   It could be that whatever, however Siri is made

00:39:16   is sort of an evolutionary dead end

00:39:18   in terms of how it's structured and programmed,

00:39:21   and there are a bunch of new branches

00:39:22   going off in other directions.

00:39:23   So, like, for Apple to say, okay, well, Apple was early

00:39:28   and they had this other thing,

00:39:30   but now they can just use the new thing.

00:39:31   Well, the problem for Apple is they can't replace Siri

00:39:34   with something built on new technology,

00:39:36   unless it can pretty much do everything that Siri does,

00:39:38   'cause you don't wanna have a regression

00:39:39   where it's like, well, you used to be able to use Siri

00:39:42   to do these hundred things,

00:39:43   but now we have a new quote unquote AI-powered Siri,

00:39:47   but it can do one fifth of the stuff.

00:39:48   So they've kind of, you know, again,

00:39:50   it's the curse of being early.

00:39:51   you build up all this functionality,

00:39:52   Siri, for all we complain about it,

00:39:54   is integrated into so many products

00:39:56   and it can do so many different things.

00:39:58   And even though we don't use all those things,

00:39:59   someone out there is relying on the fact

00:40:01   that you can ask Siri to predict the temperature

00:40:05   in a different country a month from now or something,

00:40:08   and then if you come out with a new one

00:40:10   that's powered by AI and it can't do that,

00:40:11   they've lost functionality.

00:40:12   So Apple has a difficult,

00:40:15   kind of like it was with the operating system.

00:40:16   If you build an operating system before,

00:40:18   like memory protection and preemptive multitasking

00:40:20   common and you build this huge customer base and all these apps built on it, yeah you got

00:40:24   all those years of good money but now when it comes time to have a modern operating system,

00:40:27   someone who starts from scratch right now can build an operating system with all those

00:40:31   features from day one whereas you have to kind of retrofit it and it's a more difficult

00:40:35   task so it's not that bad with Siri but I think it is actually a challenge to use better

00:40:41   more modern technologies to make a better Siri while also sort of replacing all the

00:40:46   the functionalities. You have to kind of do a piecemeal where it's like well, excuse me,

00:40:51   well unbeknownst to you when you ask this we take this path in the code and we all go

00:40:55   off into the new like large language model things but when you ask anything else it's

00:40:59   the old Siri path and they slowly replace it from the inside like this is just basic

00:41:02   you know software engineering product management stuff but it is difficult and I hope something

00:41:08   like that is happening inside Apple like there is lots of motion in the sort of AI section

00:41:14   of the company.

00:41:15   Granted, a lot of it has been related to ML, which was the other buzzword before AI, you

00:41:19   know, machine learning.

00:41:20   We've seen lots of ML-powered features being built into Apple's applications and devices,

00:41:25   particularly around the camera and, you know, or any of the stuff, even the keyboard autocomplete

00:41:30   that you're complaining about was "ML-powered," you know, and it doesn't seem to be working

00:41:34   out that well.

00:41:35   But anyway, they've been doing things.

00:41:36   It just seems like the things they've been doing have not been Siri.

00:41:40   Siri's been sitting there, being what it is, taking six weeks to rebuild its database in

00:41:44   2014, hopefully that's better now, but not really getting better, better.

00:41:48   So I mean, maybe this is the new thing that will be, you know, we have all these like

00:41:53   five to 10 year projects that we talk about in this program of like, when is Apple going

00:41:56   to do X?

00:41:57   And eventually they do do it.

00:41:58   And every time that happens, someone says like, well, now what are you going to complain

00:42:01   about?

00:42:02   Now that you've got the Mac Pro, now what are you going to complain about?

00:42:03   And you got this one, you got that.

00:42:05   It's like, there's always something.

00:42:06   There's always something, and Siri is kind of bubbling up to the top as the, you know,

00:42:11   long-term thing that Apple needs to deal with.

00:42:14   They got a new file system.

00:42:15   The Mac Pro, they build a new one, but then they forget about it for five years and we

00:42:18   freak out again.

00:42:19   So that'll be evergreen, but like, you know, they have laptop CPUs that don't overheat,

00:42:23   that are really fast, low power, like a lot of the things, they fix the keyboard, right?

00:42:28   They knock down a lot of these things.

00:42:30   They got a new operating system with memory detection and preemptive multitasking, right?

00:42:34   But there's always something else, and it seems like Siri,

00:42:37   it might be the long pole, at least until

00:42:39   the headset arrives and we have a whole new thing

00:42:41   to complain about.

00:42:42   (laughing)

00:42:43   - Which apparently is going to be in early June,

00:42:45   but we're gonna talk about that.

00:42:47   Yeah, yeah, yeah.

00:42:48   I don't know, where are we?

00:42:49   I feel like we got, oh, the ChatGBT thing, right.

00:42:52   So, anything else on that before we move along?

00:42:54   - I think we've covered it.

00:42:56   - All right.

00:42:57   Microsoft has threatened to restrict data

00:43:00   from rival AI search tools, according to Bloomberg.

00:43:03   Bloomberg writes, "Microsoft has threatened to cut off access to its internet search data,

00:43:07   which it licenses to rival search engines, if they do not stop using it as the basis

00:43:10   for their own artificial intelligence chat products.

00:43:13   The company has told at least two customers that using its Bing search index to feed their

00:43:17   AI chat tools violates the terms of their contract."

00:43:20   Whoops, please.

00:43:21   So, yeah, we've talked about this before in the context of like artists having their artwork

00:43:27   used as training data, you know, and whether that is legal, whether it should be legal,

00:43:32   whether it is ethical, and how these things are going to work themselves out in court

00:43:39   cases, depends largely on who the litigants are.

00:43:42   In this case, it's a bunch of big companies, Microsoft and the other companies that it's

00:43:45   licensing stuff to.

00:43:48   Not surprisingly, a big company thinks it's perfectly fine for them to trade their large

00:43:52   language models on every single thing they can find on the internet, but not fine for

00:43:57   somebody else to use their search index stuff to train their models.

00:44:01   So everyone wants, I, big corporation,

00:44:04   should be able to get any data I want

00:44:06   and use it to train my thing.

00:44:07   But once I've done that, nobody can use

00:44:10   what I've generated to train their things.

00:44:12   Because now, I've created value.

00:44:14   See what I've done there?

00:44:15   You can't use mine to train your thing.

00:44:17   And it kinda gets into the whole thing of like,

00:44:18   the argument added, certain, whatever it is

00:44:22   where you extend something to a certain degree.

00:44:25   Like well if there's no humans creating anything,

00:44:26   and it's just AIs creating things,

00:44:28   and they're not allowed to train off each other's data,

00:44:30   eventually they all just shrivel and die in place

00:44:32   because no one wants to share their data

00:44:33   and no one is producing any new data

00:44:35   and you've already trained on everything else

00:44:37   and they just sort of like, I don't know,

00:44:38   it's like the AI inbreeding where they just shrivel up

00:44:43   and become little shells to themselves.

00:44:45   So yeah, there was one other story here

00:44:47   that I actually didn't put out,

00:44:48   like the US Copyright Office is forming a committee

00:44:52   to discuss forming a committee to discuss researching,

00:44:56   whatever, they're doing something about like

00:44:57   the legality of copyright and AI training or whatever.

00:45:01   But unsurprisingly, big companies think

00:45:04   that they should get everything

00:45:06   and nobody should get what they do.

00:45:07   And Microsoft is trying to enforce that in their contracts

00:45:09   and we'll see how this all plays out.

00:45:11   But yeah, it's great when this stuff is up and coming

00:45:14   and it's like a free for all and like,

00:45:16   oh, nobody minds, it's just an academic project.

00:45:18   Oh, this is new and exciting.

00:45:19   Nobody really cares.

00:45:20   But then all of a sudden when there's real money to be made

00:45:22   and people are making products, they're like, wait a second.

00:45:24   These, as I said before, these products have no value

00:45:27   without good data to train them on.

00:45:29   Where does that good data come from

00:45:31   and what relationship is there between the data

00:45:34   that you're training on

00:45:34   and the product that you make from it?

00:45:37   And Microsoft is saying like,

00:45:38   "If we train on a bunch of this data,

00:45:40   "you can't take the stuff that we've trained

00:45:42   "to train your training."

00:45:43   (laughing)

00:45:44   It's like, I bet, you know,

00:45:45   'cause people are putting AI generated images on the web

00:45:48   and like in tweets and stuff like that.

00:45:50   And then other image generation AIs are being trained

00:45:55   on images generated from other AI things,

00:45:57   but they say, "Whoa, whoa, you can't train on those.

00:45:59   "There's the product of our machine learning.

00:46:01   "You can only train on things from actual human artists.

00:46:04   "You can steal that, no one cares about them.

00:46:05   "But once Microsoft uses our technology

00:46:08   "to generate something like that,

00:46:09   "you can't train on that, it's in our contract."

00:46:12   You know, and there's a bunch of court cases

00:46:14   surrounding this or whatever,

00:46:15   so it is rapidly heading towards

00:46:17   what will surely be a series of terrible court decisions

00:46:20   that we will complain about on the show.

00:46:22   - No, of course not.

00:46:24   I just love the idea of like Microsoft, Google, Apple,

00:46:27   Amazon, all like legally barring each other

00:46:30   from looking at each other.

00:46:32   Don't let your larger language models

00:46:34   look at any of my data.

00:46:35   If you do that, your data is tainted

00:46:36   and we own your whole company.

00:46:37   It's like, hey, no, that's not fair.

00:46:39   And then they'll probably do what they do with patents

00:46:40   and everything is they just have

00:46:41   these cross-license agreements.

00:46:42   So it's like, look, this is annoying, we all hate it.

00:46:44   Let's just have a giant patent cross-licensing agreement

00:46:47   that says we all agree we can use each other's patents

00:46:49   because the whole patent system is incredibly dumb

00:46:52   and would destroy the entire industry

00:46:53   if it was allowed to play out.

00:46:55   So instead we'll just say,

00:46:57   "We giant companies agree to ignore the patents.

00:46:59   "We'll only use them to crush small companies."

00:47:02   America!

00:47:03   (laughing)

00:47:04   - Yay!

00:47:04   Oh, my word.

00:47:07   You are so right.

00:47:08   All right, we have semi-breaking news.

00:47:12   WWDC has been announced.

00:47:14   - Woo!

00:47:15   - It is going to be a week long, asterisk.

00:47:17   It is going to be, as we all foretold, June five through nine.

00:47:22   It will be in Cupertino, well sort of, but mostly online.

00:47:26   But in the same vein as last year, and again, as we're told,

00:47:30   there will be a special event the Monday, June 5th

00:47:35   at Apple Park where you can sign up to get,

00:47:39   I believe it's a free ticket if you leave aside

00:47:41   the fact you have to travel there,

00:47:43   that you can request to attend.

00:47:45   They will accept requests until April 4th

00:47:47   at nine o'clock in the morning Pacific

00:47:49   or noon Eastern time.

00:47:52   And they're gonna do something presumably similar

00:47:54   to what they did last year.

00:47:55   And I put my name in the hat, we'll see what happens.

00:47:59   I genuinely don't know whether the three of us

00:48:02   are gonna be there or not,

00:48:03   and we don't necessarily need to talk about that today.

00:48:05   But I'm hopeful that all three of us would be there

00:48:09   'cause I haven't seen you two at all,

00:48:10   literally not once, since WWDC 2019,

00:48:13   and that is too damn long.

00:48:14   So we'll figure that out amongst ourselves,

00:48:17   but I'm excited that there are dates.

00:48:20   I'm excited that there's a pretty good chance

00:48:22   that I think all three of us will be there.

00:48:23   So I'm just excited.

00:48:24   This is good news.

00:48:25   - You think there's a pretty good chance?

00:48:27   Well, so we just got done complaining about Apple in Syria

00:48:29   and we are about to complain even more about Apple,

00:48:31   but in between, this is like the reverse of a (beep)

00:48:34   sandwich.

00:48:35   (laughing)

00:48:37   In this scenario, the (beep) is the bread

00:48:40   instead of the meat.

00:48:42   In the middle, what I'm gonna say is,

00:48:43   "Hey, Apple, send us press passes."

00:48:46   [LAUGHTER]

00:48:48   Because otherwise, it's a lottery.

00:48:49   Like, as Apple says, invitations will

00:48:51   be allocated by a random selection process.

00:48:53   And so we'll find out by April 5 whether we got the random.

00:48:56   I put my name in the hat as well.

00:48:57   But the odds aren't great, because it's not like WWDC.

00:49:00   It seems like it's a smaller number of people.

00:49:02   It's not 5,000 people they're getting invitations to,

00:49:04   I don't think.

00:49:04   No, it seems like it's exactly like last year, the one day

00:49:07   in person at Apple's campus with the in-person keynote

00:49:10   presentation, and then everything else is online,

00:49:13   basically.

00:49:13   The in-person thing, the old conference

00:49:16   when everyone was in the conference center

00:49:18   held about 5,000 attendees.

00:49:20   Last year at Apple it seemed to be about 1,000,

00:49:25   maybe 1,500, something in that ballpark.

00:49:28   I would expect about that same number this year.

00:49:31   Maybe a little bit more if they could fit,

00:49:33   it seemed like they might be able to fit a few more people,

00:49:34   maybe 2,000 at most, but that's probably

00:49:38   about as high as you could expect.

00:49:39   So it's gonna be same deal as last year basically.

00:49:43   Some people will get there for that one day thing.

00:49:46   If you can't get there, you don't really need to worry

00:49:49   about missing much of anything,

00:49:51   'cause all the content will be online,

00:49:53   which, and I'm very happy with that, honestly,

00:49:55   because I think this new format that,

00:49:58   COVID kind of forced them into this new format,

00:50:00   but we were kind of heading in this direction for a while,

00:50:03   and COVID just forced them to make it

00:50:05   the premium primary experience.

00:50:08   And it's so much better, honestly,

00:50:10   than the old conference sessions

00:50:12   that were performed in person,

00:50:14   just by the nature of what they can do with this new format,

00:50:16   like it's so much better as a developer resource.

00:50:21   And so I am very, very happy

00:50:22   that they are continuing to do this.

00:50:25   Also, you know, downtown San Jose,

00:50:27   most of the stores you like to close.

00:50:29   And they're worth that money to begin with.

00:50:31   - Yeah, there were only like three or four as it was,

00:50:33   and I guess most of them are not there,

00:50:35   or wildly changed.

00:50:36   Like the sausage place isn't a sausage place anymore.

00:50:39   - No, they brought the sausage back, I thought.

00:50:41   - Well, there's like a sausage.

00:50:42   - Yeah, it's more of like a burger place now.

00:50:45   At least the vegan Indian place is still there

00:50:46   and it's still amazing.

00:50:47   But at least it was last summer,

00:50:49   I don't know if it's still there now, I hope so.

00:50:51   I might take a dinner trip there this year if I can.

00:50:53   But anyway, I'm very happy they're doing this format again

00:50:57   because it worked really well last year.

00:50:59   And even though it's not the same as the old conference

00:51:01   in terms of like, there's way less reason

00:51:05   for a lot of people to gather there in person

00:51:07   and because it's held at Apple's campus,

00:51:10   which is not even itself, downtown San Jose,

00:51:12   it's close but it's not in downtown San Jose and there is almost nothing around

00:51:16   Apple's campus besides like houses and other office buildings. There's not

00:51:21   really like a downtown area to congregate there's only a couple a

00:51:24   handful of small hotels you know so there's not much of like a community

00:51:29   gathering really going on there so it's a much smaller event for the in-person

00:51:34   people. That being said as I said on Under the Radar this week if you have

00:51:39   the opportunity to go and if you can swing the cost and logistics of going I

00:51:45   would suggest it just because it is kind of a cool pilgrimage for Apple fans like

00:51:51   it's it's cool to go there it's it's amazing to actually walk into Apple Park

00:51:55   and to see the actual you know the big circle building to be in that you know

00:52:01   tremendous cafeteria you know auditorium atrium kind of thing like it's a it's an

00:52:06   an amazing experience to see this place.

00:52:08   It's a beautiful building, it's a beautiful environment

00:52:12   they've built around it, and it's just a really cool feeling

00:52:15   to be there with everybody, even though you're just

00:52:17   watching the video in all likelihood,

00:52:18   'cause that's what it was last year,

00:52:19   even if you're just watching a video

00:52:22   and sitting there getting a slow sunburn

00:52:23   'cause you forgot to put on the sunscreen,

00:52:24   they literally gave you in the bag,

00:52:26   please put on the sunscreen, it's in your bag,

00:52:28   it's in your goodie bag, just put it on.

00:52:30   But anyway, it's cool to be there with everybody,

00:52:33   it's cool to be in the crowd as everyone

00:52:36   of seeing stuff for the first time

00:52:37   and you get to feel the crowd reaction

00:52:40   like being at a live event, you know, 'cause it is one.

00:52:42   I also find it very helpful as a developer,

00:52:46   I feel like it actually motivates me a lot.

00:52:48   Like when I'm there, in that environment,

00:52:52   it's like a theme park or a pilgrimage,

00:52:55   as I said earlier, for Apple stuff,

00:52:56   and so that actually really motivates me

00:52:59   to come home and work really hard

00:53:01   on all the stuff they just announced.

00:53:03   And throughout the rest of the summer,

00:53:06   the excitement fades.

00:53:07   You actually get the beta, you realize,

00:53:10   wow, the stuff they announced really doesn't work yet.

00:53:13   Or it doesn't do what I hoped it would do,

00:53:15   or it's missing some functionality

00:53:17   that I hope it will have later.

00:53:19   Or wow, this is a real pain in the butt

00:53:21   having to deal with all these deprecations

00:53:22   that just happened, and wow,

00:53:24   and now I have to change this whole API

00:53:26   I've been using back here for the last 10 years

00:53:27   'cause they just changed it or killed it or whatever.

00:53:30   And so there's all this kind of grind

00:53:32   or pain in the butt stuff

00:53:34   that you have to deal with later in the summer,

00:53:35   but at that point, at the very first day

00:53:37   that everything's unveiled, when you're there that week

00:53:40   or that day, it's all fun.

00:53:43   It's all like, wow, look at this, it's so amazing,

00:53:45   and everyone's excited, and everything's positive

00:53:47   'cause no one's found all the crap yet,

00:53:49   and it's just a really nice experience to be there,

00:53:51   and it is very motivating as a developer to go there.

00:53:53   So I do strongly recommend, I mean, look,

00:53:57   if you don't live anywhere near California,

00:53:59   that's gonna be a lot of expense and time,

00:54:01   and so it's probably not worth it to you

00:54:02   for objective reasons.

00:54:05   But if you win the ticket lottery thing,

00:54:08   and if you get the opportunity to go,

00:54:09   and if you can handle the cost and time to get there,

00:54:13   it's a fun pilgrimage and a fun event.

00:54:15   It's not anything that you can put a monetary value on,

00:54:20   it's just fun and motivating.

00:54:22   And it's really great for that,

00:54:24   and so it's recommended if you can swing it.

00:54:26   And I'm gonna do my best to be there.

00:54:28   - Yeah, the only reason I've been considering

00:54:30   going this year is just to have the Apple Park experience,

00:54:34   that everyone had the other year.

00:54:36   'Cause I still don't really want to be traveling.

00:54:38   I don't relish being on a plane,

00:54:40   breathing other people's air.

00:54:41   I don't relish doing all that stuff.

00:54:43   I don't relish getting COVID again.

00:54:45   Many, many reasons that I would be very unlikely to go.

00:54:48   But considering I saw everybody go all last year

00:54:50   and how much fun they had,

00:54:51   I think it's worth it for me to do as an experience.

00:54:53   'Cause who knows how many more times they'll do it

00:54:55   in this exact format and who knows how long Apple Park

00:54:58   will be the way it is.

00:54:59   So I want to visit Apple Park.

00:55:01   across I'll be, I fully expect that if I do end up going, I will be very annoyed by the

00:55:06   fact that apparently you're not allowed to bring real cameras into Apple Park and I'll

00:55:09   have to take pictures of my iPhone the whole time I'm there.

00:55:12   But you know, what can you do?

00:55:13   I'm so sorry.

00:55:14   You just think of how good those pictures would be.

00:55:16   So many beautiful things and I could take cool pictures of people and crowds but nope,

00:55:19   not allowed, just iPhone only.

00:55:21   Does it actually say that or is it one of those things where it's like no detachable

00:55:24   lenses?

00:55:25   Like there's a couple of...

00:55:26   Yeah, it's basically, yeah, it's basically no real cameras.

00:55:28   I know that you can use iPhones to take pictures

00:55:31   within limited context, but I think the thing last year

00:55:33   was no, maybe they did say it was no

00:55:36   interchangeable lens cameras, but I feel like

00:55:37   they were just basically like, no cameras except iPhones.

00:55:41   Or no cameras except phones, I guess.

00:55:44   - Yeah. - Is probably what they mean.

00:55:45   - I don't remember that, I mean, I believe you.

00:55:47   I just, I do not remember that being a thing.

00:55:48   - Maybe I'm misremembering, someone from Apple can tell me.

00:55:50   But anyway, this all--

00:55:51   - No, I remember it being kind of vague,

00:55:52   like, because what they basically intend for the rule to be

00:55:57   is no professional photography.

00:55:59   But that's hard to codify,

00:56:01   and so usually it ends up being like,

00:56:02   no professional cameras,

00:56:04   and that's also hard to codify.

00:56:07   - I think they also don't want you to have zoom lenses,

00:56:09   because all the walls are glass,

00:56:11   and if you have like a big zoom lens,

00:56:13   you could like read things off whiteboards that are, you know.

00:56:16   - I wonder like, you know,

00:56:17   if the rule ends up being like no detachable lenses,

00:56:19   like well, could you bring in like the Nikon P1000,

00:56:22   and like, it's basically a telescope.

00:56:25   - Yeah, yeah, I know.

00:56:26   - You bring in a super zoom with an 800 millimeter lens,

00:56:28   it's like, hey, it doesn't detach.

00:56:29   - Oh, it's a 3000 millimeter lens,

00:56:32   but it doesn't detach.

00:56:33   - Just bring a Samsung phone,

00:56:35   it'll just make up things on Apple's whiteboards

00:56:37   that seem plausible.

00:56:38   - Yeah.

00:56:39   - Or just bring a drone, fly it right up to the edge.

00:56:40   I'm sure they won't mind that or notice that at all.

00:56:42   - Oh, I'm sure they would love that.

00:56:44   - Did you go as press last year, Marco?

00:56:46   - Yes. - Just regular, yeah.

00:56:47   So I think if you go as press,

00:56:49   you get to do and see more stuff.

00:56:50   Didn't they have like a day before thing

00:56:52   where they showed you the developer center

00:56:53   and stuff like that?

00:56:54   - Yeah, there was a developer center tour,

00:56:56   they were doing various groups for that,

00:56:58   and there was a hands-on area after the keynote

00:57:01   where we got to see the new MacBook Air,

00:57:02   and I got to be in the way of Johnny Ives' shot again,

00:57:04   or Tim Cook's shot, sorry.

00:57:06   - That's another reason that Apple

00:57:07   should give us press passes.

00:57:08   - Yeah, 'cause then all three of us could be in the way

00:57:10   for Tim Cook trying to handle the new product

00:57:12   for the camera shots.

00:57:13   - Yeah, we'll be too busy talking about all the stuff

00:57:16   we did in the hands-on room, we won't even have time

00:57:18   to complain about Apple like we're about to do.

00:57:20   (laughing)

00:57:21   - Yes, that's it.

00:57:21   - I mean, it actually is true,

00:57:23   Like that actually happens.

00:57:24   (laughs)

00:57:25   - Oh my word.

00:57:26   I don't know, I mean WWDC,

00:57:27   I granted I haven't been since 2019

00:57:28   and I miss the event.

00:57:31   Golly, do I miss seeing you two

00:57:32   and all of our other mutual friends.

00:57:35   But WWDC, at least the way I remember it,

00:57:38   and granted it is different now,

00:57:39   but it is exactly what Marco was describing.

00:57:41   It's just a really great event

00:57:43   to get you really excited about your work

00:57:46   or about Apple, even if it's not your work,

00:57:48   or if you hope for it to one day become your work.

00:57:51   It's just it's super fun and and I really hope that all three of us end up there and you know

00:57:57   Whether we get press passes or just get very lucky with the lottery or just choose to go

00:58:01   Because we haven't seen each other in a long friggin time one way or another

00:58:05   I hope that I hope that it works out, but I don't know there's that much more to say about this now

00:58:11   But I am I'm excited that they've announced it

00:58:15   I'm excited that I have a specific thing to look forward to

00:58:18   I'm excited that if I fly on Sunday,

00:58:21   apparently Richmond has some airline,

00:58:23   I forget, like Breeze or something,

00:58:24   that does direct from Richmond, Virginia to SFO,

00:58:27   which is stunning.

00:58:28   - Wow.

00:58:29   - Unfortunately, I think the return trip

00:58:30   is like every other day or something like that.

00:58:33   It's not the day I would want to take back,

00:58:35   but that's neither here nor there.

00:58:38   Anyway, I'm just excited,

00:58:39   and I really hope it works out for the three of us.

00:58:40   So anything else about WWDC?

00:58:42   We'll do more about things we expect to see

00:58:45   when we get closer to time.

00:58:47   We don't have a merchandise story for ATP for WWDC yet.

00:58:50   We are working on that, no promises,

00:58:52   but anything else about Dub-Dub?

00:58:55   - I would just say it's almost,

00:58:57   I know this is rich coming from me,

00:59:00   it's almost like seeing a live sports event

00:59:02   versus watching it on TV.

00:59:04   - That is rich coming from you.

00:59:05   - Yeah, I can't even tell you which sport.

00:59:08   (laughing)

00:59:09   But no, the feeling of being there,

00:59:12   of actually seeing the building

00:59:15   and being right there with everyone else

00:59:17   and everyone's cheering together

00:59:19   when something good is announced.

00:59:21   The feeling of being there is really cool

00:59:24   and really energizing and it's a great experience.

00:59:26   And again, and what you're paying for

00:59:29   with all the travel logistics and everything,

00:59:30   what you're paying for is that experience.

00:59:33   You're not paying for the developer content, that's free,

00:59:36   but you're paying for the coolness

00:59:39   of seeing it in person live.

00:59:41   That I think is the main selling point.

00:59:44   So that's why if you get in,

00:59:46   and if you can swing the travel logistics,

00:59:48   it's pretty cool and you should do it.

00:59:50   You don't have to do it every year,

00:59:51   but you should do it at least once.

00:59:52   - Yeah, I agree.

00:59:53   And actually, I should point out,

00:59:55   as part of the meat in this poo-poo sandwich,

00:59:58   every Apple employee that I've met is awesome.

01:00:03   Individually, Apple people are super great.

01:00:05   We may strongly disagree with the decisions

01:00:09   and policies of the organization,

01:00:11   but pretty much all the rank,

01:00:13   especially the rank and file people, are pretty awesome.

01:00:16   And I can't think of any examples that disprove the rule.

01:00:19   Like pretty much everyone I've spoken to,

01:00:21   both people that I kinda know, people that I do know,

01:00:24   and even just strangers that I find out,

01:00:25   oh, you work at Apple, they're all super chill

01:00:27   and super cool, and that doesn't necessarily mean

01:00:30   you'd meet any while you're there,

01:00:32   but just being in the proximity gives you a chance.

01:00:34   So we'll see what happens.

01:00:36   - I mean, frankly, I know we're not doing

01:00:39   a hiring ad for them, really, but of all the

01:00:42   kind of Bay Area tech company or any all the big tech company I guess candidates

01:00:48   that one could possibly go to work for over on the West Coast somewhere I would

01:00:53   go to Apple before going to any anyone else like it see it's such a cool place

01:00:57   and it attracts really good people for a reason it's it's a really great place to

01:01:03   work and and it's a really cool thing to see it's a really great team that you

01:01:07   that you work with there.

01:01:08   And so when you go there, you see,

01:01:12   again, you won't, as Keith said,

01:01:15   you're not walking through the offices or anything.

01:01:18   You don't see a single desk from somebody

01:01:21   that somebody's working at.

01:01:22   You're escorted in with event staff to the lunch area,

01:01:27   and you sit in these chairs that are set up and everything.

01:01:31   You're not walking through the design lab or anything.

01:01:33   But any person you run into there,

01:01:36   'cause you will see a lot of Apple employees there.

01:01:38   And anybody you can talk to, talk to.

01:01:42   Because it's a company that attracts really good people

01:01:45   for lots of good reasons.

01:01:46   And yeah, it's, again, it's a heck of an event.

01:01:51   I strongly encourage you to go if you can.

01:01:54   I understand if you can't because it is a huge expense

01:01:56   to get most people to California

01:01:58   and to stay in a hotel for a few days or whatever.

01:02:01   I get that, but if you can do it, it's really cool.

01:02:06   Consider becoming an ATP member.

01:02:08   Members get all sorts of fun little goodies,

01:02:11   number one of which is an ad-free version of the show.

01:02:14   You get your own private feed

01:02:15   that you can add to any podcast app you want,

01:02:17   and it's an ad-free version of our show.

01:02:20   You also get access to the bootleg feed if you'd like it.

01:02:22   The bootleg is our unedited live broadcast.

01:02:26   It's released right after we finish the show recording,

01:02:28   so it's usually the night before the main show comes out

01:02:31   in the regular feed, so you get a faster release,

01:02:34   and it contains all of our beginning and ending

01:02:38   kind of small talk stuff, the title selection process,

01:02:40   a few little bonus things here and there,

01:02:42   as well as any kind of mistimed jokes

01:02:45   that I would have fixed in the edit.

01:02:47   You get to hear me do the bell live,

01:02:49   that's not a sound effect I drop in,

01:02:50   I actually do it live.

01:02:51   You get to hear Casey swearing,

01:02:53   anything I would bleep out in the edited show

01:02:54   that comes right through in the bootleg, it's great.

01:02:56   And we also occasionally do member exclusive content.

01:03:00   That's what we have.

01:03:01   Just yesterday we released member exclusive content.

01:03:04   It is a movie club episode on John's beloved Miyazaki movie,

01:03:09   My Neighbor Totoro.

01:03:11   It's so fun, you gotta listen to it.

01:03:13   Member exclusive content is occasionally a fun thing

01:03:15   we do there, but all of this comes to you

01:03:18   for just eight bucks a month.

01:03:19   Mainly you get the ad-free feed and the bootleg feed.

01:03:21   Those are the big things, and then the occasional

01:03:23   exclusive content is kind of a fun little bonus.

01:03:26   Join us today at ATP.FM/join.

01:03:29   Again, eight bucks a month.

01:03:31   We have different currencies, annual plan if you want it,

01:03:34   But that's the gist of it.

01:03:34   Eight bucks a month, you get our ad-free feed,

01:03:36   our bootleg feed, and our occasional exclusive content.

01:03:39   It is great.

01:03:39   See for yourself, ATP.fm/join.

01:03:42   Thank you so much for listening,

01:03:43   and please consider becoming a member.

01:03:45   Thanks.

01:03:45   (upbeat music)

01:03:49   - All right, so if you are an Apple person,

01:03:52   I need you to tune out and come back to this episode

01:03:55   after WWDC, because,

01:03:58   it's that time again, fellas.

01:04:01   Let's have a chat about radar,

01:04:03   because it's time.

01:04:04   Just to very, very briefly recap,

01:04:08   Radar is Apple's internal tool.

01:04:09   It used to be also to some degree externally visible.

01:04:12   Now it's been replaced from an external perspective

01:04:14   by Feedback Assistant.

01:04:16   Radar is their internal tool.

01:04:17   It's their bug tracker or issue tracker, what have you.

01:04:20   It is basically the only way that we have

01:04:25   as external people to communicate with Apple

01:04:27   is to just throw a radar or really a feedback.

01:04:30   We'll probably use the terms interchangeably.

01:04:32   throw feedback over the wall,

01:04:33   where I guess it gets internally turned into a radar,

01:04:36   and just hope that a human eventually looks at it

01:04:38   and does something with it.

01:04:40   And it is a deeply hostile and awful approach

01:04:45   to developer relations.

01:04:51   Because, you know, I can only speak for myself,

01:04:53   I presume that Marco, you probably have,

01:04:55   and you don't have to say one way or the other,

01:04:57   you probably have at least a couple of contacts

01:04:59   in developer relations,

01:05:00   because you are on the bigger side,

01:05:02   especially as an indie person,

01:05:04   in terms of your reach and your company size and so on.

01:05:07   But I don't have any-- - You'd be surprised.

01:05:09   - Oh, okay, fair enough.

01:05:11   I can tell you, I certainly don't have any sort of contact

01:05:14   in Developer Relations that if I have a question,

01:05:15   I can ask them.

01:05:16   I have a bunch of contacts that I've made

01:05:19   completely personally that are outside Developer Relations,

01:05:21   just rank and file engineers, that are friends of mine

01:05:24   that I've made friends with,

01:05:26   not because I'm trying to use them

01:05:27   for any particular reason,

01:05:28   just because they're good people,

01:05:29   like we were talking about, and I enjoy them,

01:05:31   and hopefully they enjoy me too.

01:05:33   But I don't have any formal contacts within Apple

01:05:35   for any sort of bugs or questions or anything like that.

01:05:39   And if I have a problem, the easiest way,

01:05:42   easiest and best way for me to get an answer

01:05:44   is to tweet/tute about it, and most times I'll get

01:05:47   something that'll either push me in the right direction

01:05:49   or maybe even solve my problem.

01:05:50   But if I have a problem, if I have a demonstrated problem

01:05:54   with an Apple API, I can put together a feedback

01:05:59   which will almost certainly not get looked at.

01:06:02   If it is, I will get asked for sample code, okay fine,

01:06:06   which sometimes though takes hours to put together.

01:06:10   This is uncompensated time by the way.

01:06:12   I don't know what an iOS developer contracts for,

01:06:14   but years ago it was like $150 an hour.

01:06:16   I'm sure it's like up around $175, $200 by now.

01:06:19   So many hours of work to create a sample project for them,

01:06:24   which I understand why they ask for it,

01:06:25   but we're talking about an effective investment

01:06:28   if I was going to spend that time contracting myself out,

01:06:31   of like $1,000 of my time that could be used

01:06:35   putting together a sample project

01:06:36   for them to promptly ignore.

01:06:38   You oftentimes get asked for a cyst diagnosis,

01:06:41   which is basically like a bunch of diagnostic information.

01:06:44   Again, in and of itself, that's fair,

01:06:45   but 90% of the time it's not even necessary or useful.

01:06:50   It's just frickin' broken.

01:06:52   And you had an experience recently,

01:06:54   and I think you were talking about it here on ATP,

01:06:56   I don't think it was under the radar,

01:06:57   where you had said, "Oh, I filed a radar,

01:07:00   "and it's something that's really broken with audio stuff,

01:07:03   "if memory serves."

01:07:04   And you sent that across the wire

01:07:07   and threw it over the fence, and you didn't hear squat.

01:07:10   And this is kind of a big deal, particularly for you,

01:07:13   but arguably for Apple in general.

01:07:15   In fact, why don't I let you interrupt me,

01:07:17   and can you remind me what that bug was, if you recall?

01:07:20   - Yeah, I'll dig up the number and put it in the show notes,

01:07:22   but effectively, it was the 16.4 betas

01:07:26   were having the audio services were reset notification,

01:07:31   which I think is a background demon crash.

01:07:33   They were having that crash and reset audio services a lot

01:07:37   whenever I would begin playback in Overcast.

01:07:40   It was quite a common thing, way more than ever before.

01:07:45   It was causing problems like you'd hit play on 16.4 betas

01:07:48   and just it wouldn't play.

01:07:50   And then if you go and check the Overcast log,

01:07:51   you'd see, oh, this thing crashed like three times in a row.

01:07:54   - Cool.

01:07:55   And yeah, so it was a big problem with that

01:07:59   and I attached SysDiagnosis and the reproduction step

01:08:04   to get it which was pretty easy,

01:08:05   it's like play something on Overcast

01:08:06   and you see this in the log.

01:08:07   (laughing)

01:08:08   And I did everything I was supposed to do

01:08:11   and as far as I know that bug is still open.

01:08:14   I never heard anything.

01:08:17   I think it might be fixed in the last couple of betas

01:08:21   and now 16.4 is now out just like as of this week

01:08:24   And I think it is fixed in the release version of 16.4.

01:08:28   But it was not fixed even as recently as like two weeks ago.

01:08:31   And so I was getting a little nervous

01:08:33   it was gonna get shipped to everybody.

01:08:35   But I think it's fixed now.

01:08:36   But the bug is still sitting there, open.

01:08:38   I never heard a thing about it.

01:08:40   - Right, so here's the thing is that I was talking

01:08:44   with a friend at Apple and asked, just out of curiosity,

01:08:48   hey, can you look at feedback, blah, blah, blah, blah,

01:08:50   we'll put it in the show notes.

01:08:52   I don't need to know specifics,

01:08:53   but just, has a human being looked at this feedback,

01:08:57   yes or no?

01:08:58   And it turns out that yes, a human had looked at it.

01:09:02   And in fact, apparently there was internal activity on it

01:09:05   within 24 hours of you filing it,

01:09:08   which no bull (beep)

01:09:09   - No one told me.

01:09:10   - That is incredible.

01:09:11   That is exactly what we wanna see.

01:09:13   I am not lying.

01:09:14   I'm not being facetious.

01:09:15   Truly, that is exactly what you wanna see.

01:09:18   - That's, yeah, that's amazing.

01:09:20   - But the problem is what you just said, Marco,

01:09:23   Did you know this?

01:09:24   Were you aware of this?

01:09:25   - No.

01:09:26   - No.

01:09:26   What did you get back from Apple Marco?

01:09:27   Would you remind me real quick?

01:09:29   - Nothing.

01:09:30   - Exactly.

01:09:31   - I just loaded up the bug now in feedback system.

01:09:33   It says recent similar reports, none,

01:09:34   resolution open, no comments.

01:09:37   - It's not really tenable.

01:09:40   It's not really fair maybe.

01:09:41   I know I say that a lot.

01:09:43   I'm trying to get better about it,

01:09:44   but it's just, it ain't right.

01:09:46   And it's just, that's not a way for a company

01:09:50   who allegedly cares about developers?

01:09:53   And what is the triad?

01:09:55   It's Apple first, then users,

01:09:56   and then somewhere below that is developers.

01:09:58   You know, it's--

01:09:59   - It's not an equilateral triangle either.

01:10:01   - Well, that's also fair.

01:10:02   It's, if Apple cares about developers at all,

01:10:05   can we have some sort of communication the other direction?

01:10:09   And it's just, I'm not even getting into documentation

01:10:14   and all the other problems with Apple's,

01:10:15   you know, the whole developer story.

01:10:17   But this is not okay, that this is something

01:10:22   that could be a really friggin' big deal for Overcast,

01:10:27   like a colossally big deal.

01:10:28   - Yeah, I was really worried about this one.

01:10:30   - Right, and what are you supposed to do about this?

01:10:33   And yes, okay, let's leave aside,

01:10:36   what is it, the something incident,

01:10:38   what's the formal name for it, I always forget.

01:10:40   The DTS or something like that, what am I thinking of?

01:10:43   - Yeah, the DTS tickets, so yeah,

01:10:44   so they had, or DTS incident, whatever they call it,

01:10:47   - Yeah, so with your developer membership,

01:10:50   the $100 a year developer membership,

01:10:52   comes with two DTS, developer tech support,

01:10:56   instances, or tickets or whatever.

01:10:58   You can raise one of these two tickets a year,

01:11:00   and they don't build up,

01:11:01   like if you don't use them within the year,

01:11:03   they just expire.

01:11:04   If you run out, you can buy one for 50 bucks each, I think.

01:11:07   You can buy extra ones.

01:11:08   This actually, it gets you an actual DTS engineer

01:11:12   to look at your problem,

01:11:13   and they literally provide code-level support.

01:11:16   You can include code and they will look at it

01:11:19   and they will try it and they will try to figure out

01:11:20   what the problem is.

01:11:21   And I think, I don't know, this has probably changed

01:11:24   over the years, but I think there's some exception

01:11:26   where like if you have stumbled upon an actual bug

01:11:28   that's their fault, then they don't charge you

01:11:30   for the ticket or something like that.

01:11:32   The problem with this system, first of all,

01:11:33   is that most developers never use it

01:11:34   'cause we don't even know about it.

01:11:36   I've known about this system for years.

01:11:37   I've been an Apple developer for, oh my god,

01:11:40   how many developer members have I bought?

01:11:41   12, at least? (laughing)

01:11:43   15, whatever it's been.

01:11:44   yeah, like 15 years, something like that.

01:11:46   I've never used one.

01:11:48   Mostly because when I first heard about them,

01:11:51   and I heard that you only get two a year,

01:11:54   I thought, well, I better save that up

01:11:55   for when I really need one.

01:11:57   - Exactly.

01:11:58   - Even after I learned that you can buy one for 50 bucks

01:12:00   if you really need to, like if you run out,

01:12:02   even after that, I'm like, I still consider it like,

01:12:05   this like, you know, only in an emergency

01:12:07   would I ever use this kind of thing.

01:12:09   And so I never even think to do it.

01:12:12   I forget about it all the time.

01:12:13   I never consider it as an option.

01:12:16   And I probably shouldn't, I really should just use it,

01:12:18   but because it's so limited,

01:12:20   I completely forget that it's an option at all.

01:12:24   - Yeah, same, and I mean, some of this is on us, to be fair.

01:12:27   We should be employing, and I'm looking at Marco,

01:12:30   I'm looking in the mirror, we should be employing these

01:12:33   and seeing if it's any better.

01:12:34   But it's just so frustrating,

01:12:37   especially in the cases where one puts together,

01:12:40   and I don't know if this was the case

01:12:42   with Marko's most recent one,

01:12:43   but when one puts together a sample project,

01:12:45   and I'm about to give Jon the floor,

01:12:46   and you did this, Jon.

01:12:48   When you put together a sample project,

01:12:49   and you explain exactly what's going on,

01:12:51   here's a very simple sample project

01:12:53   that demonstrates the problem,

01:12:54   and you throw that over the wall,

01:12:57   and then crickets.

01:12:58   And crickets, and crickets.

01:12:59   And it's just, and I understand,

01:13:01   in the defense of Apple,

01:13:02   I understand that they get just an inconceivable

01:13:06   amount of issues.

01:13:07   I get that.

01:13:09   But what I also get is that the current system does not work.

01:13:15   It doesn't work for external people.

01:13:17   And from everything I've heard from the internal people, it doesn't f*cking work for them either.

01:13:23   So who is this in service of?

01:13:25   Yes, I know that Apple is a big company.

01:13:27   Yes, I know that radars go back to like literally the early 90s.

01:13:30   I get that.

01:13:32   But at some point, what is this in service of?

01:13:34   And don't even get me f*cking started about the fact that the way in which you say that you really care about something

01:13:41   is duplicating a radar.

01:13:43   It's just, "Oh, if Marco and I agree that this is a problem, well then both of you file it and that's your de facto way of voting."

01:13:49   Are you kidding me with this?

01:13:53   Just no! That is not okay!

01:13:56   That is not a mechanism by which you vote.

01:13:59   It's by throwing a radar across the wall,

01:14:01   then inevitably we'll come back with an "F you, give me a cyst diagnosed anyway."

01:14:04   This is just, this is not okay. And the fact that this is still a thing

01:14:09   blows my mind. And yes, Feedback Assistant, the app is a lot better than Radar, the web app was,

01:14:17   but it doesn't matter. It's still, the whole friggin' system is broken. It's awful. And I'm

01:14:24   gonna, I'm gonna really lose my cool, believe it or not, I haven't yet. So instead I'm gonna say,

01:14:27   say, "John, tell me your recent story about your radars and how swimmingly they went."

01:14:31   So you two are gonna make me be the big company representative again because I

01:14:37   have spent more time in big companies than both of you. You're the closest we

01:14:41   got right now. Yeah, that's true. I've spent some time, don't get me wrong, but you have certainly

01:14:45   spent a lot more. So I do understand a lot of how things work here. A couple of

01:14:50   reactions to things you two have said. When, you know, like the time we spend to

01:14:56   file a feedback, right? I would imagine that in a lot of cases the vast majority

01:15:02   of that time, at least I would hope the vast majority of that time, is the time

01:15:07   figuring out whose bug it is. Because, you know, all our programs are filled with

01:15:11   bugs, right? And we have to figure out why doesn't this work. And that takes a long

01:15:15   time depending on how thorny the bug is. But the only the only point where we're

01:15:20   gonna hopefully file something with Apple is we've done all the debugging

01:15:24   and we have determined to the best of our ability that no, I am using that API right,

01:15:29   it just doesn't work. Right? And it's hard to determine that because applications are complicated,

01:15:35   APIs are complicated, we don't have the source code to the frameworks, which also complicates

01:15:39   things. The thing that I'm used to as a web developer having the source code to all the third

01:15:42   party code really helps you determine is this my bug? Am I doing something wrong? Or is this a bug

01:15:48   with, you know, the third party thing? And it's kind of like the beginning programmer thing saying,

01:15:53   "I think I found a bug in the compiler."

01:15:54   You've almost never found a bug in the compiler.

01:15:56   (laughter)

01:15:57   It's a typical beginner programmer thing

01:15:59   because your program does something unexpected

01:16:02   and you think you understand how it should work,

01:16:03   but you really don't, right?

01:16:05   And that's how you learn and grow as a programmer.

01:16:07   So, I think, in my experience,

01:16:10   getting to the point where you're ready to file a feedback,

01:16:14   that's on us, not on Apple.

01:16:16   - Yeah, no, I agree with that.

01:16:17   - You can't charge Apple for the time

01:16:19   you spend debugging your program, right?

01:16:20   - Totally.

01:16:21   And it's hard to get to that point.

01:16:23   And that's why it feels frustrating

01:16:24   because you spend hours, days, weeks, however long,

01:16:28   trying to figure something out,

01:16:29   and eventually you figure it out,

01:16:31   I think this is not even my fault.

01:16:32   And it feels like such an injustice,

01:16:34   because normally it's your fault.

01:16:35   Like 99.9% of the time it's your fault, right?

01:16:37   You did something dumb in your program.

01:16:39   But you're like, no, I think this is a bug

01:16:41   in one of these frameworks, which does happen, right?

01:16:44   Then we go into the time,

01:16:45   if you are a conscientious bug reporter,

01:16:47   which I tried to be when I actually think

01:16:49   I have a legit bug and I'm not just complaining

01:16:50   about a feature suggestion or some crap like that,

01:16:52   which also goes into feedback,

01:16:54   which is why there's so many of them, right?

01:16:56   So I think I've got a legit issue.

01:16:58   I will go the extra mile to make a sample application.

01:17:02   This is after I've determined,

01:17:03   and probably part of the way I determine

01:17:06   that it's actually a bug in a framework

01:17:09   is by making that sample application.

01:17:10   So some of that time is attributable to me really

01:17:13   because you get frustrated, you're like, wait a second.

01:17:15   I think, forget about my app, forget about all my code,

01:17:19   forget about stuff I'm doing, fresh clean sheet of paper,

01:17:21   new project Xcode, let me see if I can reproduce this.

01:17:25   I got my real code here and I got my toy one

01:17:27   and they kind of like meet in the middle

01:17:28   until I can get the toy one to reproduce the problem

01:17:30   with the minimum number of code

01:17:31   and that's your sample project, right?

01:17:33   So some of that time building that is, you know,

01:17:36   is you figuring out where the bug is

01:17:38   because very often, as happens to me plenty,

01:17:40   I make the trivial sample project

01:17:42   or more often these days I use playgrounds

01:17:44   which has its own set of bugs, but anyway,

01:17:46   use playgrounds to do something

01:17:49   And if it works right in playgrounds

01:17:50   or it works right in your toy example

01:17:52   but it doesn't work right in your real app,

01:17:53   that's still probably your bug, right?

01:17:55   That happens all the time.

01:17:56   But sometimes it goes the other way.

01:17:57   So when it goes the other way, you're like,

01:17:59   all right, I'm gonna polish up the sample project,

01:18:01   I'm gonna file the bug.

01:18:02   At this point in your head, you understand the issue

01:18:06   because you have boiled it down, you have debugged it,

01:18:09   you have figured out.

01:18:11   When you do X and Y and Z, Q should happen,

01:18:13   P happens instead.

01:18:14   You already figured that out.

01:18:16   And you're like, I'm serving this up to you, Apple,

01:18:18   on a silver platter.

01:18:19   I've got a sample project,

01:18:21   the source code to the sample project

01:18:22   is on a public GitHub URL,

01:18:24   a zip of it is included,

01:18:26   there's a sys diagnose that the feedback app already ran,

01:18:29   I can describe in five sentences,

01:18:31   here's the problem, expected results, actual results,

01:18:33   it fits on a page, you could print this on an index card.

01:18:36   Here it is.

01:18:37   A bunch of things can happen at that point.

01:18:41   One is that you may be perhaps one of the best,

01:18:45   but one of a thousand people who sent that bug to Apple

01:18:47   they already know about it.

01:18:49   And we would hope in a sane system

01:18:50   that there would be some communication that says,

01:18:53   yeah, no, we know about that one.

01:18:54   It's been 50 people who have filed it, right?

01:18:55   - One would think.

01:18:57   - Right, yes, but anyway, that could happen.

01:18:59   The other thing that could happen

01:19:01   is that could just never get looked at

01:19:02   because it's the bottom of a big pile

01:19:04   and you would never know.

01:19:05   But usually, especially if you have a tech podcast

01:19:09   that Apple people listen to

01:19:09   or you're too devoted or whatever,

01:19:11   someone will listen to the program

01:19:13   and look at the bug or whatever

01:19:14   and then get to us your back channels

01:19:16   that X, Y, and Z is happening, right?

01:19:18   But there's a lot of these bugs and it's difficult

01:19:21   to get to all of them in a timely manner.

01:19:24   You can't expect, and honestly,

01:19:27   bugs that I file for my stupid apps,

01:19:29   Apple should not look at, for the purposes of my app,

01:19:33   because who cares about my app?

01:19:35   The only time they should look at them is because,

01:19:37   oh, all right, we don't care about John's apps,

01:19:41   because whatever, who cares, right?

01:19:42   But if this is actually a bug in our framework,

01:19:45   this could affect Photoshop,

01:19:47   or like a real app that people care about, right?

01:19:49   If it's a legit bug in the framework,

01:19:51   especially if it's a new bug,

01:19:52   it didn't exist in the last version of the US

01:19:54   and it does exist in this one,

01:19:55   it's a regression as they say.

01:19:57   That's worth lurking at, not because of my apps,

01:20:00   but because tons of apps use these frameworks.

01:20:02   I'm not using super obscure frameworks,

01:20:03   so if I have found a legit bug

01:20:06   or a legit change in behavior,

01:20:08   someone should look at that.

01:20:10   How do you determine that?

01:20:11   It's a big pile of bugs.

01:20:12   How many of them are people saying,

01:20:13   I think the color of this button should be purple,

01:20:15   and how many of them are a carefully reproduced bug

01:20:19   with a minimal sample project, right?

01:20:21   That's a legit bug that affects Microsoft's office.

01:20:23   You need people to sort through all those things,

01:20:26   triage them, and figure out which is which.

01:20:28   And Apple doesn't seem to be particularly good

01:20:30   at doing that either.

01:20:31   Forget about the communication part of it, right?

01:20:33   So the most recent one I had, and I do this,

01:20:35   I found a handful of bugs

01:20:37   and I made little sample projects.

01:20:38   The most recent one I found was interesting,

01:20:40   or the one I'm gonna talk about here,

01:20:42   was interesting in that it was a straight regression.

01:20:44   This is a thing that worked in my app in Monterey

01:20:47   and didn't work in Ventura, which right away makes me think

01:20:50   this might be some kind of behavior change.

01:20:52   And then I go look at the release notes,

01:20:53   'cause when the new versions of the OS come out,

01:20:55   they have like framework level release notes.

01:20:57   Hey, if you use this framework,

01:20:58   here's what's changed in this thing, right?

01:20:59   Because they change things.

01:21:00   So this used to do that, this is deprecated,

01:21:03   we added these function, we removed those,

01:21:05   we changed this, like nothing about this

01:21:08   in any of the release notes.

01:21:09   I look at the documentation in Ventura.

01:21:11   Documentation doesn't say anything about this.

01:21:13   This was, specifically this bug was like a,

01:21:16   there's a thing in AppKit that you can put on a view

01:21:19   that tracks when the cursor enters it,

01:21:21   and like tracks where the cursor is or whatever.

01:21:24   It's called NSTrackingArea,

01:21:26   and I use it to track when the cursor enters area.

01:21:29   I'm using it for its intended purpose, right?

01:21:32   In Monterey, it worked as I expected.

01:21:34   In Ventura, it would track the cursor normally,

01:21:36   except if you were dragging something.

01:21:40   And then if you were dragging something,

01:21:42   And it's tracking error would be like, I don't see any cursor.

01:21:44   I don't know what you're talking about.

01:21:45   It would not track at all.

01:21:46   And that broke a feature of my application,

01:21:48   because I needed to track it when things are being dragged.

01:21:51   And I was debugging it for a while.

01:21:52   I'm like, why is this not-- anyway, found the bug,

01:21:54   isolated it, made a minimal reproduction sample application.

01:21:59   You don't need to read anything.

01:22:00   I always put all the text in the app.

01:22:02   Like when you launch the app, it has text that says, here,

01:22:05   do this, do that.

01:22:07   I expect this to happen, and that happens.

01:22:09   You don't even need to read the readme.

01:22:11   Made the sample app, submitted it.

01:22:12   Again, the fact that my app broke in Ventura,

01:22:15   not a big deal.

01:22:16   But if this is a legit change in behavior

01:22:19   for NSTracking, that is a commonly used thing

01:22:22   in applications.

01:22:23   It is a, you know, it's from AppKit,

01:22:25   lots of applications use AppKit,

01:22:26   and that's a significant piece of functionality

01:22:29   that just doesn't work anymore.

01:22:31   And you know, I didn't get any response or whatever.

01:22:33   Eventually, I added a comment to the thing.

01:22:36   Someone did respond to it,

01:22:37   because I probably complained about it

01:22:39   on probably back then, Twitter or whatever.

01:22:41   And said, oh, you should, you know, what is,

01:22:44   they asked me what is your application or whatever,

01:22:46   and they said actually you should be using

01:22:48   the drag handling thing to handle drag,

01:22:50   so on and so forth, you know.

01:22:52   There was actually some feedback.

01:22:53   I'm like, okay, right, well, so I understand

01:22:56   that I could use the drag thing,

01:22:57   but like it worked in Monterey and doesn't work in mature.

01:22:59   Is NSTrackingArea just not gonna track drags anymore?

01:23:02   'Cause if that's the case, like,

01:23:04   I was trying to get like, was this an intentional change?

01:23:06   Or are you just telling me, hey, there's a bug,

01:23:08   but you can work around it in this way,

01:23:10   or are you telling me from now on,

01:23:11   NSTrackingArea will not do this, so just get used to it.

01:23:14   And if that's true, you should probably update

01:23:16   the documentation and release notes and say,

01:23:18   oh, by the way, - Oh, imagine that!

01:23:18   Imagine that! (laughs)

01:23:19   - From now on, NSTrackingArea won't do this or whatever.

01:23:22   So I added a comment to the task that said,

01:23:24   or to the feedback that said,

01:23:26   my application is whatever,

01:23:29   and I've since added a workaround

01:23:33   to using the drag handling to do this.

01:23:35   Like I said, if you're on Venturi,

01:23:37   the drag handling, you know, whatever.

01:23:39   And the next response I got in feedback was,

01:23:42   great, we've closed your bug.

01:23:44   (laughing)

01:23:46   As I said, I'll read you the text,

01:23:48   which is not great, sorry.

01:23:49   Let's see.

01:23:50   Thank you for letting us know

01:23:51   that your issue has been resolved.

01:23:53   You can close this feedback by selecting Close Feedback

01:23:55   by the actions button found above.

01:23:57   As you've indicated this issue was resolved,

01:23:58   this feedback will no longer be monitored

01:24:00   and incoming messages will not be reviewed.

01:24:02   - Cool. - Should you find the issue,

01:24:02   that the issue is still present,

01:24:04   please file a new feedback report.

01:24:06   So this is like double whammy, because one,

01:24:08   I wasn't saying it's resolved.

01:24:10   I was saying I found a workaround, which

01:24:11   programmers do all the time.

01:24:13   Like, oh, there's a bug or a change in behavior.

01:24:15   You can code around it or whatever.

01:24:18   And two, saying, oh, and by the way,

01:24:19   don't even bother responding to this,

01:24:21   because even though we can't close this issue for you

01:24:23   and we want you to close it yourself,

01:24:24   we're just never going to look at it again.

01:24:26   If you think this issue is still there, file a new bug.

01:24:28   So I did file a new bug and said,

01:24:31   I think it's still there, because basically, I

01:24:34   don't know what you're telling me here.

01:24:36   Again, my question was, is this intentional new behavior,

01:24:40   or is it a bug that you're going to eventually fix?

01:24:42   Not that I care that much, but it is basic functionality.

01:24:46   And I complained about it on Mastodon,

01:24:48   and Apple people saw it.

01:24:49   And so eventually, I got someone to respond to the task

01:24:52   and explain the situation.

01:24:53   But John, running to the press never helps.

01:24:55   Yeah, yeah.

01:24:56   Explain the situation in English, basically saying,

01:24:58   this is an intentional change.

01:25:00   We think this is the way it's supposed to work.

01:25:03   It's not going to work the other way.

01:25:05   yes, we know we haven't updated the documentation

01:25:07   or put anything in release notes,

01:25:08   and we've already filed separate internal radars

01:25:11   to deal with that or whatever.

01:25:14   Here's what I'm going to say about this particular

01:25:16   experience.

01:25:16   I'm not going to say, oh, why didn't my bug get fixed

01:25:19   and why did it take too long and all this other stuff.

01:25:20   Because honestly, who cares, right?

01:25:22   What I am going to say is that from a policy perspective,

01:25:26   one of the things that Apple should really work on--

01:25:30   setting aside all the things we already talked about,

01:25:32   like, oh, you know, be better, right?

01:25:36   When you get to a point where a human being has somehow

01:25:41   found their way to my feedback, whether it's

01:25:43   because I have a podcast and post it on Mastodon

01:25:47   or it's just random luck, at some point,

01:25:50   hey, you come up on the rotation.

01:25:52   Your feedback is being triaged.

01:25:54   A human being has now got 37 seconds

01:25:57   to look at your feedback and do something with it.

01:26:01   I think one of the worst things that they do now that they can fix without really spending

01:26:05   any more money or time or whatever is when a human does that, make sure you spend that

01:26:12   time doing something useful because that's not going to come back again.

01:26:17   You're going to have to wait.

01:26:18   Like the time gap between some of these things sometimes is like weeks or months between

01:26:21   any response, right?

01:26:23   So in that moment, when the human is looking at it for 37 seconds, please, human, spend

01:26:30   the extra five seconds to write a coherent sentence because we know no one is going to

01:26:35   look at that again for three months.

01:26:37   This is the one chance I get from my bug to have its time in the sunshine and all I wanted

01:26:42   as a human with reading comprehension skills to say to me, "This is intended behavior.

01:26:48   We're sorry that it's not documented but just FYI it's going to work this way from now on."

01:26:54   That's it.

01:26:55   And that wouldn't take any more time than the sentence that I just read you, or the

01:27:00   feedback that was in there.

01:27:01   It's actually a shorter sentence.

01:27:04   Spend the time that you have, which is small and not enough, spend that time wisely because

01:27:09   it wastes all of our time to get a response which we have all gotten, every developer

01:27:15   has gotten us, a response that makes you think the person writing it either A, isn't a person

01:27:20   and is a bot, or B, did not read anything in your feedback.

01:27:25   If a response like that is put in, ostensibly, I'm told these are all done by humans, if

01:27:30   a human spends their 37 seconds to give a response that makes the person who posted

01:27:34   it think that they didn't read the feedback, either A, they didn't read the feedback, which

01:27:39   is bad, or B, they wasted that time.

01:27:42   So my one, in this particular round of being mad about radar and feedback, my one plea

01:27:48   for Apple is, in the tiny slice of time, human's time that we get, in the too small slice of

01:27:54   human times will get, please let them do a reasonable job of feedback because it

01:28:01   will save all of us so much time. It wastes so much more time to have three

01:28:06   rounds of back and forth with a month between each round than to just have the

01:28:11   person write a coherent sentence on the first one that sounds like a human, not

01:28:15   like a PR machine, and that reflects the fact that they saw the bug.

01:28:21   I mean, that's what really bothered me about this.

01:28:23   Secondarily, if anyone is listening or whatever,

01:28:25   I think it is insane that NSTrackingArea,

01:28:28   which has an option called enabled during mouse drag,

01:28:32   now does not work during mouse drag.

01:28:34   And I await anxiously the updated documentation,

01:28:39   how are they gonna document

01:28:40   the enabled during mouse drag option?

01:28:42   They're gonna say, this used to enable during mouse drag,

01:28:45   but it totally doesn't anymore, sorry about that.

01:28:47   I think that's dumb.

01:28:48   But that's the type of thing,

01:28:49   like if it was an open source project,

01:28:51   I'd be in the issue arguing like,

01:28:53   I think this is a bad change in its tracking area,

01:28:56   you're breaking apps for no reason,

01:28:57   there's an option called enabled during mouse drag,

01:28:59   what the hell, right?

01:29:00   You can't have that argument with Apple

01:29:02   because it would be three months between replies,

01:29:04   and by the way, they already said

01:29:05   they're not monitoring this bug,

01:29:06   so you'd have to be refiling it every single time.

01:29:08   Like that's pointless, that's the time-wasting,

01:29:10   open-sourced arguing over bugs that Apple actually,

01:29:14   you know, has a happy accident

01:29:16   of their terrible feedback system avoids or whatever.

01:29:19   But setting that aside,

01:29:21   I'm willing to say, just tell me, Apple.

01:29:23   Just tell me this is the new way it is.

01:29:25   Communicate that successfully.

01:29:26   Communicate that you know that you haven't documented it.

01:29:29   And by the way, on this particular issue,

01:29:32   as far as I can tell, I mean, I don't know,

01:29:35   'cause obviously we don't know what happens on Apple

01:29:36   and we just talked about how there's no communication.

01:29:38   But it seems like what happened

01:29:41   is whatever team is responsible for NSTracking area

01:29:43   or that whole framework or whatever,

01:29:45   decided they were gonna make this change,

01:29:46   probably for some efficiency reason

01:29:48   or someone thought it would be a good idea.

01:29:50   This happens all the time.

01:29:51   We've decided that the behavior of this needs to change.

01:29:54   Even though it's going to be annoying for some people,

01:29:56   it'll have fewer bugs, it'll have better performance.

01:29:58   I get it, right?

01:29:59   So they thought they were going to do this,

01:30:01   setting aside the fact that there's an enable

01:30:02   during mousetrap auction, whatever.

01:30:03   But they were going to do this.

01:30:05   But it seemed like they were going to do it,

01:30:07   and they knew it might break some applications,

01:30:10   but they didn't want to mention it

01:30:12   and the release notes are documented

01:30:13   'cause they were like, fingers crossed,

01:30:15   maybe this won't break anyone's app

01:30:16   and I hope no one notices. - Shh, shh, shh.

01:30:17   Don't tell anyone, don't tell anyone.

01:30:18   - Yeah, don't tell anyone.

01:30:20   Which totally happens, you're like Apple,

01:30:22   the big company would never do that, trust me.

01:30:24   This absolutely happens, because the incentives

01:30:27   are structured for you not to go,

01:30:29   oh, this has to be in the release notes,

01:30:31   and then we have to make, you know,

01:30:32   we have some policy that was made 10 years ago

01:30:34   that says if you do any breaking change to AppKit,

01:30:36   you have to clear it with Microsoft explicitly

01:30:38   to make sure it doesn't break Office.

01:30:39   Can we just make the change and put it in the betas

01:30:41   and maybe just no one will notice?

01:30:43   And here I am with my dinky little app

01:30:45   that five people run saying, oh, excuse me,

01:30:47   I noticed an engineer, and they must be like,

01:30:51   damn it, somebody noticed, and it's an app

01:30:53   that doesn't matter, so who cares,

01:30:54   but it's a stupid guy with a podcast, right?

01:30:57   And I don't fault those engineers

01:31:00   for trying to slip on under, I have 100% done this

01:31:03   so many times, and I've been caught doing it,

01:31:05   and I've also snuck a bunch through,

01:31:06   and it feels really good when you see the shirt,

01:31:08   it's like, look, I'm just gonna change the behavior,

01:31:09   and I bet nothing's gonna break,

01:31:11   and if I just don't say anything, it'll be fine.

01:31:13   But I'm gonna say that's also probably

01:31:15   not the best way to do things.

01:31:15   So whatever incentives are structured to make engineers

01:31:18   want to sneak this out without putting any release notes,

01:31:22   change those incentives.

01:31:23   Because engineers should be incentivized

01:31:25   to do the right thing, which is,

01:31:26   hey, we're changing the behavior of this.

01:31:28   It's in the release notes, we updated the docs,

01:31:30   sorry if it breaks your app.

01:31:31   And if it breaks too many apps, roll it back,

01:31:33   but just wanted you to know this is a change.

01:31:34   'Cause that would have saved me a little bit

01:31:36   of my time on my app, and who knows,

01:31:37   maybe it did break Microsoft's office or something,

01:31:40   and there's some other feedback about that.

01:31:43   But yeah, this year's, this month's,

01:31:46   whatever advice for Apple is,

01:31:48   spend the tiny slice of time you have wisely

01:31:50   because it wastes most of our time if you don't.

01:31:52   And I don't think it actually requires any more time,

01:31:56   it just requires slightly differently spent time.

01:31:59   Above and beyond that,

01:32:00   not to go over all the things

01:32:01   we've talked about in the comments before,

01:32:03   but I do feel like there is a big difference

01:32:06   between a concise bug report with a minimal sample project

01:32:11   that's on GitHub from a known good developer

01:32:13   and a random feedback about the color of a button

01:32:15   from someone you've never heard of before.

01:32:17   And it's the same thing with the App Store

01:32:18   that is frustrating when a company like Panic,

01:32:21   that should be like AAA white glove gold tier status

01:32:24   in Apple Developer, can't get their game through

01:32:27   to the App Store.

01:32:29   Their famous well-reviewed game

01:32:30   that's on a million other platforms from the company Panic,

01:32:33   while scam apps sail through.

01:32:35   The goal of these faceless bureaucracies

01:32:39   is not to be a roulette wheel for outcomes.

01:32:43   There should be some, when you're triaging things,

01:32:46   you should take into account the quality of the report,

01:32:49   who it's coming from.

01:32:50   They certainly do for Adobe, Microsoft,

01:32:52   the big companies like that.

01:32:53   - Oh, funny how that is.

01:32:55   - But they should also triage based on like,

01:32:59   is there a sample project historically,

01:33:01   like give it, it's like Uber drivers and ratings,

01:33:03   give the developers a rating.

01:33:04   This person, Daniel Jalkut, five stars,

01:33:07   gives amazing bug reports, always very concise,

01:33:10   always technically accurate, hardly ever makes a mistake,

01:33:13   hardly ever blames the compiler,

01:33:14   comes with a sample project, is smart, is responsive.

01:33:17   Look at those first, you're triaging,

01:33:21   you can't look at them all, there's too many.

01:33:23   When you triage, bubble up the good ones,

01:33:25   have a system of rating them,

01:33:27   and then getting to Casey's point,

01:33:28   I have to file a second one, what can I do,

01:33:31   but public issue tracking systems have this solved,

01:33:34   we all know that, Apple's usual response is,

01:33:36   Well, we can't use public issue tracking

01:33:38   'cause Adobe sends the source code to Photoshop

01:33:40   or you can't have them up or whatever.

01:33:41   And the answer to that from us has always been

01:33:44   let people opt into it.

01:33:45   Let developers say, "I agree that anyone can see

01:33:49   "this bug report.

01:33:50   "You know what, it's on me.

01:33:51   "I'm saying market is public.

01:33:52   "I'm not gonna put Photoshop source code in here.

01:33:54   "What I'm gonna put is my little sample project that I made

01:33:56   "and I'm gonna say everybody can see this."

01:33:58   Because then they can send the link to all their friends

01:34:00   who will go to that URL and hit the me too vote up button

01:34:04   on the thing instead of having to file something themselves.

01:34:06   These are all eminently solvable problems

01:34:08   without hiring more people except for to improve the system,

01:34:11   which they already did once with the feedback thing,

01:34:13   without staffing up someone and so forth.

01:34:15   And then I guess the final thing I'll say is

01:34:18   App Review used to be, believe it or not,

01:34:20   way worse than it is in terms of

01:34:22   how long things would go through.

01:34:23   - Oh, so much worse.

01:34:25   - And then at some point, something happened inside Apple

01:34:28   and App Review got faster.

01:34:29   Did they do that by cutting corners

01:34:31   and getting worse quality?

01:34:32   Maybe, but the point is, it got faster.

01:34:35   - No, I think they actually did it by firing somebody.

01:34:37   - Yes, right, whatever had to happen.

01:34:40   I'm not saying there's easy solutions.

01:34:42   Oh, we'll just spend less time and do a worse job

01:34:44   and we can do more of them.

01:34:45   And I'm not even sure if that's the right term.

01:34:47   But the point is, it is possible

01:34:49   for big things to change inside Apple.

01:34:52   It's happened before with App Review.

01:34:54   It can happen with feedback and radar.

01:34:57   It is not a intractable problem.

01:34:59   Yes, they'll probably have to spend more money

01:35:01   and hire more staff or whatever.

01:35:02   Whatever it is that they did to make App Review

01:35:04   way, way, way faster, and I think we would all agree that that's a net win.

01:35:08   The quality still sucks, and they still do terrible things like rejecting Pank's games,

01:35:11   and all the horror stories.

01:35:13   That all is still there, but I feel like that's about been a constant, but the time has gone

01:35:17   way down.

01:35:18   So, if the time between me getting useless, incoherent feedback responses was three days

01:35:24   instead of three weeks or three months, I would feel a lot better about it.

01:35:27   So I feel like it is possible for Apple to change and improve feedback.

01:35:33   I don't know what has to happen for that to happen,

01:35:35   but I know it's possible.

01:35:37   I know you can do it at Apple, right?

01:35:39   There are basic suggestions that you ask any developer,

01:35:42   pull someone off, develop Apple, develop off the street

01:35:44   and say, "What can we do to improve feedback?"

01:35:45   They have seven ideas for you.

01:35:46   They're all good, just take them.

01:35:47   Like, they're so obvious.

01:35:49   Everybody knows what they are.

01:35:50   Doesn't mean they're easy.

01:35:51   Doesn't mean you can do them overnight.

01:35:53   Doesn't mean you don't have to hire new people.

01:35:54   Like, I understand, but this is an important part

01:35:57   of the company, kind of like App Review,

01:35:59   an important part of the developer ecosystem

01:36:01   that was really, really bad, making improvements,

01:36:04   and it pays dividends, please do this.

01:36:07   - You know, we love, the three of us in particular,

01:36:09   love to whine about, well, you know,

01:36:11   why isn't such and such better, why isn't it better?

01:36:13   And I think one of the things that we can do

01:36:17   is talk about, you know, well, how would this,

01:36:19   how could this become better?

01:36:21   And I think what you're saying, Jon, is excellent.

01:36:24   You know, if a reviewer's looking at it,

01:36:26   like, really properly look at it.

01:36:28   If it's more than just a, "This is broken,"

01:36:32   do something about that.

01:36:34   If there's a sample app, run the sample app.

01:36:37   Acknowledge that this reviewer or this submitter

01:36:40   has done good work.

01:36:41   And like you said, have a rating system

01:36:43   or something like that.

01:36:44   Oh, Daniel Jalkett always writes amazing bug reports.

01:36:47   We should pay attention to him.

01:36:48   Maybe even, and this is gonna be weird, Apple,

01:36:51   but maybe even reply to him.

01:36:54   Wouldn't that be amazing? - Whoa.

01:36:55   - I know, right?

01:36:57   This is some real new thought technology.

01:36:59   You know, Casey, all he does is fuss and moan.

01:37:02   He never includes a sample app.

01:37:03   That's not true for the sake of discussion.

01:37:05   He never includes a sample app.

01:37:06   Oh, he doesn't deserve a reply.

01:37:07   Well, you know what, okay, that's kinda deserved.

01:37:10   But another thing that I've been thinking about is

01:37:13   you're never going, Apple will never ever tell us

01:37:16   what they're doing internally,

01:37:17   which I don't think that has to be the case,

01:37:20   but you know, it's Apple, I understand it, that's fine.

01:37:22   But what if there was like some sort of indication

01:37:27   on the public facing feedback,

01:37:30   the last time that anyone within Apple has touched this,

01:37:33   we don't know what they've done,

01:37:35   we don't know if they just opened it

01:37:36   and closed it immediately,

01:37:37   but somebody touched this a week ago,

01:37:40   a day ago, two days ago,

01:37:42   and then fast forward a week,

01:37:44   and it says today,

01:37:46   then you don't have to tell me a damn thing.

01:37:49   I at least know, assuming this isn't just fakery,

01:37:52   I at least know that some friggin' human being

01:37:56   has looked at this in the last six months.

01:37:59   That would make me feel at least a little bit better,

01:38:03   probably even so much better,

01:38:05   because at least I know things are,

01:38:07   I don't know what's happening,

01:38:08   maybe I won't see the results of it for a year,

01:38:10   but at least I know somebody cares enough

01:38:12   to have looked at it in the last week.

01:38:13   And now of course I could argue on the flip side of that is,

01:38:16   well, most of these are probably gonna get looked at

01:38:17   once and ever again.

01:38:18   Well, we already know that.

01:38:20   You're not telling us anything we don't already know.

01:38:22   Like who cares put in writing maybe somebody who actually gives a up high will be able to do something about it

01:38:28   Like, ah, there's there's there's so many it's like John said there's so many ways to fix this and it's it's just

01:38:34   There's so many ways to fix this and if any one of them, you know

01:38:38   I would like I am in hell and I would love a cold glass of water, please anything

01:38:43   Anything Apple, please and my favorite by the way my favorite slap in the face

01:38:48   And I think John you just had this happen recently

01:38:51   Is oh here's something you reported years ago

01:38:56   Which kudos to Apple for keeping it around and having looked at it years later, but nevertheless

01:39:02   Here's something that happened years ago, and we think we fixed it. Hey you want to do me a solid and go check

01:39:08   Well okay fair enough, but John what happens if you don't check immediately well actually this one

01:39:13   It wasn't a years ago when it was a thing where they said and this is well

01:39:16   We'll link in the show notes to who's LapCat Software.

01:39:20   Who's that guy?

01:39:21   - Jeff Johnson, is that right?

01:39:22   - Yes, Jeff Johnson.

01:39:24   Had a similar complaint to mine, but yeah.

01:39:26   File the thing, they fixed it, or they said it was fixed,

01:39:29   and they do the thing which I think is good,

01:39:31   which is, hey, don't just close the bug when you fixed it.

01:39:34   Actually get the person who opened the bug

01:39:36   to agree with you, the yes, I agree, you have fixed it,

01:39:39   so we can close the report.

01:39:40   And that would be a good system

01:39:42   if we weren't sending messages to each other

01:39:44   by carrier pigeon, right?

01:39:46   because it takes three months for anything

01:39:47   to get back and forth.

01:39:48   So that kind of makes it dumb.

01:39:49   But anyway, I had filed something.

01:39:51   They said, we think this is fixed.

01:39:52   And Apple's thing that they've been doing lately,

01:39:54   as in the past few years, is they'll send you the "we think

01:39:58   it's fixed" message, incoherently written in a way

01:40:01   that you have to read seven times to figure out what the

01:40:03   heck they're even talking about.

01:40:04   As you'll see, like, we think this is fixed.

01:40:05   And they'll give a build number.

01:40:07   And you're like, wait, what OS is that?

01:40:08   What beta is that?

01:40:09   How do I get that?

01:40:10   How do I install that?

01:40:10   They'll tell you, basically, we think it's fixed in a beta.

01:40:13   And what they want you to do is, oh, just try your app

01:40:18   or your sample program in this beta of the operating system

01:40:22   and let us know whether it's fixed.

01:40:24   And one of mine was they sent this bug.

01:40:31   And it wasn't one of my good bug reports.

01:40:33   It was one of my bad ones.

01:40:34   It was from a user's perspective.

01:40:35   It was related to my programmer.

01:40:36   There was like, a thing is happening in system settings

01:40:39   that I don't think should be happening

01:40:40   based on what my program does.

01:40:42   And they're like, "Oh, you know, it wasn't a great bug report."

01:40:44   So they said, "We think we've got it fixed.

01:40:45   Can you confirm it in a beta?"

01:40:47   And my answer -- I didn't say anything,

01:40:48   but like my reading it was like,

01:40:51   "I don't have a good way to install a macOS beta.

01:40:53   I'm not going to install it on my main machine.

01:40:55   I don't have a drive available.

01:40:56   All my laptops have been lent out to, like, kids.

01:40:58   My son had to take two laptops to college for --"

01:41:01   So he had to run a VM on an Intel Mac,

01:41:04   but he has an ARM Mac. And anyway,

01:41:06   it wasn't convenient for me to install a beta anymore.

01:41:09   So I just had the Casey-style childlike satisfaction

01:41:13   of them saying, "Please let us know

01:41:14   if this is fixed in the last beta,"

01:41:15   and then I just ignored it.

01:41:17   (laughing)

01:41:18   I mean, 'cause like, you know, I could have responded

01:41:21   and said, "Oh, I don't have time to install a beta,"

01:41:22   but like, bottom line is, "Okay, how about I let your thing

01:41:26   sit there un-responded for six months,

01:41:27   see how you like that?

01:41:28   How about I leave your thing un-responded?

01:41:29   I know Apple, suddenly you want my feedback

01:41:32   and you're sending me emails every week saying,

01:41:33   "Hey, just so you know, your feedback, we asked for,

01:41:36   you know, we need a response from you,

01:41:38   we need some feedback, engineering needs you to confirm this.

01:41:40   They would send you emails every week or two saying,

01:41:43   letting you know that you're supposed to respond.

01:41:45   And I would just ignore them and be like,

01:41:47   I don't have, like not because I'm being mean and spiteful,

01:41:50   but I literally don't have time to install a beta

01:41:53   to figure that out.

01:41:53   And what I figured was,

01:41:55   the new public version of macOS will come out

01:41:58   and then I'll be able to see if it's fixed, right?

01:42:00   That's what I figured would happen, right?

01:42:02   Instead, what happened is after in my case, 35 days,

01:42:07   Apple said, "We're just gonna close your bug.

01:42:10   "We didn't hear from you, so later."

01:42:12   (laughing)

01:42:13   Right, and basically-- - Awesome, awesome.

01:42:14   - So the tolerance of Apple,

01:42:17   I thought it was like on a fixed timer,

01:42:18   but apparently not because Geoff Johnson

01:42:20   had only waited like 16 days.

01:42:22   Like when they ask you, "Hey, confirm that this is fixed,"

01:42:25   if you don't respond on their timetable,

01:42:28   they'll just say, "Well, we waited a while,

01:42:30   "and we didn't hear from you,

01:42:31   "so we'll just assume it's fixed, done, bye."

01:42:34   Which is not the way it should work.

01:42:36   If you're gonna have this system

01:42:37   where we don't close it until the developer confirms

01:42:39   that it's fixed, you have to wait for them to confirm.

01:42:42   And if basically no response, well I'll assume it's fixed,

01:42:46   why would you assume it's fixed?

01:42:47   Maybe I'm dead, that's why I didn't respond.

01:42:50   You don't know what's going on over here.

01:42:51   I didn't respond to your confirmation thing,

01:42:54   you can't be, that must mean it's fixed.

01:42:55   I'm closing it out, right?

01:42:57   At least let the release version of the OS come out

01:43:00   because I'll update, I'll update to the latest version

01:43:04   and then I'll be able to tell,

01:43:05   and by the way, it's not fixed.

01:43:06   But I'll paint the latest version

01:43:10   'cause 13.3 is out, it's totally not fixed.

01:43:12   It's a bad bug report, I don't wanna get into it.

01:43:14   Like I don't blame Apple, this is a really complicated issue

01:43:16   I don't even know how to fully report this

01:43:18   but it's doing something that it shouldn't be doing

01:43:20   or maybe it should be, it's hard to tell

01:43:21   without talking to a human but it's a minor issue.

01:43:23   I don't really care that much about it.

01:43:24   But like, but and Jeff Johnson has a similar story,

01:43:27   his was closed after 16 days.

01:43:29   For the same reason, he's like,

01:43:29   "I don't have a good way to install a beta,

01:43:31   "I'll just wait for the release."

01:43:32   Oh, nevermind, they closed it before the release came out.

01:43:36   Apple's internal system patience for lack of response from developers seems very low.

01:43:43   Whereas we file things and months, weeks, years go by and we hear nothing, we're expected

01:43:48   to just tolerate that, right?

01:43:50   So it is a very asymmetrical relationship.

01:43:53   And I think Apple's policy surrounding this is not great.

01:43:55   I would actually argue for inside Apple, if you think you fixed the issue, to save us

01:44:01   all time, close the bug.

01:44:02   Say closed, we think it's fixed as whatever.

01:44:04   And I know they have this weird thing where you can't reopen bugs, which just seems like a real problem with their system, but like

01:44:09   Yeah, if you wait for all developers to confirm it's gonna take you forever

01:44:14   And they don't actually wait for you to confirm. They'll auto close it after n days anyway

01:44:19   So it's like just be honest and say when you think you fix it just close it as resolved say

01:44:23   Resolved fixed in Mac OS blah blah blah blah blah because then there's a known resolution and if I come back to it

01:44:29   I'm on vacation. I come back from vacation

01:44:31   and they're like, oh, looks like my bug was closed,

01:44:33   and they said it was fixed, and macOS, blah, blah, blah.

01:44:35   I can confirm that at any time at my leisure.

01:44:37   If I care about that bug, presumably macOS has rolled on

01:44:41   since then, and I'll try to reproduce it.

01:44:43   And if it's not fixed, in a sane world,

01:44:45   I would reopen the bug, but in this world,

01:44:47   I can refile it, right?

01:44:48   You don't have to wait for me to confirm

01:44:50   and send me nagging emails and then just close it anyway

01:44:52   when I don't respond for two weeks.

01:44:54   All right, that was just a little bonus content there,

01:44:56   so I can link to the Jeff Johnson thing,

01:44:57   but that policy also seems counterproductive, let's say,

01:45:01   because it's just, it's pretending that a relationship exists

01:45:05   that doesn't actually exist, which is like,

01:45:07   we're talking back and forth,

01:45:08   we're working together on this bug, no, we're not.

01:45:10   We're throwing things over the wall,

01:45:12   and you are occasionally popping up every few months

01:45:14   to ask us to do something which may be inconvenient for us.

01:45:17   - Yeah, and as a final note on this,

01:45:20   in the same vein as what I was saying about Siri,

01:45:24   like, I think that there's a lot of people within Apple

01:45:27   would agree that radar is perhaps not optimal, but I don't think that most people, even potentially

01:45:32   the rank and file, really understand how aggressively awful it is for third-party developers.

01:45:39   And here again, it's one of those things where I think it's easy for Apple to be like,

01:45:43   "Eh, it's not great for us either," you know, or, "Eh, you know, you don't really understand how many we get,"

01:45:49   or, "Eh, it's really, really a hard problem to solve," which all those things are true.

01:45:54   But ultimately, that doesn't matter. At some point, you need to fix the problem. At some point,

01:46:03   Siri needs to be better than a pile of trash. At some point, Radar and Feedback Assistant

01:46:11   need to be better than just throwing things into Dev null and hoping for the best.

01:46:15   Like, it doesn't matter if you can excuse yourselves away from why this is bad. I don't

01:46:22   care. You all make a ridiculous amount of money to solve really hard problems. Here's a really hard

01:46:29   problem to solve. Fix radar. It's actually not that hard, but still fix radar. Please and thank you.

01:46:34   All right, can we do some Ask ATP to cheer me up please? Yes please. All right, Steven Goza writes,

01:46:40   "John and others have mentioned a bunch about Mastodon being federated and I can't wrap my

01:46:44   head around it. I ignored it for a while until last week, which was probably two months ago,

01:46:48   when John said an instance could become "unfederated."

01:46:52   Now I can't ignore my confusion anymore and would love an explanation when you have a chance.

01:46:55   Which, unfortunately, that time didn't come until now, but now we have the chance.

01:46:59   So John, what does "federating" or "federation" mean?

01:47:02   - Usually people use the email analogy,

01:47:05   and I will use that briefly here before I try a different one as well.

01:47:08   So the email analogy is you can pick an email provider, Gmail, Hotmail,

01:47:13   you can host your own email in your own domain, whatever, iCloud, Mail, all that stuff.

01:47:17   stuff, that is kind of your instance.

01:47:21   You can email me at myname@icloud.com, at myname@gmail.com, or whatever.

01:47:25   You picked Gmail, you picked iCloud, you picked Yahoo, you picked Hotmail.

01:47:29   Those are your instances.

01:47:30   But you can send email to anybody.

01:47:32   When they say, "Oh, just send me your email address," you don't look and say, "Oh, I can't

01:47:35   email you.

01:47:36   I'm on Gmail and you're on Hotmail."

01:47:38   No.

01:47:39   You know that you can send email to anyone on any other email instance, because that's

01:47:43   how email works.

01:47:44   It is a federated system, kind of.

01:47:47   So in Mastodon, we have little Mastodon addresses.

01:47:50   It's your username at your instance.

01:47:51   They look kind of like email addresses,

01:47:52   but we put another ad on the thing.

01:47:54   So mine is @Syracusa@Mastodon.social, right?

01:47:57   It's like an email address, right?

01:48:00   But I can follow people on any other instance.

01:48:03   And they can follow me the same way

01:48:05   we could email each other if we are in different email

01:48:08   instances or servers or whatever.

01:48:10   That's what Federation means.

01:48:12   The more precise or slightly more precise technical explanation,

01:48:17   which still glosses over some details, is the Mastodon.social server that I'm on.

01:48:22   It talks to other servers to find out what's happening on them.

01:48:27   And the way it does that is it looks all the people who have accounts on Mastodon.social.

01:48:31   What is the... I'm going to give you a SQL query now.

01:48:34   Give me the unique list of instances that people follow on the server.

01:48:39   So 500 people follow someone from this instance,

01:48:42   one person follows from this instance,

01:48:43   just give me the uniquefied list of all the instances.

01:48:46   Those are the instances

01:48:48   that mastodon.social needs to talk to.

01:48:50   It doesn't need to talk to all the instances in the world,

01:48:52   because if nobody on mastodon.social

01:48:53   follows someone on foobar.social,

01:48:56   it doesn't need to talk to that thing at all, right,

01:48:58   and vice versa.

01:48:59   So it can figure out who do I need to talk to,

01:49:01   and then periodically,

01:49:03   they communicate using the activity protocol

01:49:05   to exchange information.

01:49:08   tell me what the people that my people follow are saying,

01:49:10   and by the way, here's what my people are saying,

01:49:12   and I will send that to all the other instances

01:49:15   for people that follow them, right?

01:49:17   So you can sort of see how they work,

01:49:19   in the same way that an email server will receive email

01:49:21   from anybody who sends it,

01:49:22   and then will allow email to go out or whatever.

01:49:25   Unfederation is something that kind of also happens

01:49:29   in email as well, where an instance,

01:49:33   where one instance will decide, like for example, Hotmail.

01:49:36   If Hotmail decided, you know what,

01:49:38   we're not going to allow any more email from Gmail users.

01:49:41   Not that they would ever do that, but just humor me, right?

01:49:44   That would mean that if you're on Gmail

01:49:46   and you send email to someone at Hotmail,

01:49:47   it would bounce back and say, nah, sorry,

01:49:50   I couldn't deliver the email.

01:49:52   The Hotmail server said they're not accepting email

01:49:54   from Gmail anymore, right?

01:49:56   That happens with spammers.

01:49:57   We're like, there's some thing that's emailing

01:50:00   and it's spamming, right?

01:50:02   The big email services like Gmail, Hotmail, Yahoo,

01:50:05   iCloud will unfederate.

01:50:07   - By the way, I'm pretty sure it's called defederate.

01:50:09   - Yeah, the big email servers will have a denialist

01:50:12   that will remove all of the spam servers

01:50:15   that they don't wanna receive email from.

01:50:17   So you can imagine with a federated system like this,

01:50:20   especially with something like Mastodon

01:50:22   that's not quite the same as email,

01:50:24   you could draw them out as little islands.

01:50:26   You could have an island of instances

01:50:28   that all talk to each other,

01:50:30   but they don't talk to this other island of instances,

01:50:31   or maybe there's one or two connections or whatever.

01:50:34   So it's not like there's just one giant pool

01:50:36   and if you get defederated, you're off to the side.

01:50:40   It could break up into a whole bunch of little islands

01:50:42   with sparse connections between them

01:50:43   and lots of internal connections.

01:50:44   We'll see how this is gonna shake out.

01:50:46   But in the case of Mastodon,

01:50:49   defederating is considered a feature, not like a burden,

01:50:54   'cause email is supposed to be universal communication.

01:50:56   Setting aside spammers,

01:50:57   anyone should be able to email anyone else.

01:51:00   But on a social network,

01:51:02   the instance can set rules for behavior.

01:51:05   For instance, just off the top of my head you could say,

01:51:08   this instance does not allow anyone to use curse words.

01:51:12   And everybody who doesn't wanna see curse words

01:51:16   could be on that instance.

01:51:17   And they might decide, that instance might decide

01:51:19   to de-federate from any other instance

01:51:21   that allows the use of curse words.

01:51:24   If they did that, anyone who's on that instance

01:51:25   would be like, wait a second,

01:51:26   I used to follow a person over there,

01:51:28   now I can't see their stuff anymore.

01:51:29   And they say, oh well, we don't want curse words

01:51:32   coming into our instance, so we defederated from them.

01:51:34   And that person might say, oh, that sucks.

01:51:37   I'm leaving this instance and going to an instance

01:51:39   that is federated, and in that way,

01:51:41   people would sort themselves into groups

01:51:43   according to who they are willing to be federated with

01:51:47   or whatever.

01:51:48   In practice, what we mostly expect

01:51:50   is there'll be a reasonable set of behavioral rules

01:51:55   that the vast majority of people will agree upon,

01:51:57   and defederation will only happen when a Nazi

01:52:00   or a child porn server or some terrible thing comes and they'll be defederated and they'll

01:52:04   kind of be like the civilized world of people who have basic standards for behavior and

01:52:12   you know because defederating is like the nuclear option right you can always just not

01:52:16   follow people and you can block individual people and you can mute people you can do

01:52:19   all those types of things.

01:52:20   So the defederation is like look this entire instance is filled with Nazis.

01:52:23   It's so bad it's unredeemable if any good people are there I'm sorry but we're going

01:52:27   to defederate from you or whatever.

01:52:30   And so we don't know how this is gonna work out.

01:52:31   This is kind of the first time this is being done

01:52:33   at a large scale where defederation is a feature

01:52:37   of the system and not just something to use

01:52:39   to deal with spam and abuse,

01:52:40   but just a potentially ideological tool

01:52:43   or a way for people to sort them into different bins.

01:52:46   As I said in past shows, it may be the case

01:52:49   that what happens here is kind of what happened with emails.

01:52:51   You get a bunch of really big instances

01:52:53   that are kind of too big to fail

01:52:54   and that could never defederate from each other

01:52:56   because it would break the entire system

01:52:57   and that would be kind of a shame

01:52:58   because that kind of defeats the purpose of federation

01:53:01   when you have these, you know,

01:53:03   duopoly or whatever really big things,

01:53:05   but we'll see how it shakes out.

01:53:07   So that's how Mastodon is structured.

01:53:10   That's how it's supposed to work.

01:53:11   We'll see how it's gonna actually work in practice.

01:53:14   - Yeah, really quick shout out to the "Decoder" podcast,

01:53:17   which I don't personally listen to every episode,

01:53:19   but everyone I've heard has been very good.

01:53:21   They just had on Eugene Roachko.

01:53:24   I probably butchered that pronunciation.

01:53:25   I'm sorry about that.

01:53:26   he is the the benevolent dictator for life for Mastodon and it's like an hour

01:53:32   hour and a quarter it was really really really good I just listened to it a few

01:53:36   hours ago so that is worth listening to and it's it's clear that he is someone

01:53:41   who whether or not you agree with him deeply cares about Mastodon and deeply

01:53:46   cares about doing the right thing and I think I generally agree with a lot of

01:53:51   the things he said except maybe quote tweets or quote toots but that's neither

01:53:54   here nor there, but they talk a bit about federation in this and I really think it's

01:54:00   a fascinating case study on, you know, can communities really self-govern? What happens

01:54:06   if they turn kind of evil? You know, what is this all going to mean? This is what you

01:54:09   were talking about earlier, John, but the conversation with the guy who created Mastodon

01:54:14   is really, really, really interesting. And so if you have the time, I definitely suggest

01:54:18   it. We'll put a link in the show notes. All right, moving on. Eric Smith wants to know,

01:54:23   Is software getting worse?

01:54:25   And this is a blog post on the Stack Overflow blog from late January.

01:54:30   And it talks about, you know, is software legitimately getting worse?

01:54:33   And so someone, I presume John, pulled some quotes for me to read from this blog post.

01:54:39   I suspect bugs per line of code is more or less staying constant, but applications are

01:54:42   much more complex than they were, leading to more bugs in the absolute number sense.

01:54:46   Also, the threshold for a quote unquote "bug" has changed over time.

01:54:50   For example, tearing and frame latency were considered normal.

01:54:53   Now they're a bug.

01:54:54   I don't think that expanding the set of "bug" is terrible, but a byproduct of that will

01:54:59   be that it's hard to keep that set small.

01:55:01   There's an old adage in software, which is one of those old adages that no one really

01:55:07   has the wherewithal to actually test, but people hear and it rings true to them, so

01:55:10   we keep repeating it, so here I am doing that.

01:55:12   It's that the number of bugs per line of code written by a programmer hasn't changed that

01:55:18   much over the years.

01:55:20   You know, there's a range, there's a bell curve or whatever, but in general if you're

01:55:23   going to write a thousand lines of code, how many bugs on average do you expect to find

01:55:26   in a thousand lines of code?

01:55:28   And that makes some sense, obviously, not all lines are created equal, you know, it

01:55:33   depends on what language you're writing it or whatever, but in general if you're going

01:55:36   to write a thousand lines, there's some average percentage amount of bugs you're going to

01:55:41   have.

01:55:42   And then you debug and there's various software testing methodologies and QA and stuff to

01:55:46   remove those bugs and make the software as bug free as time and money will allow.

01:55:50   That is a well known thing.

01:55:51   I don't think that's changing much.

01:55:53   The reason most people believe that is because although the software industry and technology

01:55:57   advances by leaps and bounds, humans change way slower.

01:56:03   In the blink of an eye between the invention of electricity and today, human evolution

01:56:09   has done nothing, except for maybe like work up some immunities to some viruses and stuff,

01:56:15   but like just generally nothing.

01:56:17   So it's not as if, oh, because humors do faster, we'll get so much better at programming that

01:56:22   we'll make fewer mistakes.

01:56:23   Now we're pretty much making the same amount of mistakes.

01:56:25   And incidentally, this particular argument very often comes up in the context of high-level

01:56:30   versus low-level languages, which we've talked about many times in the past.

01:56:33   It's why proponents like me of high-level languages say no matter how much you love

01:56:38   C or C++ or assembly or whatever you may be arguing for our objective C or whatever, the

01:56:44   bottom line is higher level languages let you write fewer lines of code and even if you think

01:56:48   it's all syntactic sugar or it's pointless or it's taking me too far away from the internals and I

01:56:54   need lower level whatever, the the unescapable fact is the more things you type the more opportunity

01:57:01   there is to make bugs. So if you have a higher level language that lets you do things with

01:57:06   fewer lines of code with less boilerplate,

01:57:09   with less repeated things, with less worrying about details

01:57:13   that aren't important to your program,

01:57:14   you will produce fewer bugs

01:57:17   for the functionality you're making.

01:57:19   You're still doing the same number of bugs per lines of code

01:57:21   but because a thousand lines of Swift

01:57:24   does way more than a thousand lines of assembly,

01:57:27   your bug per line is the same

01:57:30   but your bug per functionality is way better.

01:57:32   As for Eric's question, is software getting worse?

01:57:36   I think basically no.

01:57:39   I think it's actually getting better.

01:57:41   Mostly because every part of the stack that we're using gets higher and higher and higher

01:57:45   level and the workhorse we're building on top of everything that has come before, we're

01:57:50   getting so much more functionality for the amount of bugs that we're getting.

01:57:55   Now that might not feel important to you because you're like, "I don't care that what Twitter

01:58:00   Quine is doing would be science fiction to someone in the 60s in terms of holy cow, what

01:58:05   What is it actually doing under the covers?

01:58:07   Networking stacks and requests and data encoding and decoding and the whole operating system

01:58:13   stack on top of it and the layout engine.

01:58:16   It is insane.

01:58:17   And then for you it's like I scroll, I see a bunch of words and I scroll.

01:58:21   It's easy for us to take that for granted, but the amount of functionality as in what

01:58:25   is it actually doing to let you use your thumb to scroll a bunch of words is phenomenal.

01:58:31   And the number of bugs that it has compared to that amount of functionality versus the

01:58:34   the number of bugs like a recipe manager on an Apple 2e had,

01:58:38   like the number of bugs is similar,

01:58:40   but the functionality is vastly greater.

01:58:43   So I would say software is getting better,

01:58:45   just maybe not in a way that users notice

01:58:47   'cause users are very quick to take for granted

01:58:50   like the magic of what these things are doing.

01:58:54   And then all we can see is like,

01:58:55   yeah, but there are still bugs.

01:58:56   There are always going to be bugs

01:58:58   unless you can spend huge amounts of time and money

01:59:00   stamping them a lot, which you can't for an application

01:59:03   and lets you scroll tweets or whatever, right?

01:59:05   But we do get more and more functionality over time.

01:59:07   So that's my take on this question.

01:59:09   - Yeah, I think there's also,

01:59:11   you have to look at what software is asked to do over time

01:59:16   and how that changes over time.

01:59:18   Our expectations are so vastly accelerating over time

01:59:23   and what we expect software to do today

01:59:26   is so different from what it did back when we had Apple IIs

01:59:30   running Printile Pro or whatever.

01:59:32   It's so different compared to that.

01:59:35   We expect our software to do way more.

01:59:37   On one level, John is right that the languages get better,

01:59:41   the abstractions we're working on get better,

01:59:42   the tools get better, and so that does,

01:59:45   that multiplies our productivity.

01:59:47   Like how good can a program be per programmer working on it

01:59:52   or per hour of time that's put into working on it?

01:59:56   We have made huge advances there.

01:59:58   But the reason it can feel like software

02:00:00   so buggy these days is that we are asking it

02:00:02   to do a heck of a lot more than we ever have before,

02:00:04   and that's always increasing.

02:00:06   As soon as software achieves some new amount

02:00:09   of functionality, we immediately take it for granted

02:00:12   and move on to asking about other things.

02:00:15   So for instance, even if, we don't have to go back

02:00:17   to the Apple II, I'll just go back to the beginning

02:00:19   of the iPhone, 'cause this is like where I've done

02:00:22   a lot of my professional software development,

02:00:24   so I've been here a while, I've seen a lot of things change

02:00:27   over time and customer expectations changing over time.

02:00:31   And when the iPhone, I'll even say when the App Store

02:00:33   first launched, so the iPhone gets its first year for free.

02:00:36   When the App Store first launched in 2008,

02:00:39   one person could make an app and the tools

02:00:43   were pretty primitive, the hardware was pretty primitive,

02:00:46   not compared to the Apple II, by any means,

02:00:47   we were way past that, compared to where we are today.

02:00:52   Things were simpler and more primitive and lower level.

02:00:55   We didn't have luxuries, I mean heck,

02:00:57   the first version of the iPhone SDK

02:00:59   didn't even have interface builder,

02:01:00   or core data, or anything like that.

02:01:02   We were super early, everything was just to see,

02:01:06   manual, retain, release, auto-release,

02:01:08   like that kind of stuff.

02:01:09   - Oh gosh, that's right.

02:01:10   - All the UI was built in code for that first,

02:01:12   there were no storyboards yet, or anything like that.

02:01:14   So it was early, it was basic,

02:01:17   and UIKit was super early, super basic.

02:01:20   But all you had to focus on

02:01:23   was whatever that app could do on the iPhone 2.0 screen.

02:01:28   And it was a single device with a single screen resolution

02:01:33   and there was no background tasking,

02:01:35   there were no extensions, there was no Apple Watch,

02:01:40   there was no iPad, there was no Catalyst app,

02:01:42   there was no widgets, like all the stuff

02:01:44   that we expect today wasn't there.

02:01:46   And back then, you expected also that the app

02:01:49   would kind of hold its own data.

02:01:51   There was a lot less expectation of cloud-based storage,

02:01:55   a lot less expectation of syncing and of cloud backup.

02:01:59   Most things didn't use accounts

02:02:01   and didn't have account mechanics in them.

02:02:04   There was no in-app purchase.

02:02:06   There were very few ads.

02:02:08   It was such a different experience back then,

02:02:12   not only because of the technological situation,

02:02:14   but also because the customer expectations were simpler.

02:02:17   Now, you are expected, if you're making an iPhone app,

02:02:21   you're expected to have all of that stuff.

02:02:24   You're expected definitely to have Sync.

02:02:27   To do Sync well, you're probably gonna need

02:02:29   some kind of basic account system,

02:02:31   and that adds a whole bunch of complexity.

02:02:32   (laughs)

02:02:33   That's a lot there.

02:02:35   You're probably gonna have interactions

02:02:37   with different web services to pull different stuff in.

02:02:39   Your customers are gonna expect you to have things

02:02:41   like an iPhone and iPad app that somehow Sync.

02:02:45   Definitely those have to be the same purchase.

02:02:47   You're gonna have to have a watch app probably

02:02:49   for certain types of things.

02:02:50   you're gonna have to have extensions,

02:02:51   you're gonna have to have share extensions,

02:02:52   you're gonna have to have a widget somewhere,

02:02:55   lock screen widget, main screen widget,

02:02:57   watch complication maybe.

02:02:59   So the expectations of what people want you to do

02:03:01   are so high that the software necessarily

02:03:05   has to be more complicated just to fit

02:03:08   what people expect all apps to do these days.

02:03:11   And so it can feel like things are worse quality

02:03:14   because we're just asking them to do so much more.

02:03:17   There are parts of what we do that are crappy quality.

02:03:20   We mentioned Siri earlier,

02:03:21   and Siri's really inconsistent,

02:03:23   but Siri is so vastly more complex

02:03:28   than anything we ask things to do,

02:03:29   even 15 or 20 years ago.

02:03:31   It's frankly amazing it works at all,

02:03:34   let alone as quote, "Well," as it actually does.

02:03:37   'Cause Siri actually does work well

02:03:38   in certain things sometimes.

02:03:40   And it's amazing to know that when you were here

02:03:44   in earlier days of technology,

02:03:45   when we didn't have anything close to that,

02:03:47   and it's remarkable that it works the way it does even.

02:03:51   But the software that we wrote back in the day

02:03:53   that might have felt more solid was a lot simpler

02:03:57   and in some ways was more solid,

02:03:59   but also now we wouldn't be happy with that

02:04:02   and customers demand so much more now.

02:04:05   You can't make software like that anymore.

02:04:07   And it's a shame because I would love

02:04:09   if software was super reliable and super simple

02:04:13   in some ways or rather I think I would love that,

02:04:15   But then my actual customer demands would be like,

02:04:19   well, I love this really simple software,

02:04:21   but can you maybe make an iPad app,

02:04:23   or can you maybe add some shortcut support?

02:04:26   There's always something that you want them to add.

02:04:29   Man, I love paying once for apps

02:04:31   and never having to pay again.

02:04:32   But oh, can you add Sync to a cloud service?

02:04:34   Like there's so many of those things

02:04:37   that modern customer expectations

02:04:40   and modern technological environments

02:04:42   in which everything has to work and interoperate

02:04:44   and meet expectations just can't support

02:04:49   that old simpler way of looking at software.

02:04:52   - Oh, you know, so if the lines of,

02:04:54   the bugs per lines of code was still the same back then,

02:04:56   part of the reason, and I think it was,

02:04:58   and that is still a factor, but part of the reason

02:05:01   some of those older programs felt more reliable,

02:05:04   especially the good ones that we remember,

02:05:06   is because the market structure and the incentive structure

02:05:10   were different back then, and not back when the iPhone

02:05:12   came out, but I'm talking farther back than that.

02:05:15   So before the internet, you had to buy your software

02:05:18   in a box on a floppy disk.

02:05:20   And if there was something wrong with it,

02:05:22   you couldn't just go download an update to that.

02:05:25   You could download it where?

02:05:26   On a CompuServe?

02:05:27   Maybe.

02:05:28   They'd have a second version on a different floppy disk.

02:05:31   So the incentives were for the software developer

02:05:34   to spend way more time making sure the software they were

02:05:37   going to pay to put on millions of floppy disks

02:05:41   was as bug-free as possible because they knew they can't just have people install a patch,

02:05:46   a day one patch when they launched the app because there was no internet, right?

02:05:49   So those incentives are different, but like that doesn't mean that the programmers weren't

02:05:53   making the exact same number of bugs per lines of code.

02:05:55   All that means is they had to spend more time and more money getting that version 1.0 to

02:06:02   a more bug-free state.

02:06:04   Like any company could do that today, it would just mean they have to ship later, right?

02:06:07   It also means if there was a bug, and there certainly was, there was always bugs, it would

02:06:12   take way longer for you to get that fixed because you couldn't just download an update.

02:06:16   They'd have to print a version 1.1, put that on floppy disk and sometimes sell it to you,

02:06:20   or you'd send a way for it and it's just like it was worse in all sorts of ways that people

02:06:24   would not tolerate, right?

02:06:25   Whereas now we just expect, "Oh, there's a bug.

02:06:27   There better be an update tomorrow when I wake up."

02:06:31   There was not going to be an update to your floppy disk when you woke up, right?

02:06:34   It was what it was.

02:06:36   So those incentives could exist today and do in different markets for example in theory,

02:06:40   I don't know if this is true, maybe Casey can tell me, but in theory military software

02:06:45   development has a much slower pace and more thorough QA process and you know whatever.

02:06:50   I hope so.

02:06:51   Yeah, well you know also more money being spent in various states to be basically a

02:06:58   jobs program where we blow people up.

02:07:00   Anyway.

02:07:01   Yeah but it's also a lot higher stakes than like you know some weather app on your phone.

02:07:04   Yeah or like self-driving.

02:07:06   think about it. That was originally the DARPA challenge, you guys don't remember that, it

02:07:10   was a good NOVA on it. Anyway, so I think like different incentive structures can and

02:07:18   do exist and those also influence perceived quality but in general the, you know, human

02:07:23   beings produce X number of bugs per Y number of lines of code is not going to change until

02:07:28   and unless human beings change and that happens really slowly. So I think any perception that

02:07:33   you have. Think about the surrounding context. You know, what functionality am I getting?

02:07:40   What incentives exist for this functionality? How fast can I get an update? Because you're

02:07:44   trading things off. You're trading off the ability to ship something with a bug in it

02:07:50   is traded off against the ability for that bug to be fixed tomorrow through the magic

02:07:54   of software updates. Right? And so that creates an incentive structure for shipping software

02:07:58   that is different from the one where you know you can't do an update until six months from

02:08:01   now.

02:08:02   - Alrighty, and then finally for tonight,

02:08:04   Thomas Alvarez writes, "John mentioned how he wants

02:08:07   "a Mac Pro with GPUs on cards, and I'm curious

02:08:09   "what use he has for GPU-powered in an ARM-based Mac.

02:08:13   "If it's gaming, aren't we past that on the Mac now

02:08:15   "with Apple Silicon and moved on to running Windows

02:08:18   "on a dedicated gaming PC?"

02:08:19   We all know how you feel about a dedicated gaming PC,

02:08:22   but would you just mind answering the question, please?

02:08:25   - Yeah, so I mean, I don't think I've actually talked

02:08:27   about my specific needing GPUs on cards

02:08:31   cards for an ARM-based Mac.

02:08:33   Mostly we've been talking about if it doesn't have GPUs on cards, how is it not just a,

02:08:38   you know, a Mac Studio with a bunch of empty space or something.

02:08:41   So anyway, setting that thing aside, I am actually interested in significant GPU-powered

02:08:47   in an ARM-based Mac.

02:08:50   Back when we were still entertaining rumors of this sort of quad SoC arrangement, and

02:08:55   I did the math on that, that would have enough GPU power to be equivalent to a pretty okay

02:09:05   external PC graphics card and I would be fine with that.

02:09:07   And why do I want that?

02:09:08   It's to play games.

02:09:09   And you'd be like, "Well, you can't play any games on an ARM Mac."

02:09:12   I still am personally holding out hope that Windows and gaming will get on the ARM bandwagon

02:09:19   eventually.

02:09:20   I don't have any reason to particularly believe that because it doesn't seem to be really

02:09:24   happening but you know it seems plausible technically it's certainly possible we know

02:09:29   that but markets move slowly and I'm not sure if there's much motion there maybe the server

02:09:34   stuff needs to go and destroy the economics of good x86 CPUs for that to happen I don't

02:09:39   know how that's all going to happen but I would like that to happen and in the meantime

02:09:43   I just you know for the same reason I have this big honking Mac that I don't need over

02:09:46   here I just like the idea of having more computing power than I actually need even if it's just

02:09:51   So I can download like some demo of some 3d program that I'm never going to figure out how to use and just play with

02:09:55   or whatever or you know, like I

02:09:57   Locked myself out of a locked note bag, you know

02:10:01   The Apple Notes like back before you could use your Apple ID to lock a note

02:10:05   You could put individual passwords on notes

02:10:07   And I had a note that I had put an individual password on that

02:10:10   I had forgotten and hadn't put it in keychain or anything like that

02:10:13   And so I used my dual GPUs to crack it like

02:10:19   - To brute force crack, yeah, you can.

02:10:21   - That's amazing.

02:10:22   - To brute force crack my own password, and guess what?

02:10:24   The fans really smiled on my Mac Pro,

02:10:27   but I did crack, 'cause it was a short,

02:10:29   like throwaway password or whatever, I just--

02:10:31   - How long did it take?

02:10:32   - I don't know, like five minutes, like it wasn't--

02:10:34   - Oh my god.

02:10:35   - I did not use a 15 character, whatever password,

02:10:37   like this is an argument for you to do the thing where,

02:10:40   I don't remember when they did that, it was ages ago,

02:10:42   where they let you do the thing like,

02:10:43   do you wanna update your thing,

02:10:44   so your Apple ID in your face can unlock all your notes,

02:10:46   or whatever, I eventually did that, right?

02:10:49   But that's just a silly example.

02:10:51   But I could have also cracked it on an iPhone.

02:10:55   I'm not pretending that I'm in a big Jeep, you do this.

02:10:57   But the same reason that people don't need a sports car,

02:11:00   'cause they're not race car drivers,

02:11:01   they just like to have one,

02:11:02   even though you can't really use it without breaking the law.

02:11:05   At least I'm not breaking the law

02:11:07   with my Mac Pro here or whatever.

02:11:09   Remember those ads when you couldn't export the G4

02:11:12   to communist countries

02:11:15   because it was like a restricted export thing?

02:11:17   I think Apple had an ad campaign about that.

02:11:18   classified as ammunition because it's too good of a computer.

02:11:20   - Yeah, anyway, I just want it.

02:11:23   Even if it's for that, you know, so I can play

02:11:24   a five year old port, a port of a five year old game

02:11:27   and high frame rates, the Mac OS version or whatever.

02:11:30   That's it, it's, you know, and it's for games.

02:11:32   Like I really hope someday, you know,

02:11:34   'cause ARM based SOCs made by Apple with lots of GPU power,

02:11:39   that's really good hardware for games.

02:11:41   Too bad no one makes games for it.

02:11:42   Too bad Windows doesn't run it, like I get it.

02:11:44   Doesn't make perfect sense.

02:11:46   And as for the dedicated gaming PC,

02:11:48   - Honestly, if I had a different house with another desk

02:11:51   for things to put on,

02:11:52   I probably would have bought a gaming PC by this point.

02:11:54   But I don't know.

02:11:55   - The desk is what's holding you back?

02:11:57   - Yeah, right?

02:11:58   - I'm gonna buy an entire Mac Pro instead of a desk.

02:12:02   - Well, no, it's not the desk,

02:12:05   it's the house where the desk would go.

02:12:06   Like there's no, there's no place,

02:12:08   it's not like I'm getting a new house anytime soon.

02:12:10   And I do have my, I've talked to those about it,

02:12:12   I do have my own irrational personal biases

02:12:14   against Microsoft and Windows

02:12:15   that will make me never wanna buy a PC.

02:12:17   and honestly at this point it's been such a long streak,

02:12:18   I feel like I need to keep it going.

02:12:20   But in theory, I also haven't bought an Xbox

02:12:24   for that same stupid reason.

02:12:25   But in theory, if I had a bigger house

02:12:27   with another desk for me to put stuff on,

02:12:28   I might actually get a gaming PC to play PC games.

02:12:31   And that's why I'm enjoying my Mac Pro.

02:12:32   I can play Microsoft Flight Simulator

02:12:34   with my fancy new GPU and it looks really cool

02:12:36   and I can't play it on any other platform.

02:12:38   I play it in Windows and I get good frame rate

02:12:40   and it looks really nice.

02:12:42   And someday soon, well, someday, I don't wanna say soon,

02:12:45   but someday that will be over

02:12:46   because I'll have an ARM-based Mac Pro.

02:12:48   But hopefully by then, like Microsoft will like port

02:12:51   Flight Simulator to Windows on ARM,

02:12:54   and I'll be able to somehow boot into Windows on ARM

02:12:57   from an ARM-based Mac Pro and play Flight Simulator

02:13:00   at even higher frame rates.

02:13:01   That is the future I believe in that I'm dreaming of,

02:13:04   and that's why I want my ARM-based computer

02:13:07   to have a beefy GPU.

02:13:09   - Good luck.

02:13:10   - You might have to keep dreaming on that one.

02:13:12   - I know, I know.

02:13:13   - Thanks to our sponsor this week, Collide,

02:13:15   and thanks to our members who support us directly.

02:13:18   You can join us at ATP.fm/join.

02:13:21   Thank you so much, we will talk to you next week.

02:13:24   ♪ Now the show is over ♪

02:13:29   ♪ They didn't even mean to begin ♪

02:13:32   ♪ 'Cause it was accidental ♪

02:13:34   ♪ Oh it was accidental ♪

02:13:38   ♪ John didn't do any research ♪

02:13:39   ♪ Marco and Casey wouldn't let him ♪

02:13:42   ♪ 'Cause it was accidental ♪

02:13:45   It was accidental And you can find the show notes at ATP.FM

02:13:52   And if you're into Twitter You can follow them @C-A-S-E-Y-L-I-S-S

02:14:02   So that's Casey Liss M-A-R-C-O-A-R-M

02:14:06   Anti-Marco Armin, S-I-R-A-C, U-S-A-C-R-A-C-U-S-A, it's accidental. They didn't mean too accidental. Tech podcast so long.

02:14:26   All right, so speaking of sponsorships and memberships and things of that nature, we

02:14:33   are trying to come up with, like we were talking about the pre-show, more ideas for member

02:14:38   specials and it would be really awesome if we could get some ideas from listeners.

02:14:44   And so it would be excellent if you used Mastodon, if possible, to toot at, and I'll take the

02:14:49   fall on this, you can toot at me or the ATP show account, but you can send it my way.

02:14:55   - In particular-- - What is the ATP show account,

02:14:57   Casey?

02:14:58   - Good question.

02:14:58   - I believe it's @ATPFM@Mastodon.social.

02:15:03   - Thank you.

02:15:04   - How does Mike do this on upgrade?

02:15:05   He has a really good way of handling it.

02:15:07   It's on Mastodon.social @ATPFM, which is what--

02:15:11   - Yeah, you don't need to overcomplicate it.

02:15:13   Most Mastodon clients and the websites,

02:15:15   the search is good enough that if you just type ATPFM

02:15:18   into your search box on any Mastodon instance,

02:15:21   it will probably find us.

02:15:22   They should do a better job.

02:15:23   I was thinking about this,

02:15:24   should do a better job of indicating the verified URL.

02:15:29   - Oh, that's a good point.

02:15:30   Yeah, like put it like as like a little subtitle

02:15:32   in the search results, like name, verified URL.

02:15:35   - Yeah, 'cause what you wanna know,

02:15:36   if there's like 15 ATP FM accounts that like impersonate,

02:15:39   which one is the real one?

02:15:40   Only one of them can possibly be verified

02:15:43   against ATP.fm, the website,

02:15:45   and clients know that and show it,

02:15:48   they should show it in search

02:15:48   'cause it would help people find things.

02:15:50   But anyway, that's the address,

02:15:51   we'll put a link in the show notes.

02:15:52   You can also go to what?

02:15:53   ATB.FM/feedback?

02:15:57   - Yes, but that sends email.

02:15:58   Nobody likes email.

02:15:59   - Oh, email's fine.

02:16:01   - Anyway, yeah, so if you have ideas for members episodes,

02:16:07   okay, so here's the criteria.

02:16:08   Like, we'll accept any ideas,

02:16:09   but generally speaking, what we're trying to figure out

02:16:12   is something that we are equipped to handle.

02:16:16   So as an example, I would love to play Destiny one time

02:16:22   with John and Marco.

02:16:23   I don't have any PlayStation in the house.

02:16:25   I don't have any video capture software in the house.

02:16:27   I don't think any of the three of us, except maybe John,

02:16:30   has the patience to put together a video

02:16:32   or do any of that sort of thing.

02:16:33   - Oh, I totally have the patience for it.

02:16:35   And like I said, PlayStation 5 is video capture built in.

02:16:37   We're very close people.

02:16:38   Marco's got a PS5, I've got a PS5, KC's the holdout.

02:16:41   But anyway, that's already on the list.

02:16:42   You don't have to suggest that one.

02:16:43   - Well, I'm just saying that like something

02:16:45   along those lines that requires hundreds of dollars

02:16:47   of investment and--

02:16:48   - Or us going on a road trip across Australia,

02:16:50   probably not gonna happen, cool idea, but you know.

02:16:52   Great idea, but not gonna happen.

02:16:53   So things that we can handle with just a microphone,

02:16:57   and we can spend some amount of money.

02:17:00   In fact, if I can convince the boys,

02:17:03   I think I have a pretty decent idea for one

02:17:05   that involves spending an absurd amount of money

02:17:07   for a little bit of food.

02:17:09   I mean those words, I choose those words very carefully.

02:17:11   But anyway, something that we can handle,

02:17:14   and one thing in particular that we've been wrestling with,

02:17:16   particularly John, is we think,

02:17:19   how did you phrase this when we were talking privately?

02:17:21   we think we can get a pretty good licensing agreement,

02:17:24   I think John had said, for top four.

02:17:26   And so what is a good top four?

02:17:29   Well, top four Apple products, sure, that's kind of obvious,

02:17:32   but there's gotta be something more creative.

02:17:34   Top four, we were talking about doing privately,

02:17:36   we were talking about top four Apple announcements,

02:17:38   but then we have to do like 35 hours of research,

02:17:41   remembering all these announcements.

02:17:42   - Or do we, haven't you listened to top four?

02:17:45   - Yeah, fair. - Yeah.

02:17:46   - Anyway, the point is, particularly in the vicinity

02:17:50   top four, but even in general, if you have a member idea, please, you know, send us a

02:17:57   toot on Mastodon or feedback if you must. I guess our feedbacks are much better than

02:18:01   Apple's, aren't they? But anyway, yeah, send it our way. We would love to hear it. And

02:18:08   we hope to, again, this is not a guarantee, but our kind of goal, which we may not always

02:18:14   achieve, is maybe one of these a month if we can. It probably won't be every month,

02:18:18   maybe every other, but our goal, our hope is about one a month. So, and we can keep

02:18:22   watching movies and I'm sure we will, but let's, let's try to figure out as a collective,

02:18:26   let's figure out some other fun stuff to do. And if enough of you do join, then yes, I

02:18:29   will buy a stupid PlayStation and I will play stupid Destiny with my stupid friends.

02:18:33   And your kids will thank you for it. And by the way, for people who don't know, Top Four

02:18:36   is a podcast that Marco does with his wife, Tiff. It is on Relay. We'll put a link in

02:18:41   the show notes. They, I don't want to try to explain it. They ostensibly list the top

02:18:45   for of something, it's an experience. It is an experience for sure. It will make you mad,

02:18:51   guaranteed. Can confirm. And I have to say I have a newfound respect for what they do

02:18:56   in that program, having done just one very easy food-based challenge on a podcast, and

02:19:02   I cannot fathom what these two champions have done to their bodies over the course of this

02:19:07   show. Very, very true. It is like I have a visceral reaction now to thinking about what

02:19:13   there going through.

02:19:15   And that's also why I'm not really gung-ho about new food-based challenges in ATP, but

02:19:19   I may be forced to do so.

02:19:23   Semi-related to membership, which Casey also skipped over because I wanted to put before

02:19:26   this, I'm going to take advantage of my podcast platform to say I'm trying to sell my old

02:19:32   cameras and no one seems to want to buy them.

02:19:34   So I'm going to put a link in the show notes to a webpage I put up for me selling my old

02:19:39   cameras.

02:19:40   I'm being silly and for now trying to sell them to people who are willing to meet up

02:19:45   in the Boston metro area to avoid shipping fees and stuff like that, that's probably

02:19:49   not going to work and it's probably going to fall through and I'm going to have to ship

02:19:51   these too.

02:19:52   But for now, if you live in the Boston metro area and want to buy any of my stuff that's

02:19:57   listed on this webpage that's going to be in the show notes, it's just my old Sony A6300

02:20:00   and a couple lenses for what I think are pretty reasonable prices and everything is in nice

02:20:05   condition, hit me up.

02:20:07   All the information is on the webpage that will be in the show notes or you can just

02:20:10   go to hypercritical.co/for-sale/camera.

02:20:16   Anyway go to the show notes.

02:20:18   And I'm not going to say this is directly related to the bad sponsorships but it's not

02:20:22   not related.

02:20:25   [BEEPING]