00:00:13 ◼ ► It's good for an end of the year episode, too. Great for an episode where we recap the top stories of 2018.
00:00:21 ◼ ► We would never do that. We don't do enough homework for that. That takes way more preparation than what we're willing to do.
00:00:30 ◼ ► Yeah, and whenever connected does theirs, I haven't listened to today's yet, but whenever they do theirs, I'm always like,
00:00:35 ◼ ► "Damn, I wish I would have thought to do that." But then I realize the reality of doing that is so much work.
00:00:41 ◼ ► Especially so much work during the holiday season. I'm already busy enough. The last thing I want to do is add more work.
00:00:47 ◼ ► Yeah, the other problem is kind of like the Oscars where it's almost like stuff that happened at the beginning of the year doesn't count anymore.
00:00:53 ◼ ► No matter how important it was, the recency bias. I was going to actually add a section which was going to be 2018 impressions.
00:00:59 ◼ ► Technology impressions. It wasn't going to be like, "Let's survey all the important stories that happened."
00:01:03 ◼ ► It's going to be more like when you think back on 2018 in the technology world, what is your impression of it?
00:01:08 ◼ ► Do you feel like this was a cool year? What are the things that stick out in your mind?
00:01:12 ◼ ► But there's such tremendous recency bias that in my mind, all I can think about are the things that happened fairly recently.
00:01:18 ◼ ► So my impressions of 2018 are like the stuff that happened with the Mac line towards the end of the year, right?
00:01:29 ◼ ► And then the other thing, which we don't talk about much on the show every once in a while, but an entire year of Facebook privacy garbage.
00:01:39 ◼ ► My impression of it was basically that a bunch of bad Facebook stuff happened. Everyone who hated Facebook still hates them.
00:01:52 ◼ ► Testing the theory that people don't care about Facebook privacy stuff. We had a whole year of the worst possible things that could happen.
00:01:57 ◼ ► Tech people ignored it because they already hated Facebook and everyone else was like, "Whatever."
00:02:03 ◼ ► And I'm sure I'm missing tons of stuff that happened in the beginning of the year, but it seems like it happened so long ago that I can't bring it to mind feeling-wise.
00:02:16 ◼ ► 2018 feelings. That can be your analog episode if you haven't already done your year-end analog episode.
00:02:25 ◼ ► I'm surprised Mike hasn't gotten you into the "year of" thing. The year of adulting or whatever the hell he's doing. Just have a year of something. The year of living dangerously.
00:02:46 ◼ ► This was a big year. That's true. I was thinking more of the future, but he ain't wrong.
00:03:10 ◼ ► Jeremy Norring writes, "Hey, did Google cripple Edge's YouTube performance? What's that all about?"
00:03:29 ◼ ► "Harden the interruption. We see you've been here before. Let's make things official. Sign up with Google."
00:03:56 ◼ ► I understand this is where you put your, like, whiny little posts about how you hate everything.
00:04:00 ◼ ► But can we find somewhere else to do that? Can we just go back to Tumblr for that or something? I don't know. Whatever. Anyway...
00:04:11 ◼ ► I have only heard people complaining about it for a long time now. Even if you get past all the popover thingies or whatever,
00:04:22 ◼ ► There's a fixed top banner and a fixed bottom banner, so you're looking at what you can see of the article through a slit in your web browser
00:04:33 ◼ ► Yeah, that's the big problem that I had. It's like, you try to view it on a phone, which I hear is a common device to read web articles on in recent years,
00:04:40 ◼ ► and you just have all this crap cluttering up the viewport that's all fixed scrolling, so you can't scroll it away.
00:04:48 ◼ ► I don't know. I've never been a huge fan of medium, and as time goes on, they seem to be getting more annoying and less relevant.
00:05:04 ◼ ► Maybe my data is old, if you will, but especially when it was brand new, it seemed like everyone loved medium.
00:05:12 ◼ ► I think in the beginning they were courting people, so they were paying people to write on medium, I think,
00:05:21 ◼ ► That was later. They've gone through six different business models, a few of which involved trying to pay publishers to be on it
00:05:30 ◼ ► and to write good stuff on medium. It's been this whole cycle of craziness they've tried, and unfortunately,
00:05:38 ◼ ► as far as I know, they have not come upon something that actually works in a sustainable, stable way yet
00:05:46 ◼ ► It always seemed like, do you want to have a blog but not own your blog and have the benefit of your blog partially or wholly accrued to someone other than you?
00:05:56 ◼ ► And the sad thing is, I published at least one, maybe two things there a while back. I forget whether there's more than one.
00:06:03 ◼ ► But there is a role for it, I think. If you want to spread a message as far as possible without having your own presence,
00:06:19 ◼ ► So that's why whenever companies post their dying sunset messages, a lot of times those are on medium.
00:06:26 ◼ ► It's almost like an op-ed page for anybody. You can write whatever you want there, and if you don't have your own site,
00:06:35 ◼ ► your own blog that you're trying to maintain or build up or draw attention to, if you just want to get a point out,
00:06:40 ◼ ► it's a decent place to do that as long as you really don't want to build up yourself anywhere else.
00:06:45 ◼ ► But it does nothing for you other than get that point to more people. So if that's the point, that's fine.
00:06:52 ◼ ► But if you're trying to build up a writing reputation or an audience, it is not the place for that at all.
00:06:57 ◼ ► And I think one way you might describe it is like a social network for blogs, because that's like the network effect.
00:07:03 ◼ ► But the social network for blogs was blog. Before any of this started, people would link to each other's blogs.
00:07:08 ◼ ► And even if your blog wasn't big, the "social network of blogs," whether it was that stupid sidebar or trackbacks or whatever,
00:07:15 ◼ ► would be some popular blogger would read your blog, think it's interesting, and post it on their blog.
00:07:19 ◼ ► And that's how you got exposure through this sort of organic internetwork of independent sites.
00:07:28 ◼ ► It seemed like it was these smaller independent sites. But there was a social network. It just wasn't centralized.
00:07:33 ◼ ► So if Medium had gotten critical mass like Twitter did, their social network of blogs might have worked out better.
00:07:40 ◼ ► But it doesn't seem like they ever quite crossed that hurdle. So you post it to Medium, and it'll get more exposure than it would on your lone blog.
00:07:47 ◼ ► But it might not get as much exposure as it would get if your lone blog got picked up by one of the big bloggers back in the heyday of blogging.
00:07:54 ◼ ► So, I don't know. I know people want the old days of blogging to come back, and maybe it wasn't as great as we all thought it was.
00:08:02 ◼ ► But Medium is not bringing enough new stuff, nor is it bringing enough, in terms of quantity, enough numbers to make up for the fact that you're not accruing all the benefit of your work when you put it there.
00:08:18 ◼ ► I understand the appeal of having a push button. Here is your blog. It takes you all of two and a half seconds to create it. And you can post something here.
00:08:32 ◼ ► And Medium aesthetically, like, how can I describe this without you guys jumping on me? I agree that there's way too much chrome and decorations around the actual content.
00:08:42 ◼ ► But it looks aesthetically fine, like leaving that chrome aside, which is bad. It is real bad. But if you can leave that aside, like, this stuff looks good.
00:08:50 ◼ ► You know, the typography is fine for my eyes anyway, and I'm no expert, but it looks good. It's not offensive. It's not super janky. But golly, I just, it seems so bad to me. It just seems so, so bad.
00:09:03 ◼ ► Speaking of being offensive in typography, look at the HTML code snippet. Look at the curly quotes. Class equals incorrectly facing curly double quote.
00:09:19 ◼ ► Alright, so I'm sorry for that. I'm sorry for the rant, but I just needed to get that off my chest. So friends don't let friends use Medium. You know, and I don't think they're, no, they're not sponsoring this week, but I hear Squarespace isn't bad. Try code ATP or maybe ATP FM. May not be terrible.
00:09:32 ◼ ► Anyway, so we talked last week about how Google had deliberately crippled the video performance of Edge, Microsoft's browser, by putting a clear div on top of it, and so Microsoft thought, "Oh, that means we can't hardware accelerate. All our, everything goes down the tubes. The performance goes down the tubes, the battery life goes down the tubes, etc."
00:09:53 ◼ ► So Jeremy Norring, who worked for, who, oh, on Edge, said, "Well..." No, he did not work on Edge. It doesn't matter where he worked. He worked somewhere where this is relevant.
00:10:05 ◼ ► He, one way or another, he realized, "Oh, I needed to get keyboard shortcuts or some such working properly."
00:10:11 ◼ ► And the only way to interrupt Edge's very, like, grabby hands on the keyboard was to put a non-selectable div on top of the video. This is to, like, get keyboard shortcuts for video playback on whatever site he was working on.
00:10:24 ◼ ► So his point was, yes, this does have the net effect of benefiting Google in the sense that it destroyed Edge's performance, but it might have been that Google just needed to do that in order to get video stuff working the way they wanted it to. So it may not be as nefarious as we thought.
00:10:40 ◼ ► I think it was broken on IE, not on Edge. I know it's confusing with all the different things in the mix. But anyway, yeah, this was one of the many things that people who make websites have to do, a workaround for some browser.
00:10:52 ◼ ► And unfortunately, it's not always easy to conditionalize whatever weird workaround you have to do to precisely the situation where it's needed, right? So you could do user agent sniffing and say, "When they come with this browser, use this hacky workaround to make it work."
00:11:11 ◼ ► The problem with figuring out what that condition should be is you can't tell the future. So whatever you make it conditional on, whatever information you can glean from the browser, you're like, "Okay, but there's going to be another version of this browser that's going to come out in the future. Maybe next week, maybe next month, but surely there will be another one. Will that one also have the problem?
00:11:30 ◼ ► Should we make the condition so that it happens in every "version of IE" but then Microsoft comes out with Edge and Edge's user agent looks different than IE or maybe it looks like IE? It's not easy. Web development is not easy.
00:11:43 ◼ ► So you make a best guess and you say, "Okay, well, I'm going to put this condition on here, and if this is true, put a transparent div over the video player." And apparently, whatever that logic is also decides to put it over when Edge loads the page.
00:11:57 ◼ ► Again, it could be nefarious. We don't know that it's not. But I thought this article was good because it was a concrete example of someone doing the exact thing that Google did for non-nefarious reasons. So at least it shows that there is at least one legitimate non-nefarious reason for this to be happening.
00:12:12 ◼ ► Yep. It was surprising. But you know, the internet loves to get up in arms about things, and so here we are. Moving on, speaking of things the internet is up in arms about, Marco, let's talk about the 2018 Butterfly Keyboard Watch.
00:12:27 ◼ ► I did not bring this up. For the record, I want to state, I did not bring up the Butterfly Keyboard.
00:12:32 ◼ ► You're going to have thoughts about it. So apparently the space bar types double spaces occasionally. John, I think this was your work. Tell me about this.
00:12:38 ◼ ► You tried to, maybe I should have put it in title case. 2018 Butterfly Keyboard Watch. Does that help?
00:13:01 ◼ ► Well, no, because it's like, what was it? Like, Goodnight San Diego or something like that, where they put the errant question mark at the end and he totally got confused.
00:13:13 ◼ ► So anyway, tell us about the Butterfly Keyboard because we haven't talked about that enough on the show.
00:13:18 ◼ ► The 2018 Butterfly Keyboard Watch is something that I've been participating in and I think we all have to some degree, which gets to the question that we posed when this keyboard was first introduced.
00:13:28 ◼ ► This is like the 2017 and 2016 butterfly keyboards, except that it's got this membrane over it.
00:13:34 ◼ ► And the question is, does the membrane make it more reliable than the 2017 and 2016 butterfly keyboard, which had problems with debris?
00:13:41 ◼ ► If it doesn't make it more reliable, and if so, by how much? Or is it the same as the other ones?
00:13:45 ◼ ► Difficult to tell from anecdotal evidence, but pretty quickly we started getting reports that some 2018 butterfly keyboards are still having the same kind of problems with, you know, getting stuck or not typing when pressed or double typing or whatever.
00:14:01 ◼ ► My personal 2018 Butterfly Keyboard Watch started when I got a 2018 Butterfly Keyboard with the 2018 MacBook Air.
00:14:07 ◼ ► And about the same time that I was thinking about this, I saw a tweet from Paul and Storm, the great musicians, on Twitter.
00:14:20 ◼ ► And I saw a Reddit post linked about this, about the space bar putting in double spaces.
00:14:26 ◼ ► And I also saw double spaces appearing when you hit the space bar on my 2018 MacBook Air.
00:14:33 ◼ ► So, yeah, what's the, you guys don't know this, I need someone who knows some references somewhere, chat room.
00:14:41 ◼ ► Please mentally insert the "my watch has ended" speech from a Game of Thrones at this point.
00:14:46 ◼ ► My watch has ended. I have a butterfly keyboard and it types two spaces when I want it to type one sometimes.
00:14:57 ◼ ► All these 2018 keyboards have the membrane in them. All of them somehow have allowed some speck of dust to come through the giant gaping holes in the membrane and get underneath something in the space bar and cause it to double space.
00:15:17 ◼ ► It's almost over. Last season's coming up, so it'll be all done so you can just watch it all at once if you want to.
00:15:26 ◼ ► Well, it's one of those shows that you can't watch with kids in the room, so that may make it more difficult to watch.
00:15:34 ◼ ► So, anyway, the membrane, it might not have been that a speck of dust has gotten past the membrane.
00:15:58 ◼ ► Remember when, this is less of a problem with the current generation, but remember when the 2016s first came out,
00:16:04 ◼ ► they had huge problems with, like, the metal would just pop and snap with thermal expansion,
00:16:10 ◼ ► and sometimes the keyboard, the thermals of the keyboard itself would affect whether the keys would press properly or not.
00:16:21 ◼ ► Like, these keyboards are just horrendously engineered from the very start, and part of that was the dust ingress issue,
00:16:29 ◼ ► So it does seem, anecdotally, like the 2018s are failing a lot less than the other ones,
00:16:35 ◼ ► but they still have some of the same problems that the other ones had, because fundamentally, this design is problematic.
00:16:42 ◼ ► It's just a little bit better protected against its own horrendous design with the 2018 membrane.
00:16:48 ◼ ► And the space bar in particular is a very big key, and people don't think about it, especially in, you know, these days of modern keyboards,
00:16:55 ◼ ► they're totally abstract what's going on there, but if you have a very, very large key, like a long key like the space bar is,
00:17:01 ◼ ► the tricky part about that is you have to make it so that if you press on the very, very left edge of the space bar,
00:17:11 ◼ ► The whole space bar has to go down when you press way over on the edge of the space bar,
00:17:15 ◼ ► otherwise, you know, weird things happen, because it would feel weird if you pressed the space bar and it tilted instead of the whole thing going down.
00:17:21 ◼ ► The old, much more chunky mechanical, I don't know, it doesn't make sense to call it mechanical keyboard, but that's what people call them,
00:17:29 ◼ ► older keyboards with larger pieces that have sort of like an anti-roll bar, like from cars, underneath the space bar,
00:17:35 ◼ ► so when you press down one side of the space bar, this little metal anti-roll bar would pull down the other side of the space bar,
00:17:40 ◼ ► so the whole space bar would go down. The butterfly keyboard and all these other much lower profile keyboards than the old chunky ones
00:17:47 ◼ ► have similar mechanisms, so that they don't feel like you're tilting it, so they don't feel like you're hitting an old trackpad and just one end is going down.
00:17:54 ◼ ► But the space and the tolerance is so thin in there, so if one key is going to have problems, like mechanical problems with that mechanism,
00:18:01 ◼ ► the up and down mechanism, it's going to be the space bar, because that is the most complicated and most difficult key to get right,
00:18:07 ◼ ► in terms of, you know, debouncing, which is presumably what's going on here, where when you press a key and it activates some kind of mechanical switch,
00:18:16 ◼ ► it doesn't just say, "Going down, going down, going down, make contact," and then a certain period of time making contact, and then stop making contact.
00:18:24 ◼ ► What actually happens is you hit the key, it goes down, it makes contact, then it bounces off, and then it makes contact again,
00:18:30 ◼ ► then it bounces off, and it makes contact again, and then it bounces off, makes contact again, and then it eventually settles down.
00:18:34 ◼ ► This happens very, very quickly. I think I've told this story before, one of the first assembly code projects I had to do in college was writing a driver for a keyboard,
00:18:43 ◼ ► and you had to debounce the keys, which means you'd wait for the key to make contact, and then you'd put in a delay to wait for the bouncing to stop.
00:18:59 ◼ ► Otherwise you'd hit A, A, A, A, A, A, A, A, A, because the computer is so fast, it sees those bounces.
00:19:13 ◼ ► And this is like a 16-bit, like 8 MHz processor, so you don't need a particularly fast computer to do this.
00:19:28 ◼ ► Anyway, I had no clue that either this was a thing or that you wrote a keyboard driver to prevent this.
00:19:33 ◼ ► Debouncing keys is a good first assembly project, especially since you're using such a wimpy computer.
00:19:41 ◼ ► You write in a 16-bit assembly, and it shows you that as slow as any computer could be, it's way faster than your brain thinks it is.
00:19:54 ◼ ► I had a lot of fun with it. We were taught what the instructor called structured assembly, which is like, "Imagine C didn't exist, and you were forced to write everything in assembly."
00:20:03 ◼ ► You'd come up with some set of rules that basically mimic the structures you have in C so you can keep things straight because you have a limited amount of characters for the labels and stuff.
00:20:12 ◼ ► And you'd just say, "Look, I consider this like, this is a for loop, this is a while loop, this is an if, this is an else."
00:20:18 ◼ ► And you'd come up with structures and standards, conventions for all the different labels. You make ASCII labels for different jump locations and stuff so you have a way to refer to them later.
00:20:28 ◼ ► And you just bang that out. Because you can type whatever you want. If you just type "assembly willy-nilly," you can just compare and jump and fly all over the place and you'll just make a giant mess.
00:20:36 ◼ ► You need some kind of structure, and the structure is you essentially invent the world's most verbose C, and you force yourself whenever I need a conditional.
00:20:44 ◼ ► I'm always going to do it this way. The label's always going to be like this, and this jump target is going to look like this, and whenever I have to do an if-else, I'm always going to do it in this order.
00:20:51 ◼ ► Because you can do them so many different ways. If you come up with a structure for it, you can write surprisingly long programs in assembly that still mostly look like line noise, but that if you look at them and squint, you can say, "Oh, I see the structure now."
00:21:05 ◼ ► At this point, I already knew Cs, but you can see the C code that this would be equivalent to, even though it's like 100 times long and everything is a bunch of three or four little letter sequences that don't make any sense if you don't know the instruction set.
00:21:19 ◼ ► I asked only because when I was in school, and I also did computer engineering, we had a single course where we were doing assembly programming on, I think it was an HC-11, and we were paired with a partner, and I don't remember if it was required or just happenstance, but I was paired with a EE, and I was a computer engineer.
00:21:37 ◼ ► And I loved doing this program. I don't think I would want to do it for a living or anything like that, but I also had an incredibly good time with it.
00:21:45 ◼ ► Now, unsurprisingly, my assembly was nowhere near as structured as yours was. It was a hodgepodge.
00:21:50 ◼ ► But it was so cool and so fun writing, like, I think we did, I wouldn't call it a device driver, but we like drove, what is it, like an eight segment LED? You know what I'm thinking of?
00:22:03 ◼ ► Yeah, yeah, okay, that's what it was. You know what I'm thinking of, where you could basically display numerals.
00:22:07 ◼ ► I did like a basic calculator. It was really, really cool and really fun, and I kind of miss it, even though I don't want to ever really do it again either, you know what I mean?
00:22:19 ◼ ► Yeah, so I don't, anyway, the debouncing thing, I don't think that's, you know, what's happening here.
00:22:24 ◼ ► I think what's happening here is a mechanical issue that is much chunkier than a typical debouncing would be.
00:22:30 ◼ ► So the debouncing, you know, firmware or whatever in the keyboard is probably working fine, it's just that these keys are doing a thing where they're going through the whole initial hit and bounce and settle, and then hitting a second time, and you're getting two spaces instead of one.
00:22:46 ◼ ► Because, again, debouncing keys is not a new technology, and it's something that is so transparent that kind of like the way DRAM refreshes itself constantly, I think we've talked about previously on this show, or maybe another one that is terrifying once you know that it happens, but if you don't know, you just think RAM is magic and you write values into it and get it back out.
00:23:04 ◼ ► It's the type of thing that you can mostly take for granted, so I'm assuming whatever is going wrong with the keyboard is in fact a mechanical issue like the ones Margo described, you know, either it's debris or metal fatigue or a plastic thing breaking or something bending when it shouldn't or something expanding thermally or contracting thermally.
00:23:20 ◼ ► Bottom line is, it's a spacebar, two spaces appear, it makes everybody sad, including me.
00:23:25 ◼ ► So, and you know, one of the problems I've had with these keyboards too is like, the spacebar and the up and down arrow keys don't feel right.
00:23:35 ◼ ► Like it seems like whatever the mechanism, you know, the waterfall mechanism here, it seems like the design of it is optimized for the size of the regular, you know, most of the keys in the keyboard are like kind of one size, and if you make a key that's substantially larger or smaller than those, like in the case of the spacebar being larger, same as like, you know, shift and return, those are also a little bit big too.
00:23:54 ◼ ► Those feel weird, they don't feel right, and I've also had the same problem with the up and down arrows as with little half height keys, they don't feel good, they feel kind of weird, sometimes it doesn't feel like they're activating correctly, like it just adds to the massive pile of problems this keyboard has, and my still very strong opinion that this never should have shipped, let alone been brought to the entire product line.
00:24:17 ◼ ► But hopefully they have heard this message by now. God, I hope we get a new keyboard. Anyway, so I'm curious, John, are you going to get yours serviced? Like are you willing to go without this computer and not have it for a day or seven?
00:24:32 ◼ ► I'm going to go with the tried and true Casey technique of tilting it at 75 degrees and blowing air into it. 75? 73? I forget, I have to look up that techno to make sure I get the right angle, otherwise this will obviously not work.
00:24:44 ◼ ► Does it matter? Spin around in a circle, do any angle you want, it's not going to make a difference, it's not going to really fix the problem. Yeah, I don't know. Make sure you defrag your hard drive too, that'll really fix it.
00:24:52 ◼ ► Oh man, savage. I mean of all the keys that are to go wonky, the space bar is one of the least objectionable because two spaces are not as egregious as two of any other letter, you know what I mean?
00:25:05 ◼ ► No, it's wrong. Input devices should be 100% reliable, period. All the time. I agree, but I'm just saying you can limp along with essentially a broken computer for much longer with a space bar that doubles than with another key that doubles.
00:25:18 ◼ ► I would rather have Casey's iMac that turns itself off like once a day randomly than a keyboard that's unreliable. Not for years. Yeah, it is 75 degrees by the way. Good luck with that. I'm sure it'll make a big difference.
00:25:33 ◼ ► We are sponsored this week by Mack Weldon, better than whatever you're wearing right now. I'm currently wearing four articles of Mack Weldon clothing.
00:25:43 ◼ ► And every day I'm wearing at least one and usually at least three because Mack Weldon's clothes are just so good. I discovered them a couple years back before they were a sponsor.
00:25:55 ◼ ► And I recently, fortunately I've changed clothing sizes. I had a diet, it went well, so I had to buy a lot of new clothes. And I went with all Mack Weldon for all of my basics.
00:26:05 ◼ ► Underwear, t-shirts, even a lot of my socks, a lot of my pants even. They have such good stuff at Mack Weldon, I'm incredibly happy with it.
00:26:14 ◼ ► And every day I'm wearing at least one thing from them. It's that good. In the summertime I cannot recommend their silver lined t-shirts enough because it's naturally antimicrobial which means in the summertime you can get all hot and sweaty and you somehow don't stink.
00:26:28 ◼ ► I don't know how this possibly works as well as it does, but it does. In the winter I really enjoy their long sleeve selection.
00:26:35 ◼ ► From the basic long sleeve Pima tee to the warm knit which is like a waffle knit shirt. And I even discovered the Merino shirt which every time I wear this it gets tons of compliments.
00:26:45 ◼ ► I call it my tomato shirt because the color name is called sun dried tomato. It's just, all of their stuff is so good.
00:26:51 ◼ ► I can highly recommend everything I've tried from Mack Weldon and at this point I have tried almost everything they sell.
00:26:57 ◼ ► And I've usually bought more than one from everything. I even bought long underwear this winter because I was cold and I learned that it's just basically a waffle shirt for your legs and it's amazing.
00:27:07 ◼ ► So I'm very happy with Mack Weldon. I hope you will be too. And if you try them, they guarantee this, they want you to be comfortable.
00:27:14 ◼ ► If you don't like your first pair of their underwear, if you try it, you can keep it and they will still refund you no questions asked.
00:27:20 ◼ ► So see for yourself for 20% off your first order. Visit MackWeldon.com and use code ATP at checkout.
00:27:32 ◼ ► Thank you so much to Mack Weldon for replacing my entire wardrobe and for sponsoring our show.
00:27:42 ◼ ► Now Marco, you also have some thoughts on input methods, but in this case it's for virtual keyboards. Tell me about your iPad.
00:27:55 ◼ ► Good luck, yeah. So I added this little teaser to the show notes here that I wanted to talk about my fingerprinty iPad and a solution that somebody recommended to me.
00:28:03 ◼ ► And I'm sorry, I forget who recommended this. I believe it was somebody on Twitter a few weeks back.
00:28:08 ◼ ► But anyway, somebody on Twitter recommended that to alleviate my complaint about my iPad Pro always being incredibly, visibly fingerprinty and just really gross looking.
00:28:19 ◼ ► And even if you wipe it on your jeans, it doesn't really take it all off. The iPad Pro just has this massive fingerprint attraction problem ever since the pencil has existed.
00:28:27 ◼ ► And so somebody recommended that I use a screen protector and they specifically recommended the PaperLike, which is at paperlike.com or paper.me, I think both.
00:28:38 ◼ ► And this is like this fancy German screen protector and it tries to simulate the texture of paper and the point of it is to make the pencil usage feel better, to make it feel more like you're drawing on paper.
00:28:50 ◼ ► But it happens to also be a pretty nice screen protector. So I ordered it and I said, you know, let's see how this looks. Let's see how it actually works.
00:28:57 ◼ ► And it has been a long time. I actually, I don't know if I've ever talked about this because it was before this show existed, but I used, on my first three iPhones, yeah, until the iPhone 4, until Retina.
00:29:19 ◼ ► I mean, I didn't think they were terrible at the time. And so the reason why, so keep in mind back then, first of all, there was no oleophobic coating back then. That was a feature that came later.
00:29:32 ◼ ► Second of all, the reason I like them, and I was able to apply them perfectly, like there was never, you know, you wouldn't like see dust or bubbles or anything like that under them.
00:29:41 ◼ ► So it looked fine. The only thing you would notice is the way that it kind of adds that rainbow noise to how it looks. It looks kind of blurry.
00:29:52 ◼ ► But pre-Retina, you wouldn't really notice that. It wasn't a big deal. It just made it look like a matte screen.
00:29:58 ◼ ► And the reason I like them is because it changes the feel completely. It's much easier to move your finger across it.
00:30:06 ◼ ► Like it has a totally different feel. And I actually greatly prefer the touching feel of a matte screen protector.
00:30:15 ◼ ► And I keep saying matte here because the glossy ones I think are not great for lots of reasons.
00:30:21 ◼ ► But the matte ones, and those mirror ones, oh god forget it. But the matte ones seem, I really enjoy the feel.
00:30:28 ◼ ► So I got this from my iPad Pro thinking, you know, I know it's not going to look great.
00:30:32 ◼ ► Like the screen contents are not going to look great. But I do want to try it and see like do I prefer it?
00:30:38 ◼ ► You know, how does it affect the fingerprintiness of it? And can I actually tolerate the look of it?
00:30:50 ◼ ► But there's some interesting things about it that I thought were worth bringing up and telling you about.
00:30:54 ◼ ► So first of all, they have this amazing like detailed procedure of how to install this thing without getting dust on it.
00:31:01 ◼ ► You know, trapped under it. And of course I got some dust trapped under it anyway because it's pretty much impossible not to.
00:31:07 ◼ ► To their credit, they go through crazy lengths to give you like little tools like different wipes and things.
00:31:16 ◼ ► And they had this whole video about how to do it in these little alignment stickers so you can do it real fast.
00:31:26 ◼ ► But I did get it on there without too many dust things under it. It's not perfect but it's close enough.
00:31:53 ◼ ► When you're displaying anything with a black background, like white text on dark background, that looks fine.
00:31:59 ◼ ► You don't really notice that it's there. But when it's displaying dark content on a light background, anything with a light background, you see that kind of like rainbow noise pattern in it.
00:32:10 ◼ ► And I would love to know, anybody out there, if you know if there are screen protectors that still have this like matte kind of feel,
00:32:28 ◼ ► I just want to know like if anybody knows exactly what I'm talking about with this rainbow noise, and if you know of anything that doesn't have it, please let me know.
00:32:40 ◼ ► And if it's bumpy, that means you know, we got the whole refraction of light going through, like it's basically a bunch of little lenses.
00:32:48 ◼ ► I can't imagine it being both bumpy and not refracting light and separating the colors.
00:32:53 ◼ ► Yeah, I mean you're probably right. There's probably a good reason why all the ones I've ever seen have this problem.
00:33:04 ◼ ► Anyway, I do really love how it feels. It is way more pleasant to touch. I really love it.
00:33:11 ◼ ► It is, as designed, really nice with the pencil. Now granted, I hardly ever draw things with the pencil, so like it's not that much affecting me there.
00:33:19 ◼ ► But it is a substantial improvement in how the whole iPad feels. And it's a substantial downgrade in how light content looks.
00:33:29 ◼ ► And there's one other side benefit to it that I thought was very interesting that I actually like quite a bit.
00:33:35 ◼ ► It has a cutout for the Face ID sensor module, the little, you know, the notch. That's not really a notch on the iPad.
00:33:41 ◼ ► The only real problem I have with this iPad is that I'm always covering up the Face ID module when I pick it up. Always.
00:33:47 ◼ ► And it so happens I'm often picking it up when I'm trying to unlock it to start using it.
00:33:54 ◼ ► I am literally holding it wrong, even though I'm holding it in the most obvious way you could hold it when it's in Apple's keyboard dock.
00:34:07 ◼ ► It should start adding like carrot weather. It should be like, "Uh, dummy, you're covering the camera again. What do you want me to do with this FaceTime? On your palm? On your thumb?"
00:34:14 ◼ ► I firmly believe that the only major design flaw of this iPad that I have found yet, in pretty heavy use over the last couple of months, whatever it's been,
00:34:24 ◼ ► the only significant flaw is I think the camera's in the wrong place. If there's only going to be one, I think that's the wrong place for it.
00:34:30 ◼ ► It could even still be on that edge, but just like if you're looking at it in landscape, instead of being in the left center, it should be in the upper left.
00:34:40 ◼ ► I think that would cover more orientation. But then I don't know if that could get in the way of the speakers. I don't know.
00:34:45 ◼ ► There's probably a reason why they centered it, but I'll tell you what, man, it's really annoying.
00:34:52 ◼ ► Or two. Put one basically where the pencil charger is. Because there's plenty of space up there. They could totally do it.
00:35:04 ◼ ► But with the screen protector on, because it has a cutout that goes around the camera area, you can see where the camera area is.
00:35:14 ◼ ► And so I have found that it's actually easier to avoid it because you can see it. It's outlined. It's clear as day where it is.
00:35:23 ◼ ► You could have put yellow pinstripe tape around the thing with the big X that says "Don't put your hand here."
00:35:30 ◼ ► No, but I've found that it's actually been easier to avoid the camera, the housing covering up in my hand, because I'm seeing where it is now.
00:35:42 ◼ ► So I actually really like the feel, the pencil behavior, and the cutout of the notch of the screen protector.
00:35:54 ◼ ► So I don't know if I'm going to keep it on long term, but I'm keeping it on for a while because it really does feel very, very good.
00:36:03 ◼ ► Well, it's nice to see you embracing the Naked Robotic Core by completing Apple's product by adding to it the accessories that you want that make it the product you want it to be.
00:36:12 ◼ ► It's already covered in their accessories. I've got the keyboard, I've got the pencil. What else do you want?
00:36:16 ◼ ► You have to accessorize and keep adding stuff to it. What is it that makes the iPad for you?
00:36:27 ◼ ► Because you can't take stuff off if Apple adds it. But if you want to add stuff, you can put a big clear thing over the front of it and a big sticky thing on the back.
00:36:34 ◼ ► A magnetic keyboard stuck to it, a sticky pencil on the top of it, and stuff stickers on the back of it and put some sort of USB-C dongle hanging off the end of it.
00:36:51 ◼ ► I was saying it's gross because I think it kind of ruins the aesthetic. And also, depending on how big it is and what shape the device is that year, you might also feel the edge and the edge can collect dust on it.
00:37:04 ◼ ► But more importantly, as you pointed out, getting covered down in the thing without any dust or bubbles gets much harder as the device gets bigger.
00:37:13 ◼ ► So putting it on a 3.5 inch phone, you're really super careful, maybe you can pull it off, especially if they send you a kit to do it that includes an inflatable clean room.
00:37:27 ◼ ► And I can't imagine doing it with something as big as an iPad. That's extremely difficult.
00:37:31 ◼ ► But it brought to mind the most difficult thing I've ever had to do related to putting thin, clear film down precisely.
00:37:41 ◼ ► It's not having to do with iOS devices, it's having to do with Lego. It's the most difficult thing I've ever had to do assembling a Lego set.
00:37:47 ◼ ► I got a really cool X-Wing, I think it's like the Ultimate Collector's Edition X-Wing Lego set.
00:37:53 ◼ ► And it has a special piece, of course, for the X-Wing canopy. I think it's a special piece, I don't know, maybe they used it other times.
00:38:06 ◼ ► Like the little Lego mini-figs, that's not what scale it is. It's scaled practically for like a 6-inch action figure doll thing to go in the seat.
00:38:15 ◼ ► It's very big. So the canopy is very big. And the canopy is 100% clear, but of course the X-Wing has the metal struts along the edges of the canopy.
00:38:27 ◼ ► So the set comes with the clear canopy, and it comes with stickers to cover the top, the right, left, and the back of this thing.
00:38:37 ◼ ► And the stickers are entirely clear. It's kind of better than what they could have done.
00:38:42 ◼ ► They could have said, "Really, all you need are dark grey lines along the edges of this shape."
00:38:48 ◼ ► So we could have given you a very, very skinny pinstripe that you put along the edges, but instead they gave you a piece that covers the entire side.
00:38:56 ◼ ► And it's entirely a clear sticker, except for the very, very edge, which is like dark grey.
00:39:01 ◼ ► And you have to put on four of these stickers, and they all have to meet exactly at the sharp edges of this thing.
00:39:11 ◼ ► And the instructions just show, like, "Peel the edges off the sticker. Put the edge down. Peel off the backing. Lay the sticker on."
00:39:21 ◼ ► And they know, like, you look in the set, and it comes with two sets of these stickers, because they know you're going to try it.
00:39:27 ◼ ► You're going to screw it up. You're going to scrape the thing off painfully, because those LEGO stickers are not easy peeling stickers.
00:39:38 ◼ ► So we'll put a link in the show notes to have some photos on Instagram of the instructions, which are just like, I mean, they look straightforward.
00:39:46 ◼ ► If you've never put stickers on a LEGO thing before, or if you've never tried to put a large clear sticker on something precisely aligned before,
00:39:52 ◼ ► and then the finished product, which looks perfect in the picture, but I assure you it is not perfect.
00:40:03 ◼ ► So I don't envy anybody who has to put a giant clear projector on any of their devices.
00:40:08 ◼ ► I know a lot of people who work in stores become, like, experts at it and have a whole system, but really the environment is such an important aspect of it,
00:40:15 ◼ ► where unless you have some system that actually filters out microscopic dust particles or blows them out just in time or something,
00:40:26 ◼ ► That looks terrifying trying to apply that. And you're right, that's probably easier than what Marco did with his iPad.
00:40:32 ◼ ► I mean, the thing with the iPads is, like, with the sticker, you don't get a chance to reposition.
00:40:38 ◼ ► You don't even get a chance to peel back with these LEGO stickers. It's not like you, like, start to put down and realize you made a mistake and can reverse.
00:40:44 ◼ ► You can't even reverse with these things, because if you try to reverse, it will, like, leave the sticky stuff down.
00:40:48 ◼ ► The LEGO stickers are no joke. They are not, like, static, cling, smooth, like, stickers that you can peel on and peel off.
00:40:55 ◼ ► They are, like, one and done. Once these two surfaces meet, that's it, unless you're in for an hour of scraping and peeling and scrubbing.
00:41:02 ◼ ► I mean, the thing with the iPad screen protector, like, you basically just kind of have to do your best to minimize dust and flaws,
00:41:09 ◼ ► but not care when you get them, because you will. Like, I tried. I fought all their instructions.
00:41:16 ◼ ► I had, like, this—I was using one of my video lights kind of diagonally across it so I could see all the dust, like, as it would fall onto it,
00:41:22 ◼ ► and use all their different cloths and tabs and everything. And, yeah, I still got dust under it.
00:41:27 ◼ ► You know, not a lot of dust. There's, like, two or three visible specks, but, you know, that's—it's not, like, causing these giant bubbles in it.
00:41:34 ◼ ► It's, like, if you look in just the right place, you might notice it, but compared to all the rainbow noise, it's all over it anyway,
00:41:39 ◼ ► and the fingerprints that are still visible on top of it, it's, you know, it kind of just blends in with everything else.
00:41:44 ◼ ► So, yeah, the solution is stop caring so much. I kind of feel like that's kind of, like, the secret to happiness in this entire area about the iPad,
00:41:52 ◼ ► of, like, it's never going to look perfect. Even if you just use it totally stock, it's going to be covered in fingerprints if you're using it at all.
00:41:58 ◼ ► So, an iPad never looks perfect. So, if it's going to be imperfect, it might as well be imperfect in ways that you like.
00:42:10 ◼ ► We are sponsored this week by Backblaze, unlimited cloud backup for Macs and PCs for just $5 a month.
00:42:18 ◼ ► I use Backblaze. I've used it since before they were a sponsor, because here's what happened.
00:42:22 ◼ ► I realized that with all the different risks we have in the world for what can happen to our data and how important our data is to us,
00:42:29 ◼ ► I wanted some kind of offsite backup. Just having hard drives connected to your computer, that's great,
00:42:34 ◼ ► but what's even better is to have the extra insurance, the extra coverage of an offsite automated backup.
00:42:41 ◼ ► And so I looked into cloud backup providers, and I'll tell you what, I tried a number of them.
00:42:45 ◼ ► Backblaze was the one I stuck with, and this was before they were a sponsor, because it was just the best.
00:42:50 ◼ ► It was simply the best, and I have tried other ones even since, and I've reevaluated, and Backblaze is still the best.
00:42:57 ◼ ► You can get Backblaze yourself. You can see what I'm talking about at backblaze.com/atp.
00:43:02 ◼ ► Here's how this works. It will back up whatever your computer has on the computer itself,
00:43:07 ◼ ► and if you want to plug in external hard drives to the computer, it will back them up too.
00:43:16 ◼ ► So if you have one computer with 10 terabytes of drives connected to it, it's still $5 a month.
00:43:24 ◼ ► You can access your data anywhere in the world, because not only can you do file restores if you lose data,
00:43:31 ◼ ► if you have a hard drive failure or something, you can also, from your phone, or from a laptop on the other side of the world,
00:43:39 ◼ ► So you basically have access to all your files wherever you are all the time. It's pretty great.
00:43:44 ◼ ► Tons of other people trust Backblaze too. Over 500 petabytes of data under their management.
00:43:49 ◼ ► Just see for yourself, trust me, you want online backup for peace of mind and for all sorts of risk assurance,
00:43:55 ◼ ► you want online backup. And if you're going to use online backup, Backblaze is by far the best.
00:44:05 ◼ ► You can see for yourself, get started today. If you're going to your parents' place or something for the holidays,
00:44:10 ◼ ► set it up for them. Give them the gift of peace of mind. Even if they don't appreciate what it is,
00:44:15 ◼ ► you'll know you did it. You'll know that their data is safe. And if your data isn't safe yet, please, please get online backup.
00:44:21 ◼ ► And if you're going to get online backup, I highly suggest you get Backblaze. Once again, backblaze.com/atp.
00:44:26 ◼ ► See for yourself today. Thank you so much to Backblaze for keeping my data safe and for sponsoring our show.
00:44:35 ◼ ► Let me butcher somebody's surname that's my favorite pastime. John G. — G. N. Andrea? Does that sound about right? Hopefully.
00:44:44 ◼ ► Yeah, yeah. This was the gentleman who was hired from Google who was head of machine learning and artificial intelligence.
00:44:50 ◼ ► He has been named to the company's executive team as senior vice president of, guess what,
00:44:55 ◼ ► machine learning and artificial intelligence strategy. And this has happened since he joined Apple in April of this year.
00:45:01 ◼ ► So stars on the rise and door. They really, really, really want Siri to not suck so bad.
00:45:11 ◼ ► And it was a good hire from Google who's, you know, acknowledged leader in this area. Apple's not doing so great.
00:45:16 ◼ ► And I don't know. I mean, I don't really know how corporate politics work at this level.
00:45:23 ◼ ► Like if you're hired and then quickly put up to a VP, is it because that was the deal coming in?
00:45:31 ◼ ► If you do a good job, you could go to VP. Or did he do such great work that they are just being courted from somewhere else and they don't want them to leave.
00:45:39 ◼ ► So they bring them up or are they just trying to show that machine learning is important to the company?
00:45:43 ◼ ► Like, is it all those things combined? I think this is a good sign because having this,
00:45:50 ◼ ► you know, having Apple acknowledge that machine learning and artificial intelligence is important to its future is good.
00:45:58 ◼ ► It's better than not doing this or not having anybody reporting and any vice president whose job it is to do this thing.
00:46:05 ◼ ► But on the flip side, I'm not really as interested in Apple showing how important something is.
00:46:12 ◼ ► I just want to see their products get better. And it was a story going on recently about one of those showdowns between like Siri
00:46:19 ◼ ► and Amazon's Echo and Google and all that other stuff where they ask it a series of questions and then give them a score based on what percentage they get right.
00:46:27 ◼ ► I've never liked those quizzes just because you can make any of them win depending on how you phrase the questions and which questions you ask.
00:46:33 ◼ ► If you ask 20 questions about something that one service knows way better than the other, but don't ask anything about that other service's strength, it can really skew things.
00:46:42 ◼ ► But anyway, they ran this test and the recent results show Siri gaining serious ground over where it used to be.
00:46:49 ◼ ► It's still in second place, I think. I think it was behind Google but ahead of Amazon barely, but it has improved by like 20% since last year.
00:46:57 ◼ ► Anecdotally, I think Siri is improving, but as I think Gruber wrote about this recently, as many people have pointed out,
00:47:04 ◼ ► lots of times you ask your HomePod to do something and it gives you the brush off and says, "But you can try that on your phone."
00:47:16 ◼ ► HomePod, you're gigantic, you weigh a ton, you're plugged into the wall. There should be nothing my phone can do with Siri that you can't also do.
00:47:28 ◼ ► So Siri gets points off for being uncooperative in areas where it thinks it's not supposed to be doing something or currently can't do something and tells you to go talk to your phone instead.
00:47:38 ◼ ► Which may or may not be fair. Isn't Siri on the phone too? Are you testing Siri or are you just testing the HomePod?
00:47:43 ◼ ► But the bottom line is, if you're just yelling something into the air, chances are good that the HomePod will hear you way better than your phone, even if you have Hidengus enabled on your phone.
00:47:52 ◼ ► Because who knows where your phone is and it could be in a pocket or it could be in another room and it just doesn't have all the microphones and the super duper always listening power of whatever the hell the HomePod has going on.
00:48:04 ◼ ► So I'm glad that Apple is taking this seriously, but show me the results. Show me the money.
00:48:17 ◼ ► Can you play podcasts on it? Because I know you had your iPad in the kitchen and it's like, "Oh, if I'm in the kitchen, I want to listen to podcasts." How easy is it for you to say, "You know what? I want to listen to podcasts, but I want them to play on the HomePod."
00:48:31 ◼ ► I bring this up because that's been my experience too with all these speakers, not just Apple. I have a bunch of speakers spread throughout my house and every time I want something magical to happen, like, "Oh, I was listening to a podcast, an overcast on the car ride over," and I get home, I want to continue listening to it in my house.
00:48:49 ◼ ► I start thinking about how I could possibly make that happen and I short circuit myself and say, "You know what? Let me just AirPlay it. Let me just use a Bluetooth speaker." I take all these supposedly intelligent speakers and I say, "Oh, just forget it. I'm just going to pretend you're a dumb speaker that understands how you can send audio from my phone through the speaker."
00:49:08 ◼ ► That's it. And that's not as good. I want to be able to say something in the air and have some podcast player have it understand it. At this point, I have so little faith that I don't even ask the HomePod to play playlists in my iTunes account because I think, "Oh, I can't forget. Does this not support smart playlists and it only supports other ones?"
00:49:27 ◼ ► I keep forgetting which of Apple's things only support albums but not smart albums or playlists but not smart playlists or I just think it won't know what the heck I'm talking about or I won't know how to phrase it or whatever, so I just don't even try, which is bad on me for not trying but bad on Apple for making me think that it's not going to work.
00:49:49 ◼ ► Actually, a little sidebar here. I recently got a Sonos amp because they're a new speaker amp because it has AirPlay 2 support and it basically is a Sonos unit that drives two speakers and can accept input from your TV or from a line-in or from AirPlay 2 and automatically switch between those things, which I absolutely love. It's awesome.
00:50:13 ◼ ► Anyway, so I now have AirPlay 2 in two places. I have actually three if I include the AirPlay Express in my office, which I keep trying to make work and keeps not working so well.
00:50:23 ◼ ► And so now I have multi-room output that is supported by both the Amazon Echo family of devices, which I also have a few of, and AirPlay 2.
00:50:32 ◼ ► I can directly compare and in some cases use the same speakers between the AirPlay ecosystem and the Sonos/Eloquia ecosystem. Neither one of these systems is reliable about multi-room audio for me.
00:50:46 ◼ ► I have tried lots of different ways to phrase things. I've tried different ways to name the devices or the rooms or whatever. It's really hard to get these systems to work reliably and I don't know who's at fault with, you know, between like the Amazon and the Sonos system.
00:51:01 ◼ ► I do definitely know who's at fault with the AirPlay 2 ecosystem though because the Apple devices are the master control devices there, so I know what's going on there.
00:51:12 ◼ ► And it is so hard. Like if I say, "Hey, dingus, play so-and-so in living room and kitchen," or something like that, no combination of phrasing and anything like that is working reliably for me.
00:51:26 ◼ ► Sometimes they'll work sometimes, not other times. One set of phrasing will work on one family of devices but not the other one. Sometimes weird things happen, like it'll say it's playing and it just doesn't.
00:51:35 ◼ ► Like it'll say it's playing in a room and you go to that room and it's not playing there and you go to the app and it's like, "I have no idea what you're talking about. You didn't ask me to do anything."
00:51:42 ◼ ► Like it's really messed up and I don't know if anyone else has similar issues as me with multi-room stuff, but man, it's rough to do it by voice.
00:51:51 ◼ ► If you do it from a device, like if you open up the Sonos app and do the Sonos version of it there, or if you open up Control Center or the Music app and you go to the AirPlay menu and you do it from an iOS device, like you pick out the devices there, works fine.
00:52:07 ◼ ► But by voice it's still a mess. Anyway, so going back, that sidebar is now over, going back to the master branch here.
00:52:15 ◼ ► The main issue I have with Siri, in addition to some of the reliability stuff, is that it's still just so much slower than Alexa.
00:52:23 ◼ ► Like in just how long it takes when I say something to get either a response or the thing happening, it's just slower.
00:52:32 ◼ ► And it's slower enough that if you are accustomed to the Amazon ecosystem, you really notice that delay.
00:52:39 ◼ ► I don't know if that's because of some kind of fundamental engineering flaw in Siri or some kind of implementation detail that they can easily fix, but changing leadership is a really good sign that they're taking this more seriously.
00:52:52 ◼ ► Elevating, I mean they didn't technically change the leadership, he was already there, but like elevating someone who's relatively new and who seems pretty good at this stuff to a higher title that is more important to the company and has more public visibility and public scrutiny, that's probably a good sign.
00:53:08 ◼ ► And so I hope this is a result of them taking Siri more seriously, and I hope that, clearly what we've seen over the last seven years, what we've seen is that Siri seems to have fundamental flaws and limitations that make it slow, inconsistent, unreliable, and in spotty areas that seem to move around a lot, kind of dumb.
00:53:36 ◼ ► And they seem to have fundamental problems with it that have prevented them from improving it for all these years in strong ways in these areas.
00:53:48 ◼ ► And now that we have both the Google, whatever, and Alexa, we can see this can be much better, this can be much faster, and much more reliable.
00:53:59 ◼ ► And in some cases smarter, it seems like the smart part is kind of catching up, but it's still not faster or more reliable. And that I really hope they work on, I really hope they're taking it very, very seriously as some kind of major effort to like redo Siri because it really needs it, it's just too slow and too inconsistent.
00:54:18 ◼ ► What I worry about is that this new person is actually going to be concentrating mostly on the things that we haven't seen released yet, whether it's like artificial intelligence for making the iPhone camera better, or machine learning for doing proactive stuff on the phone about notifications.
00:54:34 ◼ ► There's lots of areas for machine learning and artificial intelligence that are forward looking, that are not simply enhancements of last decade's war over voice personal services. Surely, of course, Siri Canon does need to get better, like we just talked about.
00:54:50 ◼ ► If you are the vice president level, you're already seeing all this stuff. So part of it is yeah, let's make Siri better, but part of it is also, let's make the next great thing with the next much more exciting thing, like maybe voice assistants of this style or old hat, and maybe it's a successor to Siri, you know, is one of those things.
00:55:07 ◼ ► I don't know. I like that's the thing about being a vice president, how you balance all, you know, you're in charge of all this stuff. How do you balance, you know, what to concentrate on more than anything else? Like what do you put your attention on? If you have any budgetary control, how much money do you put towards it?
00:55:23 ◼ ► Because surely Apple has been and continues to do tons of stuff with machine learning and artificial intelligence that is unreleased entirely and stuff that they did that was never released, like, or still hasn't been released, like all the stuff they're supposedly doing with the car stuff, the self-driving car.
00:55:38 ◼ ► That's all machine learning and artificial intelligence stuff hasn't seen the light of day. So that's just totally invisible to us. And presumably that project is still ongoing, and I'm not sure what form it takes, but there's some portion of the time of this newly minted vice president's going to be spent on that, right?
00:56:00 ◼ ► To go back a step to our complaining about Siri, I would like to pile on a complaint about Siri, please. When I use my watch, which I know is the thing that Marco is unfamiliar with, and mostly John as well, but for me, the one Apple watch user of the three of us, I will not infrequently go to do something using Siri on my watch, typically sending like Aaron a message or something like that.
00:56:23 ◼ ► And I will say, "Hey, dingus, tell my wife, please bring up a pacifier," or something like that. And often, well, maybe not often, but probably a third of the time, I see, "Hold on. I'll tap you when I'm ready." And nothing.
00:56:42 ◼ ► This is a cellular Apple watch where I have not disabled the cellular connection. Often this happens when my phone is not nearby. And what I think is the problem is that it spends way too much time trying to find my phone.
00:56:58 ◼ ► But this happens a lot in the house. And so even let's suppose if my phone was across the house, and it's nowhere nearby, it is understandable for it to take a second or two, probably just one, to see if it can talk to my phone.
00:57:13 ◼ ► But after that, I don't care how power expensive it is, I want that Wi-Fi radio powered on and I want it using my home's internet connection.
00:57:21 ◼ ► And if that doesn't work within a second, assuming you can power the radio on that quickly and get a connection that quickly, I want it to go to the damn cellular network. That's why it's there.
00:57:29 ◼ ► And the fallback procedure is infuriating because it doesn't fall back, it falls down every time. And it drives me insane.
00:57:39 ◼ ► And it makes this watch, which I otherwise quite like, I mean I'm still on my Series 3 and I otherwise really, really like it, but it just makes it infuriating. And it's these sorts of paper cuts that make me so unwilling to give Siri a chance.
00:57:52 ◼ ► Whereas it is far more rare for the Amazon lady in a tube to fail to hear me or to take forever to answer my question.
00:58:05 ◼ ► The powering out of the radios, unfortunately, I have to imagine cannot be done fast enough to satisfy your demand for action now. And they can't leave them on all the time because the batteries would be dead.
00:58:15 ◼ ► So it's, you know, I feel like this is a technology problem that is not easily solvable at this point. The only thing I can think of is, I have to add, I'm sure Marco knows more about this and I'm sure this is a ridiculous idea, one of those arm swinging powered battery chargers to your watch to make the watch even thicker.
00:58:33 ◼ ► You know what I mean? Like there is a lot of energy going to waste there, but unfortunately to make one of those it would actually...
00:58:41 ◼ ► There's a lot of energy. The problem is capturing it. I don't think you can actually capture enough to appreciably charge any kind of battery of any reasonable size.
00:58:49 ◼ ► You can capture enough to perhaps wind the world's tiniest, most delicate spring in a mechanical watch, but probably not too much charge. But there's a huge amount of energy, you're just not capturing it.
00:59:02 ◼ ► There was an episode of the podcast, of John Chigie's podcast called "Pragmatic" about this a while back. It'll take me a little while to find it. It was from a few years ago when the Apple watch first came out.
00:59:13 ◼ ► I'm pretty sure he did an episode specifically about like, could you power a smartwatch with like an automatic rotor the way mechanical watches power themselves?
00:59:21 ◼ ► And yeah, basically the answer is no, they don't produce anywhere near enough current to do it.
00:59:26 ◼ ► You could do it. You just have to have a huge weight on the thing. Like the thing is, if you actually were harvesting enough energy, it would be like regenerative braking for your arm. You would feel it. That would not be comfortable.
00:59:35 ◼ ► It would be like trying to move your hand around while it has a heavy gyroscope spinning really quickly. Have you ever had a gyroscope? Did you guys have gyroscopes as kids?
00:59:43 ◼ ► Yeah. Like you get one of those in your hand and like a big heavy one spinning really, really fast, then you move your hand, you feel the resistance. That's what it would feel like if your watch was actually harvesting the energy of your arm. You'd have to feel it.
00:59:56 ◼ ► And even if you felt it, it would still take a lot of muscle to put the gyroscope. It would be like those radios or flashlights that you crank. You got to crank and you feel the cranking. Like it's hard work.
01:00:08 ◼ ► So how do you feel about that Casey? If before you use Siri, you had to turn on, difficult to turn dial 10 times, but then it would instantly respond when you talk to it.
01:00:16 ◼ ► If it was, I mean, I know you're joking, but if it guaranteed it to respond and even in this hypothetical perfect world guaranteed it to give me a reasonable response, not like, "Hey, go talk to your phone because I'm useless."
01:00:28 ◼ ► Then I would consider it. It would almost be worthwhile. But the problem is a lot of the times when I'm invoking this, I'm saying, "Hey, dingus." Like even to my watch, I'll say, "Hey, dingus, you know, tell my wife, blah, blah, blah."
01:00:38 ◼ ► And that is just not a time when it's conducive to wind. And I know that you're being facetious, but it's a reasonable thought exercise. And in certain circumstances, heck yeah, I would wind my watch 10 times in order to guarantee a response from Siri.
01:00:51 ◼ ► Yeah. This kind of like the always-on watch face is a problem we have to wait for the technical solution to become feasible. Like an always-on watch face will be better, but can't do it because of battery.
01:01:02 ◼ ► Always-on Wi-Fi when you're at home will be better, can't do it because of battery. So, you know, time marches on and hopefully we will cross both of those milestones eventually.
01:01:11 ◼ ► Alright. Speaking of, we were talking about earlier how the internet loves to get upset about things. And I really didn't want to give this time on the show because I think it's kind of silly, except then it seemed not so silly anymore.
01:01:33 ◼ ► Yeah, that was, see, here's what I think is actually going on here is there is indeed a defect that happens to some of the iPads and they ship them that way, which they shouldn't.
01:01:43 ◼ ► And they got asked about that and the real answer is, yeah, those are defects. They should be returned and exchanged.
01:01:49 ◼ ► And you see some of the pictures online that are like very bent. And I have a feeling if Apple really did say like some slight variation is normal, I don't think that level of bending from like the Reddit pictures is what they're talking about.
01:02:01 ◼ ► And then, so today, the day we're recording, which admittedly is before many of you were hearing this, Dan Riccio in an email to somebody, like kind of Steve Jobs style, like the way to get news out there is to respond to somebody's email because they know it's going to end up on MacRumors.
01:02:16 ◼ ► So Dan Riccio from Apple said basically that some amount of bentness is normal up to 400 microns, which is 0.4 millimeters. And the pictures that you see on Reddit of the bent iPads, those are bent way more than 0.4 millimeters. Way more.
01:02:35 ◼ ► So that is not what Apple is saying is acceptable. And so, you know, 0.4 millimeters over the entire width of like a 13 inch iPad might be noticeable, maybe. Honestly, probably not.
01:02:50 ◼ ► You know, to most people, I think that probably would not be noticeable. So I think this is mostly a, it's a non-story in the sense that I don't think Apple is doing anything unreasonable here, but their communication about it was really awful.
01:03:04 ◼ ► So one thing to keep in mind with the bend is when you make a device that's entirely flat and has straight edges, any kind of bend is going to be much more apparent than the old design that had curved edges, right?
01:03:20 ◼ ► Because there was less of a hard line, less, everything doesn't have to be perfectly straight. You had a bend in a curved one because the edges were curved in this sloping Apple style, like gradual, whatever the hell it's called, the ellipsoid curve thing or whatever.
01:03:33 ◼ ► You could hide some, you know, you might not notice that the curve starts back a couple of millimeters from where it would have started because there's a slight bend in the thing.
01:03:42 ◼ ► When you make it totally flat, even, you know, even accounting for the camera bump and everything, you can put it on what you think is a totally flat table, which is another potential issue.
01:03:54 ◼ ► The table might not be flat and hold down one side of it and what you're expecting to see is nothing like the other side doesn't move or go up or be elevated at all.
01:04:04 ◼ ► And if you see it move up at all, you're like, oh my God, it's bent. And you would never notice that on a curved one.
01:04:09 ◼ ► And also the thing is so thin that proportion wise, any kind of gap that you might see looks bigger in comparison to the thickness of the device unless it was like, if it wasn't, it was much thicker, that gap would look much smaller in comparison.
01:04:23 ◼ ► So 400 microns, the measurement they give to give you an idea how big that is, it's four sheets of paper.
01:04:32 ◼ ► But when I think about it, four sheets of paper is kind of thick and I wouldn't want that to be able to fit underneath the edge of my new iPad Pro.
01:04:39 ◼ ► This is part of what Apple always does, like pushing the envelope on manufacturing, making these exquisitely finished devices to extremely high tolerances.
01:04:53 ◼ ► And the iPad Pro is really pushing the limits of that because it's the thinnest device they ever made and it's really, really big and wide.
01:05:00 ◼ ► What Casey was talking about before, like not wanting to talk about this, or people who buy one of the new iPad Pros and then like bend it over their knee and be like, look at this, it's a $1,200 thing and you could bend it.
01:05:10 ◼ ► Yep, you can bend it. You're a big, strong boy. You can, like, it's really, really skinny, right?
01:05:17 ◼ ► You can also do the same thing, by the way, to the lid of almost any of your laptops if you really feel like it. It's really skinny. You can bend it over your knee. I promise you. Don't do that because you're just breaking your stuff.
01:05:29 ◼ ► It is a concern, like with the iPhone 6, if during the course of normal use it may get bent. I think the iPad Pros are approaching the point where they may have to take a different design that accounts for the fact that there's going to be some flex.
01:05:45 ◼ ► We've talked about this before when we're talking about new materials in the post-aluminum age, whether it's like going back to plastic or carbon fiber or something that springs back from bending better than aluminum does.
01:05:58 ◼ ► Glass actually does a much better job of springing back aluminum. You can kink glass. It tends not to kink. It tends to bend until it shatters. But you can't really kink glass.
01:06:08 ◼ ► And so this iPad Pro, I don't think you can make it much thinner. I don't think it would bend in normal use, but I think it might bend mishandled inside a backpack. That's what some people have been talking about.
01:06:19 ◼ ► I put it inside my backpack with a bunch of books and I carry my book bag with me to and from school or work or whatever and I take the thing out and I notice it's slightly bent. I can see that happening in the same way that people can bend iPhone 6s in a tight jeans pocket.
01:06:36 ◼ ► So this may be, you know, either it's going to stay this thickness or get thicker or they're going to have to come up with a different design. But as for the specific design, part of pulling this off successfully is figuring out how to manufacture and assemble it so that when they come off the line at the end and ship to the customer that they are actually flat.
01:07:00 ◼ ► Now I can imagine Apple trying to correct for this. Is there like, oh, it's about like the cooling process during manufacturing, you know, materials, you know, expanding and track different rates at different temperatures because it's not all, you know, a giant thick piece of metal.
01:07:14 ◼ ► There's different things inside there. There's glass, there's plastic, there's all sorts of stuff. Right.
01:07:19 ◼ ► And you say, oh, it was coming out slightly bent just to make a minor adjustment to the manufacturing process so that, you know, cooling or whatever, you know, when it's all done contracting and coming to room temperature that it's flat.
01:07:30 ◼ ► The problem is that you don't know what temperature is going to be used at. And, you know, if you account for the conditions at the end of the line of the factory, they may not be the same as the conditions in transit, which may not be the same as the conditions where it's used.
01:07:43 ◼ ► If you, you know, use it on a cool fall day or you live in the tropics and your house is 80 degrees all the time, like basically like this is how, you know, thermostats work and lots of other things where they have put different materials right next to each other, but join them to each other and they expand and contract at different temperatures and they curl.
01:08:02 ◼ ► Like I said, the thermostats, old style thermostats used to work right. It's very difficult to take something that is very thin, very solid and made of heterogeneous materials and make it so that it doesn't deflect slightly in changes in temperature.
01:08:18 ◼ ► So I feel for Apple with their challenge here because the thing is so thin, like maybe they have to do like a floating panel type design to use like a cabinetry parlance where the inside sort of float inside the outside so they don't deflect or whatever.
01:08:33 ◼ ► Like, it's not an easy problem. I think Apple should solve it and I would be upset if I got one that was bent. If I got one that was 400 microns bent, you know, I would live with it.
01:08:41 ◼ ► My current iPod, iPad is probably 400 microns bent. That's another thing that Dan Riccio said. This variance that they allow them off the line with up to 400 microns of bend, that is tighter than the variance of previous iPads.
01:08:55 ◼ ► Previous iPads are allowed to bend even more than that and probably are. Like, don't go look at your iPad. You might not like what you find.
01:09:00 ◼ ► So I don't think this is much of a story, but I think the story here is how, like the manufacturing story, how Apple is making it harder and harder on themselves to make products because our standards for them keep going up because their products are just so much, you know, cleaner and tighter than they were back when.
01:09:22 ◼ ► Remember back when they assembled laptops out of multiple different pieces of aluminum? They looked like just ramshackle, you know, sheds compared to the solid things that we have today.
01:09:32 ◼ ► So our expectations keep going up and they keep pushing the limits. They keep making their things thinner and using designs that will draw your attention to any flaw.
01:09:42 ◼ ► And then, yeah, they kind of flub this in the PR saying, "Oh, some bending is normal." That's not really the answer.
01:09:47 ◼ ► Again, it's probably not Apple's fault because the question they were actually posed and the cases that they were addressing probably were within this.
01:09:55 ◼ ► But this story on the Internet gets put right next to the picture from Reddit of someone's iPad that they bent over their knee that's like a centimeter off the table.
01:10:03 ◼ ► And those two things aren't connected to each other, but they're next to each other on a web page. And so it seems outrageous.
01:10:08 ◼ ► I think their actual reaction should have been something along the lines of like, "If you get one that's bent, bring it back and we'll exchange it."
01:10:17 ◼ ► Like that should be the actual reaction. Like that is a defect, bring it back and we'll exchange it. But that's not what they said.
01:10:24 ◼ ► They instead said, "It's totally fine. It shouldn't affect it if it's bent." And it's like, "No, no."
01:10:29 ◼ ► What they were trying to be is honest in it like, "Oh, there is a tolerance. There is some sort of range where it's okay."
01:10:38 ◼ ► If they just told everyone, "Just bring it in and we'll return it," that would have been fine too because people who have a 200 micron bend are not bringing it back.
01:10:46 ◼ ► They're going to be like, "I guess it's flat." And people who have a huge bend are going to bring it back and either get it exchanged or argue with the person that they didn't actually bend it over their knee.
01:10:54 ◼ ► That's the tricky part with this. Unless you take it out of the box right in front of the people in the Apple store and look at it right at that moment, how do you tell whether this was in someone's book bag and bent because they threw their book bag onto the floor of the bus or something?
01:11:07 ◼ ► Or whether it came out of the box like this. It's just based on the word of the person. So this is kind of a difficult situation.
01:11:14 ◼ ► They just told everybody, "If you've got a bent one, bring it in and we'll exchange it." Maybe they didn't want the extra traffic. Maybe they didn't want to have to deal with determining whether... They have to deal with this anyway.
01:11:24 ◼ ► People are bringing in bent iPads right now to Apple stores and asking for them to be exchanged. Maybe they just didn't want to ramp that up. I don't know.
01:11:31 ◼ ► But I think Marco should now take a look at his iPad Pro. And Casey, you've got one too. See if yours are bent.
01:11:41 ◼ ► I glanced at mine earlier and it seems that I'm putting it on my glass desk and it seems pretty damn flat to me.
01:11:51 ◼ ► I did. Earlier today I did. And it does not seem bent to me. Now, as we have established not only on this episode but the last 300 and whatever episodes, I have far less of a discerning eye than most.
01:12:02 ◼ ► I'm trying to help you. When I say sight down the edge, I mean pick it up and hold it like a shark fin up to your eyeball and look along the top of it.
01:12:17 ◼ ► It's not a video podcast. I don't know what you're doing over there. I know you're putting it on the cable and trying to look for a gap between the table and the iPad, which is not what I'm asking you to do.
01:12:25 ◼ ► Alright, just like that. Where's the stupid button in Skype? Alright, now it's a video podcast.
01:12:43 ◼ ► Are you podcasting with your closet door just like half a jar? How can you handle that?
01:13:06 ◼ ► Yes. Now look along the top. Like the top of it is a little road. Hold it right up to your eyeball.
01:13:37 ◼ ► Oh god. So this is the first time, I think in 300 whatever episodes of ATP that we've had the video on for more than an accident.
01:13:47 ◼ ► Yep. Like imagine there's a car driving along the road that is the edge of your thing and you're looking directly into the tailpipe of that car.
01:13:54 ◼ ► And what you're looking for is to see, like you're using foreshortening to see if it curves to the right or the left.
01:14:03 ◼ ► And by looking down the edge of it, you are taking that long distance and you're compressing it so that any deflection is exaggerated.
01:14:18 ◼ ► Go up to your bookshelves, put your eyeball right to the edge of the shelf and look down your long bookshelf and you'll see how bowed they are from the weight of the books.
01:14:26 ◼ ► Oh god, here we go. Now it's probably that mine is bent. I just can't see because I'm blind and an idiot.
01:14:36 ◼ ► Yeah, like it's as if...are they doing the thing where like they bend it so that it lays flat with the camera bump? Is that a thing?
01:14:48 ◼ ► Alright, so now you've got it all in my head. Now I'm wondering if I'm...oh god, this is such great audio.
01:15:00 ◼ ► If you look at the other short edge...yeah, totally. The short edge is definitely bent.
01:15:14 ◼ ► Yeah, so that like if you lay it flat, the camera bump will be pushed up such that it is actually more flat without a case.
01:15:22 ◼ ► Yeah, so you may be giving it that bend by constantly putting it down on the camera bump and pressing down with your fingers.
01:15:41 ◼ ► And I don't think you could see it if you weren't doing that. Like I was giving you the way to be able to detect this with your eye.
01:15:47 ◼ ► See, here's what's interesting about it is if I put it on a desk and I like poke all four corners...
01:15:57 ◼ ► Well, but I would expect if I put it on the desk with the camera bump on the desk, it should rock. But it kind of doesn't.
01:16:13 ◼ ► Yeah, so when I hang the camera off the desk but have the rest of the iPad on the desk, you can see like if you push all four corners...
01:16:22 ◼ ► You know, the three remaining corners that are on the desk, two of them do rock very slightly, but you don't even...
01:16:27 ◼ ► It isn't even enough to feel the rocking. You only see the rocking. Does that make sense?
01:16:33 ◼ ► So it's pretty subtle. And it might even be intended to make the camera bump more level. I don't know.
01:16:40 ◼ ► Well, that could be what they're talking about with the cooling thing during the manufacturing process.
01:16:46 ◼ ► I bet if you change the temperature of the thing to be either hotter or cooler, it might undo that bend or reverse it.
01:16:57 ◼ ► I have this screen protector that's like nicely applied on here, but not that much dust on there. I wouldn't want to redo that.
01:17:03 ◼ ► Yeah, that's the other thing. You get the perfect screen protector applied and then your device dies.
01:17:25 ◼ ► We are sponsored this week by Away. For $20 off a suitcase, visit awaytravel.com/atp and use promo code ATP during checkout.
01:17:41 ◼ ► Away makes these amazing, high-quality suitcases from lightweight, durable German polycarbonate or now aluminum alloy.
01:17:51 ◼ ► So, for instance, the Away carry-on sizes, which now there's two sizes of Away carry-ons, have an optional ejectable battery.
01:18:03 ◼ ► And if for any reason you need to remove the battery, like if it's getting gate checked, it's super easy to just pop it out with one click.
01:18:09 ◼ ► All the other times, you can sit there waiting at the gate. You can charge your phone, your tablet, your e-reader, whatever it is.
01:18:26 ◼ ► There's a removable, washable laundry bag. And you get so much space in these. You'll be amazed how much you can carry.
01:18:32 ◼ ► On the outside, you have a TSA-approved combination lock and four spinner wheels that spin 360 degrees for a nice, smooth ride and easy maneuvering.
01:18:40 ◼ ► All Away suitcases are guaranteed for life, and they give you a 100-day risk-free trial.
01:18:46 ◼ ► So this means you can actually buy them and travel with them. Of course, because how else would you test them out, right?
01:18:51 ◼ ► Travel with it. And see how it works for you. See if it fits your life. See if it works for your style, your packing, whatever it is.
01:19:01 ◼ ► So check it out today at awaytravel.com/atp. And for $20 off your suitcase, use promo code ATP during checkout.
01:19:10 ◼ ► That's awaytravel.com/atp and code ATP for $20 off a suitcase. Because this season, everyone wants to get Away.
01:19:18 ◼ ► All right, let's do some Ask ATP. Mitch Cohen asks us, "Does Overcast do anything special to inhibit cookies and podcast downloads?"
01:19:29 ◼ ► I'm sure the large podcast distributors take advantage of this, and I wonder if shared hosts like Libsyn do as well.
01:19:34 ◼ ► This isn't the worst thing in the world, no worse than web cookies. I only use "private browsing" now, which at least helps with tracking.
01:19:42 ◼ ► So I wonder if removing cookies from the HTTP cookie store might be an option to improve privacy.
01:19:47 ◼ ► Secondly, I know it would be a dumb cat and mouse game, but do you think NPR or other hosts would selectively block players via the user agent string for not playing ball with their rad-like proposals?
01:19:58 ◼ ► All right, so this is a two-part question. First part, about cookies. So this is an easy one. Overcast does not persist cookies for any kind of downloads or anything.
01:20:09 ◼ ► For anything, really. My own API, the Overcast API, does not use cookies in that way, so it doesn't need them, so it doesn't do that for that.
01:20:18 ◼ ► And then for any kind of downloads from third-party servers, like for either artwork, which is sometimes third-party, or for the actual podcast downloads themselves, which is what this question was talking about,
01:20:30 ◼ ► there's an option when you create an NSURL session in the API to use ephemeral sessions, which means basically they're not saved, that no persistent data is written to disk.
01:20:41 ◼ ► This does not mean that I don't use cookies during the session, because a lot of times, when you're following arbitrary download URLs out there on the internet,
01:20:52 ◼ ► the downloading process itself should use cookies within the request that it's making. So if you request something and it's going through four different redirects,
01:21:02 ◼ ► one or two of those redirects might set a cookie, and if you don't then send that when you follow the redirect to get the next step of the redirect chain, sometimes it won't work right.
01:21:10 ◼ ► So I do use cookies within the request, but then they're not saved, and therefore they're not used in future requests to the same server.
01:21:18 ◼ ► So it's something that can only be used within the session of downloading that one file, but there's no real ability to use cross-request or long-term tracking using cookies.
01:21:29 ◼ ► Now, there's other ways they can do it, like with IP address tracking, that I can't do anything about, but cookies, I treat cookies basically in almost the most hands-off way possible.
01:21:38 ◼ ► That's something that people who aren't developers might not realize. I think of cookies as this magical thing on the web that tracks you, but when someone makes a web request for something,
01:21:48 ◼ ► like from some code you write, like when a developer writes code to make a web request, you or the framework you're using actually has to do some work to make cookies work "the normal way."
01:21:58 ◼ ► Obviously, your web browser is doing this. It's the point of a web browser. But if you're in a program, like a podcast client, and you have to make an HTTP request to download a file,
01:22:06 ◼ ► cookies are just an HTTP header that the server sends back. You don't have to do anything with that HTTP header. You can ignore it. You can never read it. You can do nothing with it and drop it on the floor.
01:22:15 ◼ ► Now, as Mark pointed out, to successfully follow a bunch of redirects and get to your download, you might have to tell your framework that you're using to make HTTP requests, by the way, take that cookie and save it or whatever.
01:22:25 ◼ ► But again, if you don't make an explicit effort to say, "Oh, and I'll also save that to disk, so that like I'm pretending I'm a little mini web browser inside there so that the next time you quit Overcast and relaunch, I still have that cookie,"
01:22:35 ◼ ► if you don't explicitly do that, or if some library you're using doesn't explicitly do that, it doesn't happen. So in other words, the websites don't have the power to make these cookies go into your program that you're writing.
01:22:52 ◼ ► Now, there is a small complication to that, and that is that depending on the framework that you're using, depending on the OS and the language and the URL accessing framework you're using, sometimes cookie storage is automatic.
01:23:08 ◼ ► Yeah, it might be the default. And that's considered a good feature because it's a pain to do it manually yourself. You're like, "Oh, it's great. The framework does it for me. I don't have to think about it." But it is still a choice.
01:23:17 ◼ ► Exactly. And then regarding the second part of the question about whether my intended non-compliance with NPR's radspec might eventually result in them user agent banning me, it's possible. Somebody might at some point do that.
01:23:32 ◼ ► I don't think I care because I don't think, first of all, I don't think there's a risk that that would be very likely. It's not good business to start blocking clients for relatively minor reasons. That's not a major reason to block somebody.
01:23:48 ◼ ► And so I really don't think that my non-compliance with their spec would even make their radar. They wouldn't even care that this podcast app with 2 or 3% market share doesn't do our spec. Because you know what? Even if the spec goes really, really well, there's still going to be 75% of the market that doesn't implement it.
01:24:06 ◼ ► So I wouldn't stick out enough for them to care. And if they did, and suppose they are real mad at me, and they wanted to block overcast, and they sent me legal things saying, "Don't change your user agent to evade this or we'll sue you. We're blocking you. That's it."
01:24:23 ◼ ► Oh well. What can I do? The fact is, of the top 50 podcasts in overcast, most of them aren't NPR. Most of them aren't any one provider. The wonderful thing about podcasts is that the audience listens to all sorts of diverse things.
01:24:40 ◼ ► And it would really hurt to lose popular shows somehow. For a popular show to not be playable in my app while remaining playable in lots of other podcast apps would be a pretty big jerk move. And it would be really inconvenient for me. And I would definitely lose some customers to it.
01:24:59 ◼ ► But it wouldn't be the end of the world. I wouldn't lose my entire customer base. I probably wouldn't even lose 20%. To give you some idea of what this might mean, it wouldn't mean that much. It would hurt, but it wouldn't be the end of my business.
01:25:13 ◼ ► And I would just keep going because it probably wouldn't affect any shows I listen to. And the reality is, if I'm big enough for them to care that I'm not implementing their spec, I'm probably also big enough for them to not want to lose my audience completely.
01:25:34 ◼ ► And also getting to the cat and mouse thing, you just change your user agent string to make it look like you're into the Explorer or some other podcast. It's just a string that you send with the request. It's cat and mouse because they could keep, "Aha, I'm banning all things that look like your user agent string." But if you make your user agent string indistinguishable from some other podcast, it's stupid. It's not a feasible thing to do.
01:25:54 ◼ ► Moving on, here's another example of, "Oh, your Twitter username is so clever. Well, Grinchreal Selkin, how do you feel now?" He or she would like to know…
01:26:09 ◼ ► You're going to feel bad when poor Grinchreal is going to be like, "That's the name my parents gave me, you insensitive clod."
01:26:19 ◼ ► This individual would like to know, "Hey, what if Intel got their act together, fixed their process issues, and launched this new architecture?" We'll put a link in the show notes.
01:26:27 ◼ ► "All in 2019 or 2020. Would Apple think twice about switching to ARM or is it too little too late?"
01:26:32 ◼ ► This is kind of interesting because this implies that the ship has sailed and Apple's definitely switching to ARM. That's kind of how I feel, so I don't blame the Grinch for saying that.
01:26:42 ◼ ► But that isn't actually true as far as we know. I concur, it seems like there's enough smoke for there to be fire here, but we don't really know.
01:26:52 ◼ ► But that being said, given my gut feeling about the situation, I do think it's too little too late.
01:26:58 ◼ ► Apple loves control, Apple wants control, and they do not want to rely on Intel anymore. Marco, what do you think about this?
01:27:05 ◼ ► I actually have not been following whatever this new architecture is, but I do ultimately think the ship has already sailed.
01:27:12 ◼ ► And while Apple could possibly, I don't know, unsale it, sync it, I don't know, whatever it is, whatever they would do in this crazy metaphor, I don't think they would.
01:27:24 ◼ ► Because ultimately, if Apple is deciding to move to ARM, they would be doing it for a few reasons.
01:27:35 ◼ ► But there's lots of other reasons too, things like control and having things be in-house.
01:27:41 ◼ ► First of all, they can almost certainly make more money per unit if they control that, because Intel's CPUs are very expensive.
01:27:55 ◼ ► The lack of the control in those technologies over the last five years or so has really gotten them burned badly from Intel.
01:28:03 ◼ ► And so if Intel suddenly stops burning them next year somehow, which honestly I think given Intel's performance recently is very unlikely.
01:28:10 ◼ ► But if Intel suddenly gets better again, and it's like, "Okay, now we're back on track, we're starting to develop things regularly again and not miss our targets and keep things moving forward and be more competitive on performance per watt,"
01:28:21 ◼ ► that still doesn't change the fact that Apple was just really badly burned for like five years.
01:28:27 ◼ ► And so I think Apple would take this as a chance if, "You know what? We now know that we don't want to depend on Intel at all, even if they 'get better.' We don't want them."
01:28:38 ◼ ► So I think that chip has already sailed, I think they're not going to sink it, and it's probably, for lots of reasons, beyond just this one architecture has taken them too long to reach.
01:28:49 ◼ ► Well, I think it kind of depends on the specifics of the thing, of the new architecture. We'll link to this Anatec article that outlines it.
01:28:58 ◼ ► But this is like a public presentation that anybody could see, and it's necessarily vague about future plans because they don't want to make promises they can't keep.
01:29:08 ◼ ► So they talk about the architecture and the plans and the different models and pretty much what each one is doing, but they don't get down to the point of saying exactly,
01:29:15 ◼ ► "Here's what the benchmarks are going to look like, here's how much power it's going to take, here's the actual nitty-gritty details of, 'Okay, so Apple, so you're making a laptop next year, and you need a part that has this performance and uses this much power,
01:29:28 ◼ ► and power is so much more complicated now because it's like this crazy curve depending on what you're doing.'"
01:29:34 ◼ ► There's nothing that concrete in this presentation. It's purely like, "This many execution units, our focus will be on single-threaded performance, new instructions for AI, blah blah blah," but very vague and high level.
01:29:47 ◼ ► But what Apple, the pitch to Apple would be, and was in the past, they get the much more concrete numbers.
01:29:55 ◼ ► Like when they switched to Intel, Intel basically gave them a preview of the core architecture. They said, "The chips we've got now might not look so hot, but here's what we've got in the pipeline.
01:30:03 ◼ ► We're going to make these core chips, and here's our pipeline for core chips, and here's what we expect. This year we're going to have this price, this size, this power, this performance, the next year this, the next year that."
01:30:12 ◼ ► And that roadmap, which I presume was way more detailed than what anyone got in the public, looked really, really good.
01:30:20 ◼ ► And Intel actually delivered on that roadmap, which helps. And so Apple was super happy, and we were all happy. We got Intel Max, they were super fast, the core architecture was a great architecture for a long time.
01:30:30 ◼ ► It was so much better than anything else available, especially in terms of power and performance, so it was good for Apple's laptops, it was good for their desktops, everybody was happy.
01:30:38 ◼ ► This presentation of "Here's what we're going to give you," doesn't have enough detail to say whether it's that great.
01:30:44 ◼ ► And looking at the high level thing, I'm not impressed by this high level plan. I don't look at this and say, "Wow, Intel's going to have some great chips."
01:30:51 ◼ ► I look at this and say, "Best case scenario, if Intel delivers everything they say here, they'll keep their head above water."
01:30:58 ◼ ► And that's not what Apple wants. They want things to kick butt. They don't want just sort of like, "We'll kind of get out of the current slump and get our momentum back and they'll be pretty good."
01:31:07 ◼ ► But if Apple wanted to win Apple back or keep Apple on board, they would have to show them very concrete numbers that are way more impressive than these vague numbers.
01:31:20 ◼ ► And Apple, to Marco's point, would have to believe them and say, you know, and Apple basically would have to compare it to, "Here's our internal projections of who we made our own chips that we could get.
01:31:32 ◼ ► Intel, you better tell us that you've got something that's way better that we can't possibly reach and also is worth the extra money you're going to charge us."
01:31:41 ◼ ► It would be a very difficult sell. So I don't think anything in this article convinces me that Apple should stick with Intel instead of its theoretical ARM chips, assuming their ARM chips are way better than this.
01:31:53 ◼ ► And unless Intel is showing Apple super-secret stuff that's way better, I don't think Apple will be swayed.
01:32:00 ◼ ► Yep. Like I said, what was it? The last thing. Apple might as well be screaming that they're going to make ARM Macs at this point. So I will be extremely shocked if 2021 comes and there are no ARM Macs.
01:32:14 ◼ ► All right. And these two, we're going to do a fourth Ask ATP, a third and fourth Ask ATP, because I think these will be quick. Marcus N. writes, "Do I over or underestimate the amount of time it takes to cut an episode of ATP?
01:32:29 ◼ ► Do you care to elaborate on the process? Also, as someone who does not listen to the live shows, how much do they differ from the downloadable episodes?"
01:32:38 ◼ ► Yeah, sure. So total editing time for an episode of ATP is generally like most of a morning.
01:32:44 ◼ ► So that's maybe a couple hours, two to three hours, depending on what else is going on and how complicated of an edit is.
01:32:52 ◼ ► John Talks is a big differentiator, because John Talks in long, uninterrupted blocks that usually don't require much editing at all. So I can skim over a lot of them.
01:33:02 ◼ ► Whether or not they require editing, they don't get editing. I think sometimes they could use some editing, but apparently they don't get editing.
01:33:08 ◼ ► And then, of course, editor's privilege is I edit myself the most. And so if it's a me-heavy episode, that takes a little bit longer. But for the most part, it ends up taking maybe about one and a half times to two times the duration of the show, in additional editing time.
01:33:26 ◼ ► And as for how it differs from the released show, it's not actually a major difference. It's mainly deciding when to start and stop it.
01:33:34 ◼ ► What is the pre-show? What is the post-show? Where does it stop? Where does it start? Putting in the ads.
01:33:40 ◼ ► And we all sound better, because all the audio files have been brought in and normalized from all of our local recordings and everything. So it's a much better sound quality.
01:33:46 ◼ ► And then content-wise, I will remove any talkovers, if at all possible. Any time that one of us talks at the same time as the other one, because we're jumping in at the same time or whatever, I'll try to remove those.
01:33:58 ◼ ► If a joke takes too long to land before Casey laughs, I'll move the laugh forward, so that way I sound funnier. And then, otherwise, you know, it's minor stuff. Like, if I repeat myself a lot in a paragraph that I'm speaking, which I do all the time, I will try to reduce that.
01:34:16 ◼ ► If I say something, if any of us say something that's just a total blunder, I will remove it or I will fix it. If a joke doesn't land, nobody laughs, I'll take it out if I can. Stuff like that, just to make us sound better.
01:34:28 ◼ ► If we ramble on about something for a long time that's really just asking, like, "Chat room, what was this thing?" And we're just talking about something that we don't even know the answer to, or what we're saying is totally wrong, I will try to cut it out completely.
01:34:40 ◼ ► That way it saves us a week of email saying, "You know, you were wrong about that," and then, you know, here's the real thing. Try to avoid, you know, potential problems before they happen.
01:34:49 ◼ ► So, for the most part, I would say what you're getting with the polished, finished, edited product compared to the livestream is you are getting, like, 95% of the value, but in a way better, nicer package with, like, 50% less of the garbage and flubs.
01:35:07 ◼ ► So, I don't think you're missing out much by not having the live shows, if you don't listen live.
01:35:13 ◼ ► Occasionally, you'll take an entire segment and move it from the beginning of the show to the end of the show or whatever, which usually, you know, you don't notice because you don't know when it was actually said, but that's rare. Usually what you hear is sort of the order that we said things in.
01:35:25 ◼ ► Oh, yeah, almost always because the problem is we refer back to things all the time, so, like, if I try to move, like, a later segment into an early part of the show or vice versa, you will hear in the conversation, oftentimes, like, you would hear us say some reference to that thing that wouldn't make sense because you hadn't heard it yet or whatever else.
01:35:44 ◼ ► So, I almost never move things that much. I will cut entire topics sometimes if they just, if it doesn't go anywhere, like, if it's, like, follow up on something really obscure that I don't think anybody would care about that we didn't have much to say about. Sometimes I'll cut that.
01:35:58 ◼ ► - And then we'll get a week of feedback about why we didn't say something about it, and then I'll yell at you in the next show.
01:36:02 ◼ ► - Yeah, pretty much. I'm sure you've all heard this before. So, yeah, that's about it. And I try to limit the length of the show to right around two hours.
01:36:13 ◼ ► I don't like going past two hours, or if I do, maybe only by, like, you know, 10 minutes. But I try not to go past that. And so, occasionally, some trimming there is necessary. But for the most part, you're not really missing that much if you don't hear the live shows.
01:36:26 ◼ ► - And we won't go back in and rerecord stuff. So, for example, in this episode, Casey unfortunately said Lady in a Tube again, and we're not going to give him a chance to correct that.
01:36:35 ◼ ► And he's just going to have to live with the fact that he said it reflexively without thinking about it, and it just went right by.
01:36:41 ◼ ► - See, I was going to edit that out, but I don't think there's a good way for me to do it.
01:36:44 ◼ ► - No, there isn't. That's what I said. You have to cut that whole segment, so we'll just know that, you know, it's difficult to speak extemporaneously on a weekly basis and not flub something.
01:36:54 ◼ ► We acknowledge that we do it. We're sorry, but sometimes it just stays in the edit, because we're not going to go back and have Casey reenact that section.
01:37:04 ◼ ► - It wouldn't sound right. It's not like we don't want to do it because we're lazy. Like, it wouldn't work. We're not actors.
01:37:14 ◼ ► - You can go with cylinders. You could go with person in a tube. We're trying to avoid, yeah, sort of the fridging of cereal.
01:37:23 ◼ ► - As soon as I said that, I was like, "Wait, this isn't right," but I just had to roll on at that point. Whoops. Sorry, everybody.
01:37:31 ◼ ► All right, and then finally, Eric asks, "Hey, for Marco and Casey," and actually, I'll throw Jon in on this too in a different way,
01:37:37 ◼ ► "Do you use Xcode in full-screen mode or with multiple windows?" For me, I have a kind of similar tile approach,
01:37:44 ◼ ► as I've talked about in the past with when I podcast. Generally speaking, well, at least when I'm writing an iOS app,
01:37:50 ◼ ► it's been different now than working on an iPad thing, but when I had, or I'm sorry, when I'm working on an iPhone app,
01:37:56 ◼ ► when I was working for Work Work, it was almost exclusively iPhone work, and so I would put the simulator in the upper right-hand corner.
01:38:03 ◼ ► I would put the terminal in the bottom right-hand corner, and I would have Xcode on the left side of the screen extended
01:38:10 ◼ ► all the way until it kisses the side of the simulator window, so it's 80% or 90% of my screen, but I can always at any moment
01:38:18 ◼ ► see the terminal, see Xcode, and see the simulator all at once, and that's how I roll. The only time I have a really useful screen,
01:38:25 ◼ ► which is very rarely, is if I'm on the adorable and I have a physical device connected for debugging purposes.
01:38:34 ◼ ► - Similar. Not only do I not use Xcode in full screen, I almost never use anything in full screen ever,
01:38:41 ◼ ► unless I'm watching a movie on a computer for some reason. That's it, because the problem is, full screen to me,
01:38:47 ◼ ► it is so different from the way I both work and the way I want things to be in a windowed Mac multitasking environment.
01:38:57 ◼ ► It's one of those things that I feel like full screen is up there with Launchpad as things that were added to the Mac
01:39:04 ◼ ► in the time where they wanted to make it more iOS-like, but that I don't think should have been added,
01:39:09 ◼ ► and I think ultimately are just weird and don't fit in. Full screen to me feels like I'm on a ride.
01:39:15 ◼ ► My computer's taking me for a ride. Look, this giant animation, whoop, moved the whole 27-inch screen over
01:39:20 ◼ ► while I wait for this animation to happen in and out. I feel like I'm being taken for a ride,
01:39:26 ◼ ► and that's not a good feeling. I feel like I am not controlling things the way I want to control them,
01:39:31 ◼ ► and the computer is trying to put on a show for me that I really don't want. Furthermore, the practical reality of it,
01:39:41 ◼ ► switching in and out of full screen is the core that I have a problem with. That's why you got all these big animations
01:39:48 ◼ ► that you got to wait for and everything. But then also, full screen work mode, I feel like that must work for some people,
01:39:57 ◼ ► because it sure looks good in people's, in product shots when you see a new version of an image editor or something comes out,
01:40:05 ◼ ► and in their product shots, it's all running in full screen. Look at this wonderful, pro, awesome, black background environment
01:40:11 ◼ ► where this creative person is getting their work done. It's the same kind of people who can do all their art on their iPad.
01:40:17 ◼ ► It's like, that's great. I am not one of these people. For me, full screen mode just does not work at all,
01:40:23 ◼ ► and I've never seen anybody for whom it does work. If you're out there, I believe you exist.
01:40:27 ◼ ► Just like John's HDMI CEC unicorns, I believe you exist out there, but I've never met one of you,
01:40:34 ◼ ► and I'm certainly not one of you. Especially when doing development, I am going between multiple apps.
01:40:39 ◼ ► I'm going between Xcode and the simulator at the very least, and I'm probably also going between Xcode,
01:40:43 ◼ ► simulator and Dash, maybe between Xcode and its own documentation window and its own organizer and devices windows.
01:40:59 ◼ ► You might be also using Terminal and also using a web browser to look stuff up or to read an article about what you're doing,
01:41:10 ◼ ► This is one of the reasons why I think Xcode on iPad is further away than most people think it is,
01:41:17 ◼ ► because development is inherently a very, very multi-app, multi-window, multitasking-heavy job to do.
01:41:36 ◼ ► It just doesn't really make sense to have this full screen takeover mode where going in and out of it requires all these big, dumb animations.
01:41:44 ◼ ► You have met somebody who works in full screen mode because I am getting real-time follow-up from Jelly,
01:41:54 ◼ ► I don't know how you do Xcode full screen in anything bigger than my MacBook, but apparently that's how Jelly rolls.
01:41:59 ◼ ► Additionally, I should add that I am a very heavy and very enthusiastic user of spaces.
01:42:06 ◼ ► Oftentimes my web browser is not on the same space as Xcode and the simulator and all that,
01:42:11 ◼ ► and it sounds like, Marco, you are not a fan. If I recall correctly, John is very much not a fan.
01:42:19 ◼ ► John, you've obliquely talked about in the past that you do a fair bit of JavaScript development these days, particularly Node.
01:42:26 ◼ ► Are you using Visual Studio Code? What are you using for that? Or BBEdit or what have you?
01:42:31 ◼ ► And are you doing that full screen, or are you doing the Windows of Syracuse accounting?
01:42:35 ◼ ► I mean, you know the answer to this. I'm obviously using multiple Windows for everything.
01:42:43 ◼ ► I can tell you, people love full screen. Now, maybe most of them are XWindows users who are just trained on everything being full screen.
01:42:51 ◼ ► If you're not, as you two both know, if you are not accustomed to dealing with a whole bunch of windows
01:42:58 ◼ ► and embracing the idea that you get to decide where they belong and having faith that the applications that you use
01:43:03 ◼ ► will honor and remember your manipulations of those windows, and it won't just be throwaway work, right?
01:43:13 ◼ ► But if you don't, it feels insecure. It feels like just a bunch of loose garbage on your screen.
01:43:18 ◼ ► You're just like, "I just need to get this full screen so I'll feel like I'm anchored to the walls and I'm safe again."
01:43:23 ◼ ► I see people use everything in full screen. Their web browsers are full screen. Their IDEs are full screen.
01:43:35 ◼ ► No, I'm talking about actual full screen where you can't get the window widgets unless you jam the cursor against the top
01:43:40 ◼ ► and wait for the little things to slide down. Like, full screen, full screen on the Mac.
01:43:44 ◼ ► And I don't know how they do it. It's not how I work either, but I do know that people love it.
01:43:48 ◼ ► And again, very often, a web browser, full screen on a 15-inch MacBook Pro is an absurd waste of space.
01:43:57 ◼ ► Because it's not comfortable to read lines that long, and if the lines aren't that long, then you're just wasting space on either side or on one side or the other.
01:44:05 ◼ ► So no, I don't do anything full screen. I think the people who use their IDEs full screen, like you mentioned in Visual Studio Code,
01:44:10 ◼ ► a lot of the IDEs for web development let you do, like, you'll have a file browser in your left-hand pane,
01:44:23 ◼ ► And they like the fact that they three-finger swipe to the next space, which is really the next full screen app, and that's their browser.
01:44:31 ◼ ► Or they use multiple monitors, but if they use multiple monitors, they say, "Ah, multiple monitors.
01:44:35 ◼ ► Finally, I can see more than one thing at once." Because they don't understand that you can have more than one thing on a single monitor if you put them in Windows.
01:44:39 ◼ ► Anyway, everyone has their own habits. I think full screen is incredibly popular on the Mac.
01:44:45 ◼ ► If I took a survey of all the developers who use Macs at work, I think the majority of them would use at least one app full screen all the time.
01:44:58 ◼ ► Thanks to our sponsors this week, Away, MacWeldon, and Backblaze. And we will see you next week.
01:45:39 ◼ ► E-Y-L-I-S-S, so that's Casey Liss, M-A-R-C-O-A-R-M-N-T, Marco Armin, S-I-R-A-C, U-S-A, Syracuse.
01:45:53 ◼ ► It's accidental. (It's accidental) They didn't mean to. Accidental. (Accidental) Tech podcast. So long.
01:46:10 ◼ ► Some post-show backup follow-up. I don't know how you call this. Yeah, ain't no follow-back though.
01:46:15 ◼ ► When last we left, my absurd backup situation. I don't know if you remember the situation, but I will describe what it was.
01:46:28 ◼ ► So Crash Plan was getting rid of their home plan, and they had the small business plan, and they had a deal
01:46:34 ◼ ► where you could switch from home to small business for a discount rate, and the discount rate would expire.
01:46:38 ◼ ► That discount rate has recently expired, and so I'm up to paying much more than I would for something like backblaze for my Crash Plan backup.
01:46:46 ◼ ► But Crash Plan will also back up network volumes, and backblaze will not, so I was in this quandary.
01:46:51 ◼ ► And in preparation for this quandary, I entered this absurd situation on my wife's computer, which hosts our photo library of...
01:46:58 ◼ ► I already mentioned I've got the photos on an external SSD because they outgrew the main one. That's fine.
01:47:04 ◼ ► Backblaze will do directly attached disks, but I also have my Synology mounted on there, and backblaze won't back that up.
01:47:10 ◼ ► So on her computer, I was running Crash Plan and having it back up both of the attached drives and the mounted Synology.
01:47:20 ◼ ► Also running backblaze, which was backing up both of the mounted drives but not the mounted Synology.
01:47:28 ◼ ► Also running Time Machine, which was backing up to the Synology and also to a directly attached drive.
01:47:33 ◼ ► So she had three... Oh, and sorry, finally, Google Backup and Sync backing up her photo library to Google Photos.
01:47:42 ◼ ► So that's one computer running four programs for the purposes of backup, and one of them just had a price increase, and now it's like three times as expensive as it used to be.
01:47:54 ◼ ► And one of those backup programs, namely Crash Plan, occasionally pisses me off by going wonky.
01:48:01 ◼ ► It's still a Java program, and its main challenge seems to be that it doesn't expect to back up quite as many files as I have.
01:48:16 ◼ ► Yeah. The Synology is easy. The Synology, because they're big video files, right? You know, like TV shows and movies.
01:48:22 ◼ ► That's actually pretty easy, because the number of files is small, and each file is very, very large.
01:48:26 ◼ ► That's ideal for a backup program. You would think, "Oh, a backup program. It's going to really choke on those big files."
01:48:30 ◼ ► Backup programs love big files. What backup programs hate? Hundreds and hundreds and thousands and thousands of small files and directories.
01:48:38 ◼ ► So I have literally multiple millions of files that are being backed up, and Crash Plan chokes on them.
01:48:45 ◼ ► And if you go to their help document, they'll be like, "You need to give Crash Plan more memory so it doesn't choke on these files."
01:48:51 ◼ ► With the wonders of Java, if you've ever done anything with Java, and you use the Java command to run like a jar file or whatever, you know that you can specify on the command line how much heap space it's supposed to use.
01:48:59 ◼ ► I feel like I shouldn't have to specify this, because I feel like I'm back in classic Mac OS telling my applications how much RAM they have.
01:49:04 ◼ ► Just allocate the memory you need, I feel like saying. But no, that's not the way Java works for reasons I don't understand.
01:49:14 ◼ ► So Crash Plan suggests if Crash Plan is freaking out and isn't able to launch and never seems to back up anything, get it to launch somehow.
01:49:22 ◼ ► I'm not sure how they expect you to do this. And then go into an obscure preference thing and increase its memory allocation.
01:49:29 ◼ ► And I've been doing that for a couple of years, because eventually it just stops backing up because it can't handle it, and then I somehow get into the program and get it to launch and increase its memory.
01:49:37 ◼ ► So right now I'm paying $15 a month and I'm giving Crash Plan. Guess how much memory I'm giving Crash Plan?
01:49:46 ◼ ► Jet room on a going over under? What am I giving my backup program to backup my files and my amount of technology?
01:50:13 ◼ ► Even with 5 gigs, occasionally I will come and find my computer freaking out and the process hung and be unable to launch the thing.
01:50:31 ◼ ► I don't want to, I know it's not using all 5 gigs all the time, but I don't want to even say, "Reserve or carve out 5 gigs of RAM for my backup program."
01:50:43 ◼ ► So I uninstalled Crash Plan. I didn't cancel my account yet because I have to wait until I replace it with a different backup thing.
01:50:50 ◼ ► But I uninstalled Crash Plan and my new plan is I'm going to backup my Synology directly from it to B2, Backblaze B2, for about $12 a month.
01:51:01 ◼ ► So it's not cheap, but I feel like I'll get the computer, I'll get my wife's computer out of the business of dealing with the backing up of Synology and let the Synology back itself up.
01:51:13 ◼ ► I didn't have to allocate 5 gigs of RAM. I don't even know if my Synology has 5 gigs of RAM total.
01:51:18 ◼ ► I just installed the Cloud Sync thing, I pointed it to the B2 backup, and it's just shoving files. It's doing it right now.
01:51:27 ◼ ► And Backblaze has a nice control panel and you can set limits on how much you want it to spend and it has a nice price calculator.
01:51:40 ◼ ► I don't feel great about it, but I feel good about getting Crash Plan with the computer.
01:51:45 ◼ ► So suddenly I have 5 gigs of new RAM and I have one fewer process slaughtering my disk.
01:51:58 ◼ ► It never does anything to my computer untoward. It doesn't cause it to hang. I can schedule it to run at night.
01:52:03 ◼ ► It is always backed up. I look at it and I'm like, "Are you sure you're backed up? I never see you running. How can you be backed up?"
01:52:08 ◼ ► It's like, "Yep, totally backed up. Last backup was like 2 hours ago. I got all your files."
01:52:15 ◼ ► That's how it should work. It should just work. I don't have to think about it. It just runs. Much nicer.
01:52:23 ◼ ► I don't particularly trust that it's getting all of them, but that's my 5th level backup of stuff.
01:52:29 ◼ ► Eventually I will cancel CrashPlan. The good thing about CrashPlan is it will retain old versions forever.
01:52:34 ◼ ► It had a cascade where it would be like, "We'll keep every hour for the first day and every day for the first month and every week for the first two months and then every month."