ATP

00:00:00 ◼   ► I was going to buy a streaming box for a fleeting moment

00:00:02 ◼   ► when I thought--

00:00:03 ◼   ► - When you were becoming a vlogger, right?

00:00:04 ◼   ► - Well, no, I'm still becoming a vlogger

00:00:07 ◼   ► just exceptionally slowly, but I thought about

00:00:10 ◼   ► buying a streaming box when I thought I was going

00:00:11 ◼   ► to get into like Switch streaming with iMic

00:00:15 ◼   ► and then never ended up doing so.

00:00:17 ◼   ► - I wonder which of us makes our second YouTube video first,

00:00:21 ◼   ► you or me?

00:00:21 ◼   ► - It's a race to the bottom, I assure you.

00:00:24 ◼   ► (laughing)

00:00:24 ◼   ► (electronic music)

00:00:26 ◼   ► All right, let's start with some follow-up.

00:00:27 ◼   ► Guess what, we don't really have any.

00:00:29 ◼   ► I think that's because in part we just released, or Marco just released last week's episode

00:00:37 ◼   ► earlier today because we are recording the episode from the future in the past or something

00:00:44 ◼   ► because it is currently the 20th of July.

00:00:47 ◼   ► We are recording the episode that will be released on or around the 26th of July because

00:00:52 ◼   ► our vacations are in serial.

00:00:54 ◼   ► And so we didn't really have time to amass follow-up.

00:00:57 ◼   ► than to say we are going to be doing a Q&A episode, I believe—what did you say, Jon?

00:01:03 ◼   ► Master of Ceremonies, Syracuse? You said the fourth—

00:01:06 ◼   ► Jon Streeter I think it's recording on August 4th, and

00:01:08 ◼   ► we're doing a double recording with back-to-back episodes, which is why we're doing the Q&A,

00:01:12 ◼   ► because we're just going to answer questions until we all die, and then Marc will make

00:01:15 ◼   ► two shows out of it.

00:01:16 ◼   ► [Laughter]

00:01:17 ◼   ► Jon Streeter Sounds wonderful. So anyway, so yes, please

00:01:21 ◼   ► pile up questions. At this time, the only way to get a question put in to be a contender

00:01:28 ◼   ► for the show is to tweet with the hashtag #AskATP. Please do not email us because it

00:01:35 ◼   ► will get lost to the ether. And if we were cooler like Hello Internet, we would have

00:01:40 ◼   ► asked for postcards, but we are not that cool nor that organized. So tweet with—

00:01:44 ◼   ► That would be a terrible mistake if we learned nothing from Hello Internet. Who wants to

00:01:48 ◼   ► to go to the post office or PO box

00:01:50 ◼   ► and bring home a bunch of pieces of paper

00:01:52 ◼   ► and then transcribe the things on them, terrible mistake.

00:01:55 ◼   ► - Yeah, so please tweet #askATP

00:01:59 ◼   ► and we will hopefully select your question.

00:02:02 ◼   ► Anyway, that's it for follow-up, right?

00:02:05 ◼   ► That might be record time.

00:02:07 ◼   ► We should have somebody go through and chronicle

00:02:10 ◼   ► how long we spend on follow-up.

00:02:11 ◼   ► Like, what does the trend line look like?

00:02:12 ◼   ► - I'm sure there's like one angry fan

00:02:14 ◼   ► who has already done that.

00:02:15 ◼   ► - Is that fan you?

00:02:17 ◼   ► Because no one else cares.

00:02:20 ◼   ► Sometimes there's none at all.

00:02:22 ◼   ► Sometimes there's just one thing.

00:02:24 ◼   ► Sometimes there's a lot.

00:02:25 ◼   ► If we record two days before we do a second recording, usually not that much.

00:02:29 ◼   ► Funny how that works out.

00:02:30 ◼   ► We should go on Adobe schedule vacations more often.

00:02:34 ◼   ► Oh man, all right.

00:02:36 ◼   ► So let's start with the topics, and first thing is the Apple A10X is 10 nanometer.

00:02:43 ◼   ► That's cool.

00:02:44 ◼   ► So tell us about this, John.

00:02:45 ◼   ► Remember we were talking about Intel and their process advantage and how a Taiwan semiconductor

00:02:51 ◼   ► was supposedly catching up, but Intel still had a lead and all sorts of stuff.

00:02:57 ◼   ► And we were seeing, we were going to see Intel, it was going to be first at 10 nanometer with

00:03:02 ◼   ► a significant type of chip, not just memory or something like that.

00:03:06 ◼   ► But as people who took apart the new iPad with the Apple's A10X and it discovered the

00:03:14 ◼   ► the ATENX system on the chip is using Tylon Semiconductor's 10nm process.

00:03:22 ◼   ► And ATENX is not a tiny chip, I mean, in the grand scheme of things it's not as big as

00:03:27 ◼   ► a Xeon, but it's pretty complicated, right?

00:03:29 ◼   ► It's got multiple cores and a GPU and I/O, it's not as simple as a memory chip.

00:03:34 ◼   ► So I haven't been keeping up with Intel and its various woes that much.

00:03:40 ◼   ► But bottom line, of all the Apple devices we care about, which includes PCs and iPads

00:03:48 ◼   ► and watches and phones and all sorts of other things, the first one, as far as I'm aware,

00:03:53 ◼   ► to get a 10nm chip is this one, and the chip is not from Intel.

00:03:57 ◼   ► So whether or not Intel has "lost its lead in process," it continues to be embarrassing

00:04:05 ◼   ► that Apple's system on a chip, a XX line of chips, are so fast and so low power and now are on a

00:04:16 ◼   ► smaller process than the chips that are in all of Apple's Macs. So it's kind of exciting/depressing,

00:04:23 ◼   ► depending on how you look at it. So remind me, and I'm not being funny at all,

00:04:27 ◼   ► what is the 10 nanometer represent? The lines within the processor, like the quote-unquote

00:04:33 ◼   ► wires is that correct or is that not even close?

00:04:35 ◼   ► Oh I don't know they call it feature size I'm sure there's some standard for measuring

00:04:38 ◼   ► from you know this thing to that thing like I think it was like a dot pitch on monitors

00:04:43 ◼   ► I'm sure there's some standard because you say well what is a feature like it's from

00:04:46 ◼   ► you know from one transistor to another how closely you can pack them together or like

00:04:50 ◼   ► I have no idea. Maybe it's like the center of one to the center of the next one? Yeah

00:04:54 ◼   ► but but there's different geometries like this is their FinFET thing like they're actually

00:04:57 ◼   ► 3D if you look at the shape of where the source and drain and gate and everything is they

00:05:02 ◼   ► They all have all these clever arrangements of layers of stuff.

00:05:05 ◼   ► Someone should look up, you can probably just go to the Wikipedia and find out, but basically

00:05:08 ◼   ► if you think of it like dot pitch on a monitor, it's from, you know, how, from the center

00:05:14 ◼   ► of, like what Mark was saying, from the center of one thing to the center of the other, the

00:05:16 ◼   ► question is how do they determine that because I think there are different sizes and configurations

00:05:20 ◼   ► of things that you can put down on a chip.

00:05:23 ◼   ► I'm assuming it is like the smallest distance between two features.

00:05:28 ◼   ► So anyway, Taiwan Semiconductor doing pretty well.

00:05:33 ◼   ► Apple with its A-series chips doing pretty well.

00:05:36 ◼   ► Intel, I don't know, doing okay.

00:05:41 ◼   ► Kind of stumbling, it was the recent announcement

00:05:43 ◼   ► they were getting out of the wearables market

00:05:45 ◼   ► that they were never really in.

00:05:46 ◼   ► - Yeah, I was gonna say,

00:05:47 ◼   ► what did they have in the wearables market?

00:05:49 ◼   ► - They had a program and an initiative

00:05:53 ◼   ► and an attempt, like they've done many times.

00:05:55 ◼   ► Like remember when they tried to get

00:05:56 ◼   ► into the standalone GPU market with,

00:05:57 ◼   ► what the hell was that called?

00:06:00 ◼   ► Chat room will tell me in seven seconds or so.

00:06:02 ◼   ► - Yeah, like about five years ago maybe?

00:06:04 ◼   ► - Yeah, where they were taking a bunch of little x86 cores

00:06:07 ◼   ► and put them on a Larabee.

00:06:09 ◼   ► I got it before the chat room, haha.

00:06:11 ◼   ► And they were gonna make a run at the GPU market

00:06:14 ◼   ► and they had evangelists and they had APIs

00:06:16 ◼   ► and they had some silicon and it just never came together

00:06:19 ◼   ► and they said, "Oh, okay, well we're not gonna do that."

00:06:21 ◼   ► And then they concentrated on their internal GPUs,

00:06:25 ◼   ► like the integrated ones, and they've come a long way on that,

00:06:28 ◼   ► partially, supposedly, rumor has it,

00:06:31 ◼   ► at the behest of Apple asking for more powerful

00:06:34 ◼   ► integrated GPU so they don't have to include a discrete GPU

00:06:37 ◼   ► on every single Mac just to have any graphics performance

00:06:40 ◼   ► worth a damn.

00:06:40 ◼   ► They were gonna get into wearables,

00:06:43 ◼   ► and they had an initiative, and they had a team,

00:06:45 ◼   ► and they had a blah, blah, blah,

00:06:46 ◼   ► and now they've just bailed out.

00:06:48 ◼   ► They also had a bunch of things like getting into

00:06:51 ◼   ► set-top boxes back in the day,

00:06:54 ◼   ► What it's like of all the Intel failed programs

00:06:56 ◼   ► where they were going to enter a new market.

00:06:57 ◼   ► I mean, if you want to trace it all the way backwards,

00:06:59 ◼   ► it's when they divested of all their ARM holdings.

00:07:01 ◼   ► They had both at the-- another thing-- XScale.

00:07:04 ◼   ► They had ARM CPU.

00:07:06 ◼   ► And they said, now we're going to get out of that business.

00:07:09 ◼   ► Probably it's going to turn out to have been a bad idea.

00:07:12 ◼   ► But in theory, like-- I mean, it didn't have to be.

00:07:15 ◼   ► It could have been they got out of that business

00:07:17 ◼   ► and they made a bunch of their own x86 chips

00:07:21 ◼   ► that were just as good as the things that Apple's making out.

00:07:24 ◼   ► But they didn't, they didn't make those.

00:07:26 ◼   ► And Apple did, and so that's how you lose.

00:07:28 ◼   ► So you can trace it back and say,

00:07:29 ◼   ► "Oh, you should've stopped."

00:07:30 ◼   ► There's no guarantee that if they kept Xscale

00:07:31 ◼   ► that they would've been able to do

00:07:33 ◼   ► as good a job as Apple is doing now.

00:07:34 ◼   ► So it's a weird world we live in

00:07:36 ◼   ► for someone who grew up in the Wintel era.

00:07:38 ◼   ► But I'm still rooting for Intel

00:07:40 ◼   ► because they're all the CPUs and all the Macs,

00:07:41 ◼   ► and I want a really good Mac, so let's get going.

00:07:45 ◼   ► - Yeah, that's pretty much why I care about this stuff

00:07:48 ◼   ► so much, like I know that it's probably,

00:07:51 ◼   ► for lots of reasons we've discussed before,

00:07:52 ◼   ► it's probably unlikely that we're ever gonna see,

00:07:55 ◼   ► or at least I guess never say never,

00:07:56 ◼   ► but that anytime soon we're gonna see Apple

00:07:59 ◼   ► bring the A line of ARM CPUs to the Mac and have ARM Macs.

00:08:04 ◼   ► And that's kind of a shame because these CPUs are awesome.

00:08:08 ◼   ► Like they keep making these amazing chips

00:08:12 ◼   ► and with their in-house designs

00:08:14 ◼   ► that are kicking the butts of everything else

00:08:15 ◼   ► in the middle market.

00:08:17 ◼   ► And Intel is really having a hard time

00:08:21 ◼   ► getting incremental performance gains every time they release something, and they were

00:08:24 ◼   ► even having a hard time releasing anything.

00:08:27 ◼   ► So I kind of wish that more of this excitement would be available on the Mac side, because

00:08:33 ◼   ► who knows what they could do.

00:08:35 ◼   ► Right now they have this amazing A10X chip, based on the amazing A10 chip, making the

00:08:41 ◼   ► iPad very competitive with low-end Macs in CPU performance.

00:08:46 ◼   ► And imagine what they could do with that

00:08:48 ◼   ► if they let it run in a 150 watt desktop enclosure.

00:08:52 ◼   ► That would be, that could be really amazing.

00:08:55 ◼   ► But for lots of reasons we discussed before,

00:08:57 ◼   ► basically making ARM Macs is neither easy nor likely.

00:09:02 ◼   ► But I do think that's kind of a shame

00:09:04 ◼   ► because there's all this exciting stuff happening

00:09:06 ◼   ► over in the ARM world that us Mac fans

00:09:09 ◼   ► are just kind of not a part of or minimally a part of.

00:09:14 ◼   ► I continue to think that there is nothing specifically magical about ARM that guarantees

00:09:18 ◼   ► that if Apple did have a team that said, "Hey, you've got 150-watt power budget.

00:09:22 ◼   ► Go nuts."

00:09:23 ◼   ► They would just have to end up doing all the same things that Intel does with its Xeon.

00:09:27 ◼   ► Only Intel probably has a bunch of them patented, and Apple doesn't have experience with them,

00:09:30 ◼   ► because you do end up making different choices when you have a ridiculously high power budget

00:09:33 ◼   ► like that.

00:09:34 ◼   ► And there's no secret sauce that only Apple knows.

00:09:37 ◼   ► And a lot of the tricks they know from their experience with low-power stuff are not applicable

00:09:41 ◼   ► to 150-watt power envelopes.

00:09:43 ◼   ► So I think they would make good chips, no doubt.

00:09:46 ◼   ► And maybe they would make ones that are actually better

00:09:48 ◼   ► than what Intel offers.

00:09:49 ◼   ► But I don't feel like they would be crushing them

00:09:51 ◼   ► in the same way that they're crushing on the low power side.

00:09:55 ◼   ► Because this is not Apple's first chip.

00:09:57 ◼   ► What was it, the A6 was their first sort of their own design?

00:10:02 ◼   ► They've had years and years to get this good at system

00:10:04 ◼   ► on a chips in tablets and phones.

00:10:07 ◼   ► They have no experience so far in making a Xeon class

00:10:10 ◼   ► processor.

00:10:11 ◼   ► So even if they did make one, the best

00:10:13 ◼   ► could hope for is probably like parity or a slight edge on Intel's things because they

00:10:16 ◼   ► can make different choices. But give them four or five years and I think they'd be doing

00:10:20 ◼   ► really well. But who knows, then we wouldn't be able to run Windows as well as we can now,

00:10:25 ◼   ► so that would be sad.

00:10:26 ◼   ► Who runs Windows anymore? What year is this?

00:10:28 ◼   ► I do. Casey does.

00:10:29 ◼   ► I used to. No, I don't. I don't anymore. I used to.

00:10:31 ◼   ► Yeah, Casey's a real developer now.

00:10:33 ◼   ► Ooh, sick burn. But yay!

00:10:36 ◼   ► Yeah, my apologies to everyone who has to develop on Windows. I feel bad for you.

00:10:41 ◼   ► - Hey man, I maintain C# is an unbelievably good language,

00:10:44 ◼   ► it's just everything around it that's garbage.

00:10:46 ◼   ► - I agree, I completely agree.

00:10:48 ◼   ► Yeah, and Microsoft has almost always been really good

00:10:52 ◼   ► at developer tooling, like they're really good at that.

00:10:55 ◼   ► Just Windows is terrible,

00:10:57 ◼   ► but the developer tools are usually great, but yeah.

00:11:00 ◼   ► And one thing before we cap off this topic is that

00:11:04 ◼   ► I do think that even though my example of the 150 watt

00:11:07 ◼   ► Mac Pro slash high-end iMac chip

00:11:11 ◼   ► would maybe be hard for them to move to.

00:11:14 ◼   ► The A10X, the iPad class of chip that they're making now,

00:11:18 ◼   ► which is not that different from the iPhone class

00:11:20 ◼   ► chip they're making, is probably suitable

00:11:24 ◼   ► for at least the bottom two lines of MacBooks.

00:11:27 ◼   ► At least the MacBook 12 inch and at least the,

00:11:30 ◼   ► what I call the MacBook escape,

00:11:31 ◼   ► the non-touch bar 13 inch MacBook Pro.

00:11:34 ◼   ► I bet they could make an A series chip

00:11:36 ◼   ► that would be very competitive in those two form factors.

00:11:41 ◼   ► - Yeah, I mean, they already do, practically.

00:11:43 ◼   ► - Yeah, exactly.

00:11:44 ◼   ► - It's just a question of the GPU.

00:11:45 ◼   ► And by the way, it's making me think about Mac losing out,

00:11:48 ◼   ► like, oh, so they're missing out on Apple's,

00:11:50 ◼   ► on the potential of Apple's expertise

00:11:52 ◼   ► in the areas where they've been shown

00:11:53 ◼   ► they can make a really good chip,

00:11:54 ◼   ► and their potential expertise for higher bandwidth chips.

00:11:57 ◼   ► And also, Apple thus far is missing out

00:12:00 ◼   ► on the other place that the action is happening

00:12:01 ◼   ► that's relevant to the Mac in terms of silicon chips

00:12:05 ◼   ► that are inside there.

00:12:06 ◼   ► Nvidia GPUs.

00:12:07 ◼   ► That is currently where a lot of the exciting high-end action

00:12:10 ◼   ► is happening on desktop computers.

00:12:13 ◼   ► And Apple is thus far pretending it doesn't exist

00:12:17 ◼   ► and continuing to ship AMD/ATI GPUs, which are fine,

00:12:21 ◼   ► but it's not-- it's like the excitement is happening

00:12:24 ◼   ► on the ARM side with Apple's own things.

00:12:26 ◼   ► And in the GPU land, the discrete GPU land,

00:12:28 ◼   ► the excitement is happening on Nvidia's side of the fence.

00:12:31 ◼   ► And Apple is like, no, we're just going to go Intel and AMD,

00:12:34 ◼   ► and we'll just ignore all those fireworks

00:12:36 ◼   ► and then the happy laughter happening in other areas.

00:12:39 ◼   ► - And I do think ultimately one of the biggest reasons

00:12:44 ◼   ► why we're unlikely to see ARM Macs

00:12:46 ◼   ► for the foreseeable future is I really don't think

00:12:50 ◼   ► that a lot of Mac apps would be rewritten

00:12:54 ◼   ► or recompiled for ARM in a prompt manner.

00:12:57 ◼   ► I think the Mac is, some parts of the Mac are very healthy,

00:13:02 ◼   ► content creation, high-end productivity tools,

00:13:05 ◼   ► these are fairly healthy,

00:13:07 ◼   ► but there's a whole lot of Mac apps out there

00:13:11 ◼   ► that are basically unmaintained,

00:13:13 ◼   ► and some of which are very, very old.

00:13:14 ◼   ► And they continue to work now,

00:13:16 ◼   ► just because the architecture hasn't changed in a long time,

00:13:18 ◼   ► in over 10 years, and so they continue to work,

00:13:23 ◼   ► but I worry if Apple were to force an incompatible move

00:13:28 ◼   ► to the basis of the operating system,

00:13:30 ◼   ► I worry about what would happen to the app ecosystem.

00:13:34 ◼   ► I bet we'd lose a lot of apps.

00:13:36 ◼   ► As we're seeing right now with, I mean look, iOS,

00:13:38 ◼   ► in this way, iOS is way healthier than the Mac OS.

00:13:43 ◼   ► And the transition to iOS 11 is losing lots of apps.

00:13:46 ◼   ► Just from the 64-bit change, it was only a few years ago.

00:13:49 ◼   ► For the Mac to change CPU architectures,

00:13:53 ◼   ► it would be, I think, really difficult

00:13:57 ◼   ► to move a lot of Mac software over.

00:13:59 ◼   ► just because so much of it's un-maintained.

00:14:01 ◼   ► - Numerically, iOS obviously will lose more

00:14:03 ◼   ► 'cause it's numerically got so much more.

00:14:05 ◼   ► But on the Mac, I have some confidence.

00:14:07 ◼   ► The Mac has gone through this multiple times

00:14:09 ◼   ► before in the past that, yes, you always leave stuff behind,

00:14:11 ◼   ► but the apps that we care about would be ported,

00:14:14 ◼   ► just like they were ported from 68K to PowerPC,

00:14:17 ◼   ► and those same apps that we care about

00:14:19 ◼   ► that still aren't reported from PowerPC to Intel,

00:14:21 ◼   ► and they were ported from Intel 32-bit to 64-bit,

00:14:23 ◼   ► and they would be ported from 64-bit.

00:14:26 ◼   ► The apps we care about would be ported.

00:14:27 ◼   ► This is maybe like one or two random abandoned apps that you would lose.

00:14:31 ◼   ► But numerically speaking, on iOS there's just so many apps and we all have some.

00:14:36 ◼   ► I think, do you think you have more apps installed on, cumulatively on your iOS devices or on

00:14:41 ◼   ► your Macs?

00:14:42 ◼   ► Like numerically, each application icon counts as one thing.

00:14:45 ◼   ► Oh, it's no question, iOS.

00:14:47 ◼   ► I would guess the same.

00:14:49 ◼   ► Yeah, yeah, yeah, probably.

00:14:51 ◼   ► Maybe not on my iPad.

00:14:52 ◼   ► On my iPhone, yes.

00:14:53 ◼   ► I think my iPad is relatively lean and mean,

00:14:55 ◼   ► but iPhone, yeah, I think Marco's right.

00:14:58 ◼   ► - How many applications are in your applications folder?

00:15:01 ◼   ► On your Mac, just pick a random Mac.

00:15:03 ◼   ► - I have 129.

00:15:05 ◼   ► - 112.

00:15:06 ◼   ► - Although many of these I barely even recognize,

00:15:08 ◼   ► so I long show infrequently.

00:15:10 ◼   ► - I know, we all have junk in there.

00:15:12 ◼   ► Oh, this is reassuring.

00:15:14 ◼   ► So in my applications folder,

00:15:17 ◼   ► I have two applications called system preferences,

00:15:20 ◼   ► which apparently have exactly the same name.

00:15:22 ◼   ► - That's good.

00:15:24 ◼   ► - This does not make me feel good.

00:15:27 ◼   ► This is from 2014, and the second one is from 2015 and 2016.

00:15:31 ◼   ► And one of them does not have a .app extension.

00:15:37 ◼   ► That makes me feel good.

00:15:39 ◼   ► - I gotta clear a lot of this crap out of here, my God.

00:15:42 ◼   ► - Anyway, you wanna guess how many applications

00:15:44 ◼   ► I have in my application?

00:15:44 ◼   ► - I mean, if I have 129 and I keep things pretty minimal,

00:15:48 ◼   ► I'm gonna guess you have 200 maybe?

00:15:51 ◼   ► - Yeah, I'm gonna say 250-ish.

00:15:52 ◼   ► Is it Price is Right rules?

00:15:54 ◼   ► - No.

00:15:55 ◼   ► - Oh, okay, then I don't have to bet one application.

00:15:58 ◼   ► - 311.

00:16:00 ◼   ► - Oof.

00:16:01 ◼   ► - Wow.

00:16:02 ◼   ► - Goodness.

00:16:03 ◼   ► - But how many do you actually use on a regular basis?

00:16:06 ◼   ► - Not 311.

00:16:07 ◼   ► - I probably use something like 20 or 30 on a regular basis.

00:16:12 ◼   ► I mean, the Mac tools tend to be larger in scope

00:16:16 ◼   ► and fewer in number for a lot of people.

00:16:20 ◼   ► I would bet.

00:16:22 ◼   ► There's a lot more like large productivity tools,

00:16:25 ◼   ► large creation tools, large library tools,

00:16:27 ◼   ► as opposed to an iOS where you have a whole bunch of stuff

00:16:30 ◼   ► that does like little things or one thing at a time.

00:16:33 ◼   ► Not to mention, iOS has games,

00:16:35 ◼   ► and the Mac unfortunately doesn't.

00:16:38 ◼   ► - I do have a subfolder for games, let's see what's in that.

00:16:42 ◼   ► - I have a subfolder for Mac ports that I haven't touched.

00:16:44 ◼   ► - I mean, I have a Steam icon there too, all right?

00:16:47 ◼   ► So the games folder only have 42 items,

00:16:49 ◼   ► but there are subfolders in there,

00:16:50 ◼   ► So it just goes on and on and on.

00:16:52 ◼   ► I got a lot of-- here's the thing.

00:16:54 ◼   ► The reason I have so many applications

00:16:55 ◼   ► is they don't-- on the Mac, applications

00:16:58 ◼   ► aren't in your face like they are in iOS.

00:17:00 ◼   ► Like in Springboard, they're in your face

00:17:02 ◼   ► unless you bury them in a folder, like all of them are.

00:17:05 ◼   ► I don't watch applications if I go to the application folder.

00:17:07 ◼   ► Nobody does.

00:17:08 ◼   ► You do Command Space or whatever, right?

00:17:09 ◼   ► Yeah, of course.

00:17:10 ◼   ► And whenever I use Disk Inventory X or whatever

00:17:13 ◼   ► to find out where all my space is being taken up,

00:17:15 ◼   ► it's never with applications.

00:17:16 ◼   ► Because in the grand scheme of things, they're tiny.

00:17:18 ◼   ► my one terabyte drive, even a huge application.

00:17:21 ◼   ► Maybe occasionally, like Marco, I

00:17:22 ◼   ► get cranky at some multi-gigabyte sound samples

00:17:25 ◼   ► that are part of some stupid Apple application.

00:17:27 ◼   ► But other than that, all the data is elsewhere.

00:17:31 ◼   ► So I never ended up deleting applications.

00:17:33 ◼   ► I look in here, and I've got-- I don't know.

00:17:36 ◼   ► Multiple old-- I've got Delicious Library 2 and 3.

00:17:39 ◼   ► I've got many applications where they're

00:17:41 ◼   ► like the history of the application

00:17:42 ◼   ► is present in the folder as the versions and the numbers

00:17:45 ◼   ► and the icons go up.

00:17:46 ◼   ► And I never ended up deleting them because it's not as,

00:17:49 ◼   ► and don't get me wrong, my iOS home screens

00:17:51 ◼   ► are also a giant graveyard of ancient files,

00:17:54 ◼   ► which is why I'm kind of dreading

00:17:55 ◼   ► when they drop 32-bit support, 'cause it's gonna,

00:17:57 ◼   ► well, dreading and looking forward to it,

00:17:59 ◼   ► 'cause it'll reduce the number of applications by half,

00:18:01 ◼   ► because I think I still have applications there

00:18:03 ◼   ► from my original iPod Touch that have just been

00:18:06 ◼   ► carrying along from iTunes backup to iTunes backup.

00:18:08 ◼   ► - I only had, I think, two.

00:18:10 ◼   ► Yeah, I have Scorekeeper XL and iCast Pro.

00:18:14 ◼   ► and the scorekeeper developer said he's updating it.

00:18:17 ◼   ► iCast has already done a new version

00:18:19 ◼   ► that's like a million dollars, so,

00:18:20 ◼   ► you know, they're not actually really going to be lost.

00:18:24 ◼   ► - Is iCast the thing you use to do

00:18:26 ◼   ► the live broadcasts from the road?

00:18:29 ◼   ► - I used to, yes.

00:18:30 ◼   ► It's an iCast server, or an iCast broadcaster, rather.

00:18:34 ◼   ► So I used to do that from the road.

00:18:36 ◼   ► Now I just bring a Mac and use

00:18:38 ◼   ► Vrgaminba's nicecast app, 'cause it's,

00:18:40 ◼   ► 'cause it really, I mean, what I learned, you know,

00:18:43 ◼   ► while doing our live show and setting up for that

00:18:46 ◼   ► and doing various things over the last few years,

00:18:48 ◼   ► I've learned basically that I greatly prefer Macs

00:18:52 ◼   ► to be in those kind of production roles

00:18:55 ◼   ► and not just for superstition of Macs being reliable

00:18:58 ◼   ► and iOS not being, but just for the flexibility they offer.

00:19:01 ◼   ► And you can do these things on iOS if you need to,

00:19:05 ◼   ► but I'm much happier and more versatile

00:19:08 ◼   ► when I'm doing it on the Mac, if that makes sense.

00:19:10 ◼   ► I mean, we all know that you can't get work done on iOS,

00:19:13 ◼   ► so that's why you're doing it on a Mac.

00:19:15 ◼   ► - Oh my God, I still have Firefox?

00:19:17 ◼   ► (laughing)

00:19:19 ◼   ► Does anybody still use Firefox?

00:19:21 ◼   ► - So we've gone during the pre-show

00:19:23 ◼   ► that hopefully has not hit the regular show

00:19:26 ◼   ► of watching TV as a group,

00:19:28 ◼   ► and now we're letting all the people listen to us

00:19:31 ◼   ► plunge through and go spelunking

00:19:33 ◼   ► through our application folders as a group.

00:19:35 ◼   ► This is the best--

00:19:36 ◼   ► - I still have MakeMKV.

00:19:36 ◼   ► I haven't ripped a Blu-ray in years.

00:19:38 ◼   ► - I just did that yesterday.

00:19:40 ◼   ► - I have Firefox, I also have MakeMKV.

00:19:43 ◼   ► - Yeah, I just ripped Days of Thunder yesterday.

00:19:46 ◼   ► It's a great movie.

00:19:47 ◼   ► - See, I don't need to rip anything,

00:19:48 ◼   ► I can just play some off of your Plex server.

00:19:51 ◼   ► - I know, that's exactly what I was doing.

00:19:53 ◼   ► We bought Days of Thunder on Blu-ray,

00:19:55 ◼   ► but Blu-rays are barbaric, and so I immediately ripped it,

00:19:59 ◼   ► and it is now on Plex.

00:20:00 ◼   ► - Ooh, I still have OpenTTD, I should play that.

00:20:02 ◼   ► The Transport Tycoon.

00:20:03 ◼   ► - Oh, so good.

00:20:04 ◼   ► I have applications with the circle with a line through it,

00:20:09 ◼   ► I should just probably delete those.

00:20:10 ◼   ► - I don't think I've ever even seen that.

00:20:12 ◼   ► - Did not run on my fancy new version of Mac OS

00:20:17 ◼   ► called El Capitan.

00:20:18 ◼   ► - Man, I can't even imagine cleaning out

00:20:20 ◼   ► John's 10 year old computer like that.

00:20:23 ◼   ► What you must find there.

00:20:24 ◼   ► - Just never clean it, that's the secret.

00:20:26 ◼   ► - We are sponsored this week by Fracture,

00:20:30 ◼   ► a photo printing company that prints photos

00:20:32 ◼   ► directly onto glass.

00:20:33 ◼   ► Visit fractureme.com/podcast for 10% off your first order.

00:20:38 ◼   ► Fracture is a photo decor company that's out to rescue your images from the digital ether.

00:20:43 ◼   ► They don't, they know, if you just post a picture to Facebook or whatever, it's gone

00:20:46 ◼   ► off people's timelines in what, a day at most? And you never see it again. Fracture wants

00:20:51 ◼   ► you to take those favorite pictures and get them printed and hang them in your house or

00:20:55 ◼   ► give them to people as gifts. They make wonderful gifts for any occasion for family and friends

00:21:00 ◼   ► and they really make your house look lived in and they show your memories to you and

00:21:04 ◼   ► everyone who comes over. We have these all over our house, they get tons of compliments.

00:21:08 ◼   ► It's nice to get photos printed once in a while.

00:21:10 ◼   ► It really is.

00:21:11 ◼   ► And Fracture is the best and easiest way to do that.

00:21:13 ◼   ► And these are printed on glass.

00:21:15 ◼   ► So it goes edge to edge.

00:21:17 ◼   ► The colors look great.

00:21:18 ◼   ► They really pop.

00:21:19 ◼   ► The contrast really pops.

00:21:20 ◼   ► And it just looks sleek and it looks modern.

00:21:23 ◼   ► You don't have to get them framed because they are their own independent object.

00:21:25 ◼   ► It's wonderful.

00:21:26 ◼   ► They have a laser cut rigid backing.

00:21:29 ◼   ► So the whole thing isn't glass front to back because that would be really heavy and fragile.

00:21:33 ◼   ► It's glass in the front and there's rigid backing on the back.

00:21:36 ◼   ► so it's very lightweight and easy to hang.

00:21:38 ◼   ► They even include a wall anchor right in the box.

00:21:41 ◼   ► It's so easy to get these things.

00:21:42 ◼   ► All you have to do is upload a picture and pick your size.

00:21:45 ◼   ► That's it.

00:21:47 ◼   ► So you can bring a special memory to life with these prints,

00:21:49 ◼   ► give them as unique gifts, or decorate your home

00:21:51 ◼   ► with the moments that tell your story.

00:21:53 ◼   ► Every Fracture print comes with a 60-day happiness guarantee

00:21:56 ◼   ► so you're sure to love your order,

00:21:58 ◼   ► and they're all handmade in Gainesville, Florida

00:22:00 ◼   ► from US-sourced materials in a carbon-neutral factory.

00:22:03 ◼   ► For more information and 10% off your first order,

00:22:06 ◼   ► visit fractureme.com/podcast.

00:22:09 ◼   ► The last you wish podcast sent you, tell them ATP,

00:22:12 ◼   ► and that'll help support the show.

00:22:13 ◼   ► So thank you very much to Fracture

00:22:15 ◼   ► for sponsoring our show once again.

00:22:17 ◼   ► - So in the show notes, near-ish the top,

00:22:25 ◼   ► but never ever ever high enough for us

00:22:28 ◼   ► to actually talk about it, has been three words,

00:22:31 ◼   ► and it has been here four months.

00:22:35 ◼   ► And those three words are Netflix credits skipping.

00:22:39 ◼   ► And apparently this is enough, and Jon is flustered enough,

00:22:43 ◼   ► that he would like to talk to us about skipping credits

00:22:47 ◼   ► in Netflix.

00:22:48 ◼   ► I don't know why I'm in for this ride as much as you guys are.

00:22:52 ◼   ► So let's all buckle up together.

00:22:54 ◼   ► We are really scraping the bottom of the barrel here.

00:22:56 ◼   ► We really are.

00:22:57 ◼   ► You know why this is in there.

00:22:59 ◼   ► I mean, the link there explains it.

00:23:01 ◼   ► So this is from March.

00:23:02 ◼   ► So yes, it has been there for many, many months.

00:23:05 ◼   ► The story-- we'll link to the Verge thing--

00:23:07 ◼   ► is by now, that's the only URL we have for the story,

00:23:09 ◼   ► is that Netflix, in their normal way,

00:23:11 ◼   ► was testing this new feature on some platforms

00:23:15 ◼   ► and some clients for Netflix.

00:23:16 ◼   ► That when you're watching a show on Netflix

00:23:19 ◼   ► and the credits come on, the opening credits to the show,

00:23:22 ◼   ► there would be a button in the interface somewhere

00:23:25 ◼   ► that says skip credits.

00:23:27 ◼   ► And if you click that, it would skip it.

00:23:28 ◼   ► When we were in Chicago, we were finishing up House of Cards.

00:23:31 ◼   ► And I had done that by connecting my MacBook Adorable

00:23:34 ◼   ► to the TV.

00:23:35 ◼   ► And I saw and used that button.

00:23:37 ◼   ► And so that was on the web client.

00:23:39 ◼   ► Yeah, and I'm not sure where they're rolling it out to

00:23:41 ◼   ► and what clients or whatever.

00:23:43 ◼   ► But it's not everywhere.

00:23:44 ◼   ► And it seems like it's the thing they're testing.

00:23:45 ◼   ► Now, already for several years, Netflix has had a thing

00:23:48 ◼   ► where if you're binge watching and you watch one episode

00:23:50 ◼   ► and the next episode's gonna come,

00:23:52 ◼   ► that it will not show the credits to the subsequent episode.

00:23:55 ◼   ► So it'll just get right to the program.

00:23:56 ◼   ► I've seen that.

00:23:57 ◼   ► I don't know if that's show specific or whatever,

00:23:58 ◼   ► but it's been doing that for a long time.

00:24:00 ◼   ► So if you sit down and watch five episodes of a program,

00:24:03 ◼   ► you don't have to see the opening credits five times.

00:24:04 ◼   ► But you would have to see it once for the first one.

00:24:07 ◼   ► And now with the skip credits thing,

00:24:08 ◼   ► it'll be like if you don't want to see the opening credits

00:24:10 ◼   ► to House of Cards because you're afraid you'll die of old age,

00:24:15 ◼   ► and you're on the right interface,

00:24:17 ◼   ► you can click this button to skip it.

00:24:19 ◼   ► And the reason I put this in the notes so long ago is that--

00:24:26 ◼   ► four years ago?

00:24:27 ◼   ► Four years before this feature was added to Netflix,

00:24:30 ◼   ► I, in one of my rare fits of blogging,

00:24:32 ◼   ► wrote a blog post about this very thing,

00:24:34 ◼   ► about the idea of credit skipping

00:24:38 ◼   ► in Netflix-type environments specifically.

00:24:40 ◼   ► It was also in the context of the PlayStation 4,

00:24:42 ◼   ► which was not yet out, but had been announced,

00:24:44 ◼   ► and the features had been announced,

00:24:45 ◼   ► and the title of this post is Annoyance-Driven Development,

00:24:48 ◼   ► the idea that you look at a technology product

00:24:52 ◼   ► or something that you're working on

00:24:53 ◼   ► and think about the things that annoy you and other people

00:24:57 ◼   ► about your product and try to knock those down one by one,

00:25:01 ◼   ► even if they seem silly.

00:25:03 ◼   ► I think-- refreshing my memory on this thing--

00:25:05 ◼   ► that the PlayStation examples were things

00:25:08 ◼   ► like being able to pick up a game right where you left off

00:25:10 ◼   ► and not having the thing start downloading a software update

00:25:15 ◼   ► as soon as you turn it on.

00:25:16 ◼   ► And so they did a lot of things in the PlayStation 4

00:25:18 ◼   ► to remove the annoyances of the PlayStation 3

00:25:21 ◼   ► by giving it a low power mode and an auxiliary processor

00:25:24 ◼   ► that let it do useful things when it's quote unquote off.

00:25:26 ◼   ► Like the fans are off, it's sitting there,

00:25:28 ◼   ► you think it's completely off, but it's not.

00:25:30 ◼   ► It's downloading software updates in the background.

00:25:32 ◼   ► It's preserving the state of your game

00:25:34 ◼   ► so you can pick up right where you left off.

00:25:36 ◼   ► Even Nintendo has come on this bandwagon

00:25:37 ◼   ► a little bit with the Switch

00:25:38 ◼   ► where you can put the thing into sleep mode,

00:25:41 ◼   ► pick it up, get it out of sleep mode,

00:25:42 ◼   ► and you are exactly at the second you left off

00:25:44 ◼   ► in your Zelda game,

00:25:46 ◼   ► which makes you wanna pick it up and play

00:25:48 ◼   ► more than if you had to wait through a boot sequence

00:25:50 ◼   ► launch the game again or something like that.

00:25:52 ◼   ► And software updates, which are ubiquitous nowadays, having to download multiple gigabytes

00:25:58 ◼   ► of stuff while you sleep so that when you wake up it's all just already there for you

00:26:02 ◼   ► is a great example of annoying driven development.

00:26:04 ◼   ► Because it's not the end of the world, like, oh, software updates have to happen and there's

00:26:06 ◼   ► no getting around the fact that you've got to have multiple gigabytes and we can't make

00:26:09 ◼   ► your internet connection faster.

00:26:10 ◼   ► It's like, but can we design the PlayStation 4 in a way to make this less annoying for

00:26:15 ◼   ► people?

00:26:16 ◼   ► so far I still have a separate mode with a separate CPU and a low power state of like

00:26:21 ◼   ► it complicates the hardware tremendously but it's a big win.

00:26:24 ◼   ► So in terms of streaming video, I was using House of Cards as an example because the credit

00:26:29 ◼   ► sequence is really, really long.

00:26:30 ◼   ► I was saying how Netflix has changed how you watch television and how you have to reconsider

00:26:35 ◼   ► every part of the experience and think about how annoying it is.

00:26:37 ◼   ► And the opening credits aren't annoying in the old world of television where there were

00:26:41 ◼   ► no DVRs and you just watched television when it was on and that was it.

00:26:46 ◼   ► were annoying but that's all a separate issue. But already with people with Netflix and streaming

00:26:50 ◼   ► video they understand that people do binge watching and they did that thing where it

00:26:53 ◼   ► will skip the credits to the next episode. They just need to go to the next step. Also

00:26:57 ◼   ► by the way, releasing entire seasons all at once instead of doling them out an episode

00:27:00 ◼   ► at the time. They reconsidered so many things and this seemed like the very next thing on

00:27:04 ◼   ► the chopping block which is, hey, credits. They make sense when you are watching television

00:27:09 ◼   ► in the old world. In the new world do they make as much sense to have opening credits

00:27:13 ◼   ► before every single episode or before any episode.

00:27:16 ◼   ► And I know, as I said in this post,

00:27:17 ◼   ► there's lots of things about union contracts

00:27:20 ◼   ► for screen actors and writers and everything.

00:27:22 ◼   ► And then there's the other angle of like,

00:27:24 ◼   ► oh, people should get credit for their work and blah, blah,

00:27:25 ◼   ► blah, like basically the motivation behind

00:27:28 ◼   ► all those union rules and everything.

00:27:29 ◼   ► But I think you can't, you have to change,

00:27:34 ◼   ► you have to change your outlook and the rules

00:27:37 ◼   ► surrounding all this stuff as the technology changes.

00:27:39 ◼   ► And technology is changing how we watch television.

00:27:42 ◼   ► And so every single aspect of how we watch television should be up for grabs for reconsideration to say is this the best way to do it?

00:27:49 ◼   ► Or is there a better way given how technology has changed how we watch television apparently Netflix eventually how many years later?

00:27:56 ◼   ► four years later eventually more or less agrees with me that

00:28:00 ◼   ► Maybe the credits are a little bit too much and maybe we should give people the option of skipping them entirely if they want

00:28:06 ◼   ► There's already no commercials. So they've already got that going for them

00:28:09 ◼   ► So anyway, I was excited enough about this in March to put it in the in the show notes and to leave it there

00:28:15 ◼   ► And now I finally get to talk about it

00:28:17 ◼   ► And I feel like this is I'm not gonna say it's vindication because I know a lot of people

00:28:22 ◼   ► agreed with me when I originally posted this but it's more of a celebration to say

00:28:26 ◼   ► You know, I'm I'm actually getting the thing that I was asking for and it seemed inevitable because Netflix is very aggressive these type of

00:28:32 ◼   ► Things I'm just I'm actually surprised it took four years for them to do it, but more like this, please

00:28:37 ◼   ► Wow.

00:28:38 ◼   ► Cool.

00:28:39 ◼   ► I feel better for having listened to that.

00:28:46 ◼   ► Nothing you said was wrong, but I felt like this has been sitting in the show notes, like

00:28:50 ◼   ► almost there, but not quite there for months, and I was expecting some sort of level.

00:28:56 ◼   ► It's not timely.

00:28:57 ◼   ► It's not like, you know, it doesn't go bad.

00:28:58 ◼   ► Oh, Jon, I love you so much.

00:29:00 ◼   ► And you don't have a product to speak of now that...

00:29:04 ◼   ► What did Dr. Wave say to your application?

00:29:07 ◼   ► was? Was it a fart app? Oh, yes. I was so confused.

00:29:12 ◼   ► Well, it's kind of like I got farting. I don't know. Anyway, Marko's got a problem. Anybody

00:29:17 ◼   ► who's listening to this who has a product or is involved in making a thing that people

00:29:20 ◼   ► use, it's a potential edge in the market. Think about what is annoying about using products

00:29:27 ◼   ► of your type, and in the next version of your product, try to eliminate those annoyances

00:29:31 ◼   ► and don't keep any sacred cows like, "Oh, we can't change that." You'll find a lot of

00:29:36 ◼   ► You'll find a lot of people hold sacred many things they won't even consider changing.

00:29:40 ◼   ► Like, you have to consider everything and reconsider it in the face of technology and

00:29:44 ◼   ► find the annoying things and be willing to do what Sony did.

00:29:49 ◼   ► Make your product way more complicated and way more expensive, potentially introducing

00:29:54 ◼   ► bugs, because the payoff is when people turn on the PlayStation 4, the 11 gigabytes of

00:29:59 ◼   ► the Destiny 2 beta is already downloaded and the system update has already happened and

00:30:04 ◼   ► they can start playing immediately.

00:30:05 ◼   ► Even if it takes forever to download, if it does it while you're sleeping, you don't care.

00:30:09 ◼   ► And it's 100% worth it to add an entire separate CPU and low power state and special OS mode

00:30:14 ◼   ► for this.

00:30:16 ◼   ► And also skip those House of Cards credits because come on, we've all seen them enough

00:30:19 ◼   ► now.

00:30:20 ◼   ► So long.

00:30:21 ◼   ► I mean, they're beautiful, but they're so long.

00:30:22 ◼   ► Oh my word, it takes forever.

00:30:24 ◼   ► So long.

00:30:25 ◼   ► So long.

00:30:26 ◼   ► All right, so let's talk about something that always cheers us up.

00:30:31 ◼   ► Twitter!

00:30:32 ◼   ► And a while back, there was...

00:30:33 ◼   ► God, I can't even remember when this was.

00:30:35 ◼   ► How long was it? How long was a while ago, Casey?

00:30:37 ◼   ► Apparently it was October of 2016.

00:30:40 ◼   ► Why are we talking about a news article from October 2016?

00:30:42 ◼   ► I have a reason for this one too.

00:30:44 ◼   ► Oh my god.

00:30:45 ◼   ► Bring it out, you're dead. That's right.

00:30:47 ◼   ► I'm so sorry, listeners, for this entire episode.

00:30:51 ◼   ► You apologize here. Where do episodes come from, Marco? When three podcasters love each

00:30:56 ◼   ► other very much? Somehow, every week there's an episode, and it's because somebody, some

00:31:03 ◼   ► elves go into the show notes and find a bunch of topics and put them in there. And if you're

00:31:08 ◼   ► not going to do it, guess what? I'm going to do it. And then you don't get to complain

00:31:11 ◼   ► about the things that we talk about and try to apologize for the listeners for things

00:31:14 ◼   ► we're talking about. Because guess what? I build the show in the notes and that's what

00:31:18 ◼   ► we're going to talk about. If you want to talk about something, feel free to rotate

00:31:20 ◼   ► something up to the top and then we'll talk about that instead. In fact, I will offer

00:31:24 ◼   ► that opportunity to you now because I just talked about Netflix credit shipping. Is there

00:31:27 ◼   ► something that you would like to talk about that you would like to move up?

00:31:31 ◼   ► Let me see.

00:31:32 ◼   ► Overcast Springboard Crasher stories? That's a good one.

00:31:34 ◼   ► It's kind of boring.

00:31:35 ◼   ► No, it's not.

00:31:36 ◼   ► I think it's exciting.

00:31:37 ◼   ► It's exciting.

00:31:38 ◼   ► It was exciting when it happened, but now I've forgotten most about it and probably

00:31:41 ◼   ► you do too. But it was cool.

00:31:42 ◼   ► It's not even old. It's not even old by comparison to October of 2016.

00:31:46 ◼   ► It was a couple of months ago.

00:31:49 ◼   ► In a surprising turn of events, Marco did not immediately Command-A and then delete

00:31:53 ◼   ► the entire show notes in a fit of rage just moments ago.

00:31:58 ◼   ► (laughing)

00:31:59 ◼   ► - Oh my word.

00:32:00 ◼   ► - The secret is he doesn't look at them most of the time.

00:32:01 ◼   ► So that's-- - That is accurate.

00:32:03 ◼   ► - So down, if we scroll down a couple pages,

00:32:04 ◼   ► we have an item that says-- - Oh God.

00:32:06 ◼   ► - TSMC ahead of Intel on process soon, question mark?

00:32:09 ◼   ► - Look at that, the show notes.

00:32:10 ◼   ► (laughing)

00:32:11 ◼   ► They're like a crystal ball.

00:32:13 ◼   ► (laughing)

00:32:15 ◼   ► - Look at the slug on this.

00:32:17 ◼   ► 10 nanometer chip foundry process coming to Apple partner

00:32:21 ◼   ► TM TSMC ahead of Intel.

00:32:23 ◼   ► - And it was from the ever reliable news source.

00:32:26 ◼   ► AppleInsider.com.

00:32:27 ◼   ► This is dated September of '16.

00:32:30 ◼   ► Oh my word.

00:32:31 ◼   ► We didn't talk about it, but guess what?

00:32:33 ◼   ► They did it.

00:32:34 ◼   ► It came true.

00:32:35 ◼   ► Maybe this is how we can make things come true.

00:32:36 ◼   ► Just put them in our show notes.

00:32:37 ◼   ► Put them in the show notes and it's like burying a time capsule.

00:32:42 ◼   ► Our show notes are the crock pot of Apple news.

00:32:46 ◼   ► Just wait long enough.

00:32:47 ◼   ► It's going to be there.

00:32:48 ◼   ► Oh my God.

00:32:50 ◼   ► All right, so let's talk about Disney since Marco hasn't come up with something better.

00:32:54 ◼   ► It'll be the Marco show after this when we talk about overcast and springboard.

00:32:57 ◼   ► Marco should start remembering now all the details of the springboard graduate stuff.

00:33:02 ◼   ► Did you know that letterpress moved off of Game Center?

00:33:04 ◼   ► I did.

00:33:05 ◼   ► I didn't know that.

00:33:06 ◼   ► That's down here too.

00:33:07 ◼   ► We should just—you know what we should do?

00:33:12 ◼   ► If we get really desperate, we should go from the bottom up and just say, "We're dedicated.

00:33:16 ◼   ► We have to do it.

00:33:17 ◼   ► We have to do it.

00:33:18 ◼   ► Just start at the bottom."

00:33:19 ◼   ► Oh, my word.

00:33:21 ◼   ► This is so bad.

00:33:22 ◼   ► Anyway, all right, I'm just trying to bring this back around.

00:33:25 ◼   ► So at some point one day eventually, or years ago it seems, there was some talk in late

00:33:31 ◼   ► 2016 about Twitter potentially looking to be bought.

00:33:36 ◼   ► And there were some potential suitors, I don't remember who they were other than Disney.

00:33:39 ◼   ► What, some Salesforce in the running?

00:33:42 ◼   ► Ah, maybe.

00:33:43 ◼   ► And so what ended up happening, and I didn't get a chance to look at these show notes and

00:33:46 ◼   ► refresh my memory, so being Chief Summarizer-in-Chief is a little dangerous today, but my recollection

00:33:52 ◼   ► is that Disney was kind of sniffing around, saying, "Eh, maybe we do want Twitter," and

00:33:58 ◼   ► then realized, "Oh, wait, Twitter's a cesspool of disgusting, terrible people, so yeah, we

00:34:03 ◼   ► don't want anything to do with that."

00:34:04 ◼   ► Well, the reason I put this in there, not because I don't care about this, whatever,

00:34:08 ◼   ► people buying Disney, that's old news.

00:34:11 ◼   ► I don't know why it stayed in the notes, I probably would have deleted it, but I saw

00:34:13 ◼   ► it down there, and it brought to mind something, what was it, it was relevant, it's eternally

00:34:20 ◼   ► relevant. I forget what the specific inciting incident was, but it's about when you have

00:34:25 ◼   ► an online community of any kind. Oh, I know what it was. It was the PUBG thing. You guys don't

00:34:33 ◼   ► follow this. Do you know what that is? Nope.

00:34:35 ◼   ► Anyways, it's a new game, Player Unknown Battlegrounds. It's a game. It's a cool game.

00:34:40 ◼   ► You'll like the premise of it. We love games.

00:34:43 ◼   ► Yeah. Well, no, it's a game mode. I think it might've started as a mod and now it's

00:34:48 ◼   ► it's his own game, I'd always get confused about the details.

00:34:50 ◼   ► But anyway, you are like an army guy.

00:34:52 ◼   ► You parachute into this fairly realistic looking world,

00:34:57 ◼   ► like a big open field with some trees and buildings

00:34:59 ◼   ► and stuff like that, and you have to sort of scrounge

00:35:02 ◼   ► for weapons and bandages and stuff like that,

00:35:05 ◼   ► and it's fairly realistic in that you can't take

00:35:07 ◼   ► 500 bullets, like a few bullets kills you, right?

00:35:10 ◼   ► And there are teams, and there is sort of a gamey,

00:35:16 ◼   ► shimmery circle, ring, dome around the entire play area that starts off really, really big,

00:35:22 ◼   ► but it slowly shrinks over time, eventually corralling all the people together. And eventually

00:35:28 ◼   ► there's one person who's left alive, right? So your strategy could be, I'm just going to crawl

00:35:32 ◼   ► in the grass and hide behind this tree and be like a mile away from everybody, but the circle

00:35:35 ◼   ► will shrink and shrink and shrink. And if you're not out there fighting, you're also not getting

00:35:38 ◼   ► better weapons and getting healing supplies and stuff like that. So it's a fun game mode.

00:35:45 ◼   ► And they're fun videos to watch because there are teams in this friendly fire and there's no radar or anything like that.

00:35:50 ◼   ► You're just crawling around and going through buildings and some people camp out in buildings and you just go around a door and there's a guy.

00:35:55 ◼   ► Anyway, it's a very popular game right now.

00:35:57 ◼   ► And the guy who made the game has community standards for his game, basically saying if you intentionally kill people on your team because friendly fire is a thing, that's against the rules and you get banned.

00:36:09 ◼   ► Temporarily banned and I assume you keep doing it permanently banned.

00:36:12 ◼   ► Because that's bad behavior. We don't want someone who comes to you join a game and the guy in your team kills everybody on his own

00:36:16 ◼   ► Teams takes their stuff like no, we don't have that right and he did that to like a popular streamer like twitch streamer

00:36:23 ◼   ► and the switch streamer was cranky about it and

00:36:26 ◼   ► Said some jokey things to him about like how he's gonna kick him in the face or something like that

00:36:30 ◼   ► and the guy

00:36:33 ◼   ► Because he had temporarily banned him and he explained to him

00:36:36 ◼   ► Why he doesn't tolerate even jokey sort of threats of violence or whatever

00:36:40 ◼   ► And anyway, that was the story like game creator,

00:36:44 ◼   ► bands popular streamer and has to explain to him why he doesn't

00:36:48 ◼   ► think threats of violence are funny, you know, counter to everything

00:36:52 ◼   ► and macho stupid gamer dude culture. Right.

00:36:55 ◼   ► And it made me think of Twitter and anybody who has any kind of community,

00:36:59 ◼   ► whether it's the game that you made and the standards

00:37:01 ◼   ► you make for the people who play your game online or Twitter,

00:37:03 ◼   ► who makes a service that they let people use or, you know, just think of anything

00:37:07 ◼   ► like if you have a website and you have comment section

00:37:10 ◼   ► or you have the old style web bulletin boards or whatever,

00:37:14 ◼   ► online communities and the policing of them.

00:37:17 ◼   ► It always amazes me, or Reddit is a great example,

00:37:20 ◼   ► it always amazes me how reluctant people are

00:37:24 ◼   ► in these communities to enforce their will.

00:37:29 ◼   ► So we all have opinions about what we think is good

00:37:34 ◼   ► and what's not good,

00:37:35 ◼   ► but the standard sort of non-oppressed majority opinion

00:37:40 ◼   ► of like, I'm not in a group

00:37:47 ◼   ► that is frequently the target of abuse,

00:37:51 ◼   ► or I'm at the top of the pyramid,

00:37:52 ◼   ► therefore I think everyone should be able to do everything.

00:37:55 ◼   ► So even though I think something is bad,

00:37:58 ◼   ► I'm not going to tell people in my community

00:38:00 ◼   ► that they can't do that thing

00:38:02 ◼   ► because who am I to judge, right?

00:38:04 ◼   ► And even removing the motivation, just the whole idea of like, I want my community to

00:38:09 ◼   ► be a place where everyone is free to be themselves and do whatever the heck they want.

00:38:13 ◼   ► And it almost seems to me at various times on Twitter and other places that there is

00:38:18 ◼   ► nothing so terrible that the community, the people who run the community would decide

00:38:23 ◼   ► that it's not allowed, right?

00:38:24 ◼   ► So it's like, I'm trying to think of something for the show that I could use as an example.

00:38:27 ◼   ► Like someone comes in and says, I'm in favor of eating babies.

00:38:31 ◼   ► I love to eat babies.

00:38:33 ◼   ► They come out and I just eat them instantly, right?

00:38:35 ◼   ► Perfectly healthy babies, I eat them, right?

00:38:37 ◼   ► And people are like, you know, we don't like baby eating.

00:38:40 ◼   ► Can you ban all the people who promote baby eating

00:38:44 ◼   ► in this community?

00:38:44 ◼   ► 'Cause they're really annoying

00:38:45 ◼   ► and we're all against it, right?

00:38:47 ◼   ► Like everyone here in this community of, you know,

00:38:50 ◼   ► knitters or woodworking people or, you know,

00:38:55 ◼   ► go-kart enthusiasts or whatever,

00:38:59 ◼   ► can we just say that if you promote the eating of babies,

00:39:02 ◼   ► you get banned, like no baby eaters, right?

00:39:06 ◼   ► And people around the community are like, "Well, I don't like eating babies.

00:39:10 ◼   ► And I know most people don't like it, but those people should have the right to talk

00:39:13 ◼   ► about baby eating."

00:39:14 ◼   ► It's like, "What do you mean the right?

00:39:17 ◼   ► It's your community.

00:39:18 ◼   ► It's a website that you made.

00:39:19 ◼   ► It's a product that you made.

00:39:21 ◼   ► You're not the US government.

00:39:23 ◼   ► They can talk about eating babies all they want on their front lawn.

00:39:26 ◼   ► I don't want them in this community."

00:39:28 ◼   ► And so they decide to make a community where everyone is allowed to do everything.

00:39:31 ◼   ► Because that feels like egalitarian for them and you know, it's freedom and even if they don't invoke freedom of speech

00:39:35 ◼   ► It's like this is the kind of community we want

00:39:37 ◼   ► we want a community everyone is feels free to say anything they want even is about baby eating and

00:39:41 ◼   ► There's a place for that type of thing. I think you usually call that place 4chan, right, but whatever

00:39:46 ◼   ► But it it it's

00:39:51 ◼   ► Continues to be amazing to me that you see so little of the other thing where they say, you know

00:39:56 ◼   ► What this is my community and I don't like baby eating and if you talk about baby eating here

00:40:00 ◼   ► You're banned forever and they're like that's not fair just because you don't like yeah. Yeah, just because I don't like baby eating

00:40:07 ◼   ► That's the only reason and this is my community right whether it's this guy with his game

00:40:11 ◼   ► And that's why you get a situation where this person saying hey

00:40:15 ◼   ► Killing people on your team is bad and threats of violence on Twitter are bad and you're gonna get banned for it

00:40:20 ◼   ► That's why it's a story like that's a story in the gaming news. It's it's a man bites dog story, right?

00:40:27 ◼   ► Like whoa someone someone decided like their own personal opinion and they enforce their own personal opinion on their community

00:40:33 ◼   ► Is this person the worst person ever enforcing their own personal opinion now in Twitter?

00:40:37 ◼   ► There's no personal opinion or because Twitter is a giant massive corporation, but they have the same problem that they seem

00:40:42 ◼   ► unable to choose to enforce

00:40:45 ◼   ► You know any kind of standards was like well

00:40:49 ◼   ► We don't want to enforce our standards and other people on the exam

00:40:51 ◼   ► I use baby eating because I was hoping to find something that everyone or audience would be like, yeah

00:40:54 ◼   ► against baby eating. Now the pro baby eating people go ahead and email us, whatever, but

00:40:58 ◼   ► now I'm going to say one that I would have used years ago in place of baby eating,

00:41:02 ◼   ► feeling confident that everyone who's listening to my voice would be like, "Yeah, totally baby

00:41:05 ◼   ► eating is bad." But unfortunately, I cannot think that these days. And that is the example

00:41:10 ◼   ► from Twitter where Twitter is like, "Hey, are you the neo-Nazi party? You want to promote a Nazi

00:41:17 ◼   ► agenda? You're fine on Twitter. As long as you don't threaten violence or blah, blah, blah,

00:41:22 ◼   ► or violate one of our things that we decide we're okay with Nazis like if they just want to talk peacefully about Nazi stuff

00:41:28 ◼   ► That's fine with us as long as they don't threaten violence, right?

00:41:30 ◼   ► Isn't all of their speech kind of well

00:41:33 ◼   ► Like I would just hope that Twitter as a company be like, you know what?

00:41:37 ◼   ► Even if Nazis are just like talking calmly about you know, they're they're not see agenda

00:41:43 ◼   ► No, no Nazis like it not like are you in a group? Are you in the new Nazis? You're in an organization, right?

00:41:50 ◼   ► Are you affiliated with them? Do you promote that in any way? You're banned. It's like, whoa, you can't ban them. That's freedom

00:41:56 ◼   ► It's like no no, they can have their meetings like they can talk about their Nazi things. They just can't be on Twitter

00:42:01 ◼   ► but

00:42:03 ◼   ► Twitter seemingly cannot bring itself to do that because it would be seen as like oh

00:42:08 ◼   ► Twitter don't go there because they don't let you if they don't like what they're saying they ban you right and not just like

00:42:12 ◼   ► You know things that we all agree are bad

00:42:14 ◼   ► But just like political ideas like the Nazi Party like some people are just into that man

00:42:19 ◼   ► Why can't you just like can't we just say no and Twitter and many other companies inability to draw the line

00:42:25 ◼   ► Yeah, and drop whatever Twitter wants to draw it like they I feel like you know, you lose all the Nazi people

00:42:32 ◼   ► Yeah, the Nazi people will go elsewhere. They'll have to go someplace else like well

00:42:35 ◼   ► But if I ban if I ban too many things everyone will leave and our daily active users will go

00:42:39 ◼   ► It's like you have you have to make that choice. I feel confident if I was running Twitter that if we ban all the Nazis just

00:42:46 ◼   ► on principle, like they don't have to do anything bad, they're just gone, right?

00:42:50 ◼   ► That you won't lose too many users, and the users you do lose, it's good that you're losing them,

00:42:56 ◼   ► right? And so it frustrates me as a user of Twitter and many other services to see the

00:43:02 ◼   ► unwillingness to enforce personal standards. And even if more people did this, that would mean

00:43:10 ◼   ► people are going to have weird personal standards. They're going to be like, "Sorry, nobody's name

00:43:13 ◼   ► name begins with the letter J, and then I can't join their service and I would be mad

00:43:16 ◼   ► about it, right? But it's their service. Like, they can make up their own stupid rules, right?

00:43:20 ◼   ► I'm just saying, like, maybe the baby eaters and the Nazis, maybe get rid of them. And

00:43:24 ◼   ► it's like, well, who cares? Twitter is being more successful if they let everybody in.

00:43:28 ◼   ► But things like these rumors about, oh, someone was going to buy you, but your service is

00:43:32 ◼   ► a cesspool, that, you know, your choices have a material effect on your company, not just

00:43:39 ◼   ► on the users, which you should care about, like, what's it like to use Twitter? The decisions

00:43:43 ◼   ► about what is and isn't allowed affect what it's like to use Twitter. They also affect

00:43:47 ◼   ► potentially who's willing to buy you, right? And they affect your image like there's tangible

00:43:52 ◼   ► and intangible repercussions to this. So I continue to hope that more services will act

00:43:59 ◼   ► like small services with one opinionated person, like the guy who runs the, you know, Player

00:44:05 ◼   ► unknowns battlegrounds to say I have rules and these are my rules and you

00:44:10 ◼   ► might not like the rules that's my damn game you don't like the rules go

00:44:13 ◼   ► somewhere else make your own game right you don't like the rules in Twitter

00:44:16 ◼   ► Nazis we're banning you all from Twitter go make your own service go someplace

00:44:21 ◼   ► else go do another thing because we're not the US government so anyway I'm

00:44:25 ◼   ► against baby eating I'm against Nazis I'm also against player killing in

00:44:29 ◼   ► in Battleground.

00:44:30 ◼   ► - How do you feel about people who's name is being with Jay?

00:44:34 ◼   ► - Neutral.

00:44:34 ◼   ► (laughing)

00:44:37 ◼   ► A lot of people's name is being with Jay, come on.

00:44:41 ◼   ► - We are sponsored this week by Eero.

00:44:43 ◼   ► Go to Eero.com and use code ATP for free overnight shipping.

00:44:47 ◼   ► Eero is distributed wifi made easy for you at home.

00:44:51 ◼   ► The single router model just does not work

00:44:54 ◼   ► to cover houses well enough.

00:44:55 ◼   ► We all know that you buy router covered in antennas

00:44:58 ◼   ► and it says it's gonna cover your whole house

00:45:00 ◼   ► and it still doesn't.

00:45:02 ◼   ► What you need is a distributed wifi system.

00:45:04 ◼   ► That's what offices and businesses have used for years.

00:45:07 ◼   ► With Eero, you can install an enterprise grade wifi system

00:45:10 ◼   ► in your house in just a few minutes.

00:45:12 ◼   ► All you gotta do is download the Eero app

00:45:14 ◼   ► on your iOS or Android device

00:45:16 ◼   ► and it walks you through every step of the process.

00:45:18 ◼   ► It's quick, easy, and very painless.

00:45:20 ◼   ► 'Cause current wifi routers are really, really hard

00:45:23 ◼   ► to manage and to optimize.

00:45:25 ◼   ► The Eero app does all this for you

00:45:27 ◼   ► right in the palm of your hand.

00:45:28 ◼   ► You can see how many devices are connected at any given point.

00:45:30 ◼   ► You can see your internet speed.

00:45:32 ◼   ► It tests it for you.

00:45:32 ◼   ► It does QoS for you.

00:45:34 ◼   ► You can do a guest network if you want.

00:45:35 ◼   ► It's full of advanced features, but all exposed through a

00:45:38 ◼   ► very, very easy to use app.

00:45:40 ◼   ► And it's all protected with state-of-the-art encryption.

00:45:43 ◼   ► And Eero also has incredible customer

00:45:45 ◼   ► support if you need it.

00:45:46 ◼   ► They've really invested heavily in this.

00:45:47 ◼   ► You can call, and you can get a hold of a Wi-Fi expert within

00:45:50 ◼   ► 30 seconds.

00:45:51 ◼   ► If you have any concerns about any devices during setup, one

00:45:54 ◼   ► of their experts can walk you through everything.

00:45:56 ◼   ► They can even help if you just wanna know

00:45:58 ◼   ► how many you should buy for your house,

00:45:59 ◼   ► just give 'em a call.

00:46:01 ◼   ► I've tried Eero and it's wonderful.

00:46:03 ◼   ► I could not believe how easy it was to set up.

00:46:05 ◼   ► It literally took minutes to set up

00:46:07 ◼   ► and the app is truly wonderful.

00:46:09 ◼   ► And they just released their second generation product.

00:46:11 ◼   ► So second generation Eero's now,

00:46:13 ◼   ► they have an additional five gigahertz radio,

00:46:16 ◼   ► so they have even better coverage.

00:46:18 ◼   ► It's now tri-band, it's twice as fast as the predecessor.

00:46:21 ◼   ► It includes a new thread radio

00:46:22 ◼   ► for certain low power devices

00:46:24 ◼   ► such as locks and doorbells and sensors.

00:46:26 ◼   ► And the new Eero Beacon, which is a new separate type

00:46:30 ◼   ► of just a repeater unit that they have,

00:46:32 ◼   ► is even half the size now.

00:46:35 ◼   ► So it's even smaller.

00:46:36 ◼   ► These things were already small and discreet before

00:46:38 ◼   ► and they looked beautiful pretty much anywhere in your house

00:46:41 ◼   ► and now the additional one is even smaller.

00:46:44 ◼   ► It plugs directly into any wall outlet.

00:46:46 ◼   ► It looks kinda like a nightlight.

00:46:47 ◼   ► It's really small.

00:46:48 ◼   ► It even has a nightlight built into it

00:46:50 ◼   ► 'cause they realize, well, we're here,

00:46:52 ◼   ► might as well do that too.

00:46:53 ◼   ► It's pretty great.

00:46:55 ◼   ► So check out Eero today.

00:46:56 ◼   ► Once again, if you go to Eero.com to learn more,

00:46:59 ◼   ► that's E-E-R-O.com, use promo code ATP at checkout

00:47:04 ◼   ► for free overnight shipping.

00:47:06 ◼   ► Once again, Eero.com, E-E-R-O.com,

00:47:09 ◼   ► and use promo code ATP for free overnight shipping.

00:47:12 ◼   ► Thank you very much to Eero

00:47:13 ◼   ► for sponsoring this show once again.

00:47:15 ◼   ► - Marco, you had some very interesting

00:47:24 ◼   ► overcast crashes that you had a little bit of trouble figuring out and justifiably because

00:47:30 ◼   ► having heard this story back a couple of months ago when you did discover what this fix was,

00:47:35 ◼   ► I thought to myself, my goodness, I don't know if I would have ever been able to figure

00:47:38 ◼   ► this out. But I think it's a really interesting, to the best of your recollection, story of

00:47:45 ◼   ► both what was going on, why it was going on, and how you discovered it. So can you take

00:47:49 ◼   ► us through this, maybe starting with like, how did you discover there was an issue in

00:47:53 ◼   ► in the first place.

00:47:54 ◼   ► - iTunes Connect, the backend for the App Store,

00:47:58 ◼   ► reports your number of crashes for each version of your app.

00:48:01 ◼   ► And this is as reported by people who opt in

00:48:04 ◼   ► to sharing the stats with Apple and developers

00:48:07 ◼   ► during the iPhone setup process.

00:48:09 ◼   ► I started seeing a big uptick in crashes,

00:48:11 ◼   ► and I started noticing a separate problem too,

00:48:13 ◼   ► which is my app would crash in the background a lot,

00:48:17 ◼   ► and I'd get all these crash logs for background crashes.

00:48:20 ◼   ► I didn't immediately identify it as such.

00:48:22 ◼   ► What I immediately saw was my app would not resume playback

00:48:27 ◼   ► if you hadn't been running it in a while.

00:48:29 ◼   ► So if you sent a play command,

00:48:31 ◼   ► let's say you get in your car and the Bluetooth turns on

00:48:34 ◼   ► and it sends a play command,

00:48:35 ◼   ► or you put on Bluetooth headphones

00:48:36 ◼   ► and you hit the play button on them.

00:48:37 ◼   ► I noticed that the app was not in memory anymore.

00:48:40 ◼   ► I started getting reports that were auto-generated

00:48:43 ◼   ► from Springboard that it was terminating me

00:48:45 ◼   ► in the background for holding onto a shared lock

00:48:50 ◼   ► on the database file.

00:48:52 ◼   ► So I did a number of things to try to fix this.

00:48:55 ◼   ► I read somewhere that if you open up

00:48:58 ◼   ► a background task assertion, so in,

00:49:01 ◼   ► I forget how, if the Mac does this,

00:49:04 ◼   ► but if the, on iOS you do the UI background task ID

00:49:08 ◼   ► and you say begin background task with identifier

00:49:11 ◼   ► and it basically gives you a little handle

00:49:13 ◼   ► and you do your task and then you tell it,

00:49:15 ◼   ► okay, I finished that, end this task.

00:49:17 ◼   ► And then if your app is about to be terminated

00:49:20 ◼   ► for whatever reason, and you have one of these open,

00:49:23 ◼   ► the system will give you a little bit of extra time

00:49:25 ◼   ► to go complete that.

00:49:27 ◼   ► So, somewhere I read some articles that was, you know,

00:49:31 ◼   ► that I can't find now, you know, which is probably good

00:49:33 ◼   ► to protect the not so innocent,

00:49:36 ◼   ► but somewhere I read some article that said,

00:49:38 ◼   ► in order to fix these crashes

00:49:40 ◼   ► about holding onto the lock too long,

00:49:41 ◼   ► just open up a background task

00:49:43 ◼   ► for every database query that you make.

00:49:46 ◼   ► 'Cause the way SQLite works is,

00:49:49 ◼   ► every query it temporarily locks the database file,

00:49:53 ◼   ► does what it needs to do, and then unlocks it.

00:49:55 ◼   ► And this is required because that's how you have

00:49:58 ◼   ► different processes sharing a file,

00:50:00 ◼   ► or rather sharing a database through SQLite

00:50:03 ◼   ► without some kind of server managing the background,

00:50:04 ◼   ► like MySQL would have a server, but SQLite doesn't.

00:50:07 ◼   ► So anyway, it needs to do file locking to do this stuff.

00:50:10 ◼   ► So I got this tip from some random blog post that,

00:50:14 ◼   ► oh, wrap every database query in a background task assertion

00:50:19 ◼   ► and then you won't have this problem anymore.

00:50:21 ◼   ► That was technically true,

00:50:26 ◼   ► in the sense that-- - Oh, here we go.

00:50:28 ◼   ► - When I shipped the version that included this fix,

00:50:32 ◼   ► it was technically true that I was no longer

00:50:34 ◼   ► getting killed with that background code.

00:50:36 ◼   ► However, what was happening is I noticed

00:50:40 ◼   ► that my crash rate, rather than going down,

00:50:44 ◼   ► went way, way up.

00:50:47 ◼   ► It took me a while to figure out what was going on.

00:50:49 ◼   ► And during the time that I was trying to figure out

00:50:52 ◼   ► what was going on, I also noticed that as I would be

00:50:54 ◼   ► listening to Overcast, like in the car on long car trips,

00:50:57 ◼   ► sometimes my entire iPhone would crash.

00:51:01 ◼   ► And it would do what nerds might know as a re-spring,

00:51:04 ◼   ► which is the phone does not reboot,

00:51:06 ◼   ► but it looks like it's rebooting.

00:51:07 ◼   ► And it just comes back up faster.

00:51:09 ◼   ► And you see the little spinner, all your apps quit,

00:51:13 ◼   ► and then eventually you get back to your lock screen.

00:51:16 ◼   ► And the way you can tell that it wasn't a real reboot

00:51:18 ◼   ► is Touch ID is still on.

00:51:20 ◼   ► 'Cause when you reboot for real,

00:51:22 ◼   ► you have to enter your passcode the first time.

00:51:24 ◼   ► But if it does a little spinner and back to the lock screen

00:51:26 ◼   ► and all your apps are gone and Touch ID works,

00:51:28 ◼   ► that was a springboard crash, not a full reboot.

00:51:31 ◼   ► And I noticed my phone was doing this more and more.

00:51:34 ◼   ► And it was very inconvenient at times such as

00:51:36 ◼   ► when I was driving and following directions in ways.

00:51:40 ◼   ► (laughs)

00:51:41 ◼   ► And my directions were just gone for a little while

00:51:43 ◼   ► while all the apps got killed

00:51:45 ◼   ► and I had to restart everything.

00:51:47 ◼   ► But anyway, and I started getting reports from people

00:51:49 ◼   ► saying my phone keeps crashing with Overcast,

00:51:51 ◼   ► but I can't figure out why.

00:51:54 ◼   ► And it took until one guy sent me a video

00:51:58 ◼   ► that he showed something amazingly useful to me.

00:52:01 ◼   ► He said, "Look, Overcast, if you pause it in the background

00:52:06 ◼   ► "and you bring up control center,

00:52:07 ◼   ► "so you can see that Overcast is still

00:52:09 ◼   ► "in the now playing thing in control center,

00:52:12 ◼   ► "after 10 minutes exactly,

00:52:14 ◼   ► or no, I think it was like three minutes.

00:52:16 ◼   ► Anyway, after some short duration of minutes,

00:52:19 ◼   ► it disappears, reliably, every time.

00:52:22 ◼   ► So that led me to finally catch the inner debugger

00:52:25 ◼   ► and to see that I was being killed

00:52:27 ◼   ► with a certain crash with a certain code.

00:52:29 ◼   ► And it still was not enough to actually figure it out

00:52:31 ◼   ► because the code I was getting killed with was,

00:52:34 ◼   ► you've exceeded some limit.

00:52:36 ◼   ► (laughs)

00:52:37 ◼   ► It's like, okay, well, helpful, maybe, thanks,

00:52:41 ◼   ► but I have no idea still.

00:52:44 ◼   ► And I tried a few things, trying to figure out

00:52:47 ◼   ► what limit am I exceeding here, what does this code mean?

00:52:50 ◼   ► Eventually I realized that this would also

00:52:53 ◼   ► sometimes correspond with one of those springboard crashes,

00:52:57 ◼   ► with the whole phone crash.

00:52:58 ◼   ► And so I thought, I wonder if this is related.

00:53:01 ◼   ► I direct messaged somebody who I know

00:53:05 ◼   ► on the springboard team.

00:53:06 ◼   ► I basically said, look, I'm at my wits end here,

00:53:08 ◼   ► I can't figure this out.

00:53:10 ◼   ► is there anything that would cause this crash code

00:53:14 ◼   ► to be sent to my app that might sometimes

00:53:17 ◼   ► crash Springboard?

00:53:18 ◼   ► Little while later, I hear, oh, there's actually

00:53:22 ◼   ► a bug report internal to Apple talking about

00:53:26 ◼   ► how Overcatch is causing this problem.

00:53:28 ◼   ► (laughing)

00:53:31 ◼   ► I was like, nice.

00:53:31 ◼   ► Oh?

00:53:32 ◼   ► - Do you think they ever would've told you?

00:53:35 ◼   ► They're just like--

00:53:36 ◼   ► - I don't, so there is actually,

00:53:38 ◼   ► There is a process that happens sometimes inside of Apple

00:53:43 ◼   ► where if you have an app that's causing lots of problems

00:53:46 ◼   ► for Apple, there is sometimes you will hear

00:53:51 ◼   ► from one of the developer evangelists about it.

00:53:54 ◼   ► This has happened to me maybe once or twice ever,

00:53:57 ◼   ► so this isn't a thing that happens often.

00:53:59 ◼   ► But this was happening in the lead up to WWDC.

00:54:03 ◼   ► This is only a couple weeks before WWDC, I think,

00:54:05 ◼   ► if I remember correctly.

00:54:06 ◼   ► And so the problem with that is that this is a time

00:54:09 ◼   ► when the entire, like all of Apple is basically

00:54:11 ◼   ► freaking out, rushing to finish the WBC betas.

00:54:16 ◼   ► 'Cause that's like the first time the OS

00:54:19 ◼   ► is really gonna see any kind of wide attention.

00:54:21 ◼   ► It's gonna be installed on a bunch of developer devices

00:54:23 ◼   ► and a bunch of enthusiasts who say they're developers

00:54:25 ◼   ► who aren't and just use the developer betas anyway.

00:54:28 ◼   ► So there's a massive rush right before WBC

00:54:31 ◼   ► for anybody who works in anything near these new OSs

00:54:35 ◼   ► to just get it done, to get that first beta ready

00:54:38 ◼   ► for the public.

00:54:40 ◼   ► And so this was happening during that big crunch time,

00:54:43 ◼   ► so I imagine it slipped through the cracks.

00:54:45 ◼   ► You know, I'm sure it was on somebody's to-do list

00:54:47 ◼   ► to maybe see if they could reach out to me or something,

00:54:49 ◼   ► and they didn't.

00:54:50 ◼   ► Because you can imagine Apple is not so keen

00:54:52 ◼   ► on an app that can crash Springboard.

00:54:54 ◼   ► Like, that's not something that apps

00:54:56 ◼   ► are supposed to be able to do.

00:54:57 ◼   ► So basically, so there was this internal thing

00:55:02 ◼   ► that I just happened to, like, I happened to ask somebody

00:55:05 ◼   ► who started looking around who happened to find this.

00:55:07 ◼   ► And then basically the notes that Apple Engineering

00:55:12 ◼   ► had noticed about it was that I was taking a large number

00:55:17 ◼   ► of background assertions in the app

00:55:20 ◼   ► which was apparently causing Springboard,

00:55:23 ◼   ► Springboard has like a watchdog that goes along

00:55:25 ◼   ► and checks various limits every few minutes.

00:55:27 ◼   ► And that watchdog when it looked at how many

00:55:30 ◼   ► background assertions I had because I was doing one

00:55:33 ◼   ► for every database query.

00:55:35 ◼   ► (laughs)

00:55:36 ◼   ► So like, you know, a typical load of the app

00:55:39 ◼   ► is gonna make probably a few hundred database queries

00:55:41 ◼   ► over the course of navigating the first few screens.

00:55:43 ◼   ► So, you know, we're talking over the course

00:55:47 ◼   ► of typical usage, I might make a few hundred

00:55:49 ◼   ► to a few thousand database queries

00:55:51 ◼   ► over like a well-used session of the app.

00:55:54 ◼   ► The background task API was designed

00:55:57 ◼   ► back in the early days of iOS

00:56:00 ◼   ► for you to take out one background task assertion

00:56:04 ◼   ► if you're finishing uploading a file to a web service

00:56:08 ◼   ► or doing some kind of sync operation.

00:56:11 ◼   ► Like, okay, take a background assertion,

00:56:12 ◼   ► finish the sync operation, and then give it back.

00:56:14 ◼   ► And so apparently this was designed for something more

00:56:19 ◼   ► on the order of hundreds of these to happen

00:56:23 ◼   ► during a process, not thousands of these to happen

00:56:26 ◼   ► the way I was doing it with every database query.

00:56:30 ◼   ► And this was causing two problems.

00:56:31 ◼   ► Number one, it would cause this watchdog

00:56:34 ◼   ► that goes around to check to see

00:56:37 ◼   ► if you've exceeded your limits,

00:56:38 ◼   ► it would cause it to kill my app to say,

00:56:41 ◼   ► hey, you've exceeded your limits.

00:56:42 ◼   ► But apparently there was also a bug in that watchdog

00:56:46 ◼   ► in older versions of iOS before 11

00:56:48 ◼   ► that sometimes if you had such a ridiculous number

00:56:53 ◼   ► of them as I did, that watchdog would crash.

00:56:56 ◼   ► And because it's a part of the system,

00:56:58 ◼   ► If that crashes, basically all of Springboard comes down

00:57:01 ◼   ► and is brought back up.

00:57:03 ◼   ► This has been fixed in iOS 11,

00:57:05 ◼   ► but it was not fixed for at least the earlier parts of 10.

00:57:08 ◼   ► I don't know if the very last version of 10

00:57:10 ◼   ► might have fixed it, I don't know, honestly.

00:57:12 ◼   ► So basically, Overcast was not only exposing

00:57:15 ◼   ► its own limit crossing behavior,

00:57:17 ◼   ► and by the way, I don't know what that limit is.

00:57:19 ◼   ► Nobody will tell me what that limit is.

00:57:21 ◼   ► It doesn't really matter for my purposes.

00:57:23 ◼   ► All that matters is background tasks

00:57:25 ◼   ► for something to be used not in the thousands.

00:57:28 ◼   ► (laughs)

00:57:29 ◼   ► But basically, I was causing Springboard to crash.

00:57:33 ◼   ► And I greatly, greatly appreciate both the user

00:57:38 ◼   ► who sent me the video who showed that it would be terminated

00:57:41 ◼   ► after a certain number of minutes,

00:57:42 ◼   ► 'cause that allowed me to get the crash code

00:57:44 ◼   ► that was reliably causing this problem.

00:57:46 ◼   ► And I also greatly appreciate anybody inside Apple

00:57:49 ◼   ► who helped me track this down.

00:57:52 ◼   ► And I really don't appreciate the blog post

00:57:54 ◼   ► I read back forever ago and I'm glad I can't find it now.

00:57:57 ◼   ► So I'm telling you all now, and you developers listening,

00:58:00 ◼   ► don't use background task IDs for frequently occurring

00:58:03 ◼   ► things like database queries.

00:58:05 ◼   ► That is too granular.

00:58:07 ◼   ► And the solution to the problem

00:58:09 ◼   ► ended up being something totally different.

00:58:11 ◼   ► We are sponsored this week by Betterment,

00:58:14 ◼   ► investing made better.

00:58:15 ◼   ► Go to betterment.com/ATP to learn more.

00:58:18 ◼   ► Betterment is a smarter way to invest your money,

00:58:21 ◼   ► providing investing advice through smart technology,

00:58:23 ◼   ► automated investing and human advisors.

00:58:26 ◼   ► They have changed the industry by using the same strategies

00:58:28 ◼   ► that financial advisors use with clients

00:58:30 ◼   ► who have millions of dollars,

00:58:31 ◼   ► but now making those available to everyone.

00:58:34 ◼   ► This includes advanced strategies such as smart rebalancing,

00:58:37 ◼   ► global diversification, tax harvesting, and so much more.

00:58:41 ◼   ► Betterment's mission is to help you manage

00:58:43 ◼   ► and grow your investments to build your financial future

00:58:46 ◼   ► and all with low fees,

00:58:47 ◼   ► so you lose less of your money along the way.

00:58:50 ◼   ► 'Cause if you do the math, typical investment fees and costs

00:58:53 ◼   ► really add up over time and Betterment's fees are a fraction of the cost of other financial

00:58:57 ◼   ► services. And Betterment's so easy to use that they've won awards for their customer

00:59:02 ◼   ► experience. You can log into a demo account right on their website if you're curious to

00:59:05 ◼   ► see it. And ATP listeners can get up to one year managed for free. For more information,

00:59:11 ◼   ► visit betterment.com/ATP. Terms and conditions apply and investing involves risk. To learn

00:59:17 ◼   ► more visit betterment.com/ATP. That's betterment.com/ATP. Betterment, investing made better. So, going

00:59:30 ◼   ► back to the original problem. The problem was I was holding on to a shared file lock

00:59:37 ◼   ► in the shared app container because the database file, so quick, very quick overview, I apologize

00:59:43 ◼   ► to non-programmers, this is probably very boring. I've been trying to not talk about

00:59:46 ◼   ► for weeks and John's finally making me this week.

00:59:48 ◼   ► But anyway, so--

00:59:49 ◼   ► - And not just him.

00:59:50 ◼   ► - Okay, you too.

00:59:52 ◼   ► I voted.

00:59:53 ◼   ► So anyway, long explanation, slightly less long again.

00:59:57 ◼   ► iOS apps are totally isolated from each other.

01:00:00 ◼   ► They, you know, for security reasons,

01:00:01 ◼   ► for control reasons, for user-friendliness,

01:00:03 ◼   ► there's lots of good reasons.

01:00:04 ◼   ► iOS apps, if your app wants to read or write a file

01:00:09 ◼   ► for its own data purposes,

01:00:11 ◼   ► it writes it in a special sandbox folder

01:00:12 ◼   ► that it can't read or write anything outside of that

01:00:15 ◼   ► and no other apps can read its files.

01:00:18 ◼   ► And this is extremely enforced such that

01:00:21 ◼   ► even if you have multiple apps from the same developer,

01:00:24 ◼   ► they are isolated from each other,

01:00:26 ◼   ► even if you have extensions of your app,

01:00:29 ◼   ► because an extension is technically running

01:00:30 ◼   ► as a separate process on the system,

01:00:33 ◼   ► the extension can't read and write the same data

01:00:35 ◼   ► as its parent app even.

01:00:37 ◼   ► Like that's how walled off everything is in iOS

01:00:39 ◼   ► for various good security reasons.

01:00:42 ◼   ► Which by the way, I love and I kinda wish

01:00:44 ◼   ► that the Mac had more of that.

01:00:46 ◼   ► Anyway, Apple created a system called

01:00:48 ◼   ► Shared App Group Containers to try to solve the problem of,

01:00:52 ◼   ► first of all, if you have multiple apps

01:00:55 ◼   ► from the same developer, can they share any information?

01:00:57 ◼   ► And then second of all, if you just have an app

01:00:59 ◼   ► with an extension, it's pretty hard to make

01:01:02 ◼   ► a useful extension that can't read or write

01:01:05 ◼   ► any of its parent app's data.

01:01:07 ◼   ► The shared app container is basically a,

01:01:10 ◼   ► it's literally what it sounds like.

01:01:10 ◼   ► It's a shared container that apps and extensions

01:01:14 ◼   ► can all read and write from, you know,

01:01:16 ◼   ► separately, you know, separately from their main data space

01:01:19 ◼   ► so they can share data.

01:01:21 ◼   ► I had the bright idea back when the,

01:01:23 ◼   ► when watchOS first came out,

01:01:25 ◼   ► and I had to make my watch extension,

01:01:26 ◼   ► which was Overcast's first extension,

01:01:28 ◼   ► what if I just move the database file

01:01:31 ◼   ► into the shared app container,

01:01:32 ◼   ► and then the extension can read and write

01:01:34 ◼   ► the same database as the app?

01:01:35 ◼   ► And I can copy the same model files over

01:01:37 ◼   ► and have the extension have a lot of code sharing

01:01:40 ◼   ► with the main app.

01:01:41 ◼   ► The thing with iOS is that processes can be randomly

01:01:45 ◼   ► suspended and terminated without their consent or whatever

01:01:51 ◼   ► as the system deems them no longer necessary

01:01:53 ◼   ► if the user quits, see whatever else.

01:01:55 ◼   ► And this is especially enforced on extensions.

01:01:58 ◼   ► If you open up, say, a widget,

01:02:00 ◼   ► like the today, it used to be called today widget now,

01:02:03 ◼   ► and I think they're just called widgets now, right?

01:02:05 ◼   ► - I don't even remember.

01:02:06 ◼   ► - Whatever they're called in this iOS release.

01:02:09 ◼   ► So widgets have very, very low limits

01:02:11 ◼   ► because you might have a screen

01:02:13 ◼   ► that has like five of them on it.

01:02:14 ◼   ► Like if you open up your whatever notification center

01:02:16 ◼   ► is called now and you have the part of the notification center

01:02:18 ◼   ► that used to be called today's screen,

01:02:19 ◼   ► who knows what that's called now.

01:02:21 ◼   ► Anyway, and so you have all these things

01:02:22 ◼   ► that might be called widgets on that screen.

01:02:25 ◼   ► (laughs)

01:02:26 ◼   ► And so each one of those is running a separate process

01:02:29 ◼   ► and so they have very, very aggressive limits on those.

01:02:32 ◼   ► In the context of a widget

01:02:34 ◼   ► or actually any extension technically,

01:02:36 ◼   ► there is no such thing as a background task assertion

01:02:39 ◼   ► in the typical sense where you say,

01:02:40 ◼   ► begin background task and I'll tell you when I'm done.

01:02:43 ◼   ► They have this other thing, this NSProcessInfo API

01:02:45 ◼   ► that is horrendous and really hard to use.

01:02:48 ◼   ► But basically, they very aggressively terminate

01:02:52 ◼   ► the processes of those extensions

01:02:54 ◼   ► whenever you're done with them.

01:02:56 ◼   ► So the problem is if you have a shared app group container

01:03:00 ◼   ► and you have the ability for one of these processes

01:03:03 ◼   ► to take a lock on a file in that.

01:03:06 ◼   ► You have a problem then, what do you do

01:03:09 ◼   ► if say the widget has a lock on the database file,

01:03:12 ◼   ► which is the single copy that's shared

01:03:14 ◼   ► between the parent app and everything else.

01:03:16 ◼   ► If the widget takes a lock on that file

01:03:18 ◼   ► and then gets terminated before the lock is released,

01:03:21 ◼   ► what should happen?

01:03:22 ◼   ► Should the lock not be released?

01:03:25 ◼   ► Then the app can't run because it's deadlocked.

01:03:27 ◼   ► It's waiting for that lock to be released

01:03:28 ◼   ► from the extension that's not running right now.

01:03:31 ◼   ► Or should you just kind of force the lock to be released?

01:03:34 ◼   ► Well then you have the risk of data corruption

01:03:36 ◼   ► because then if the widget then relaunches,

01:03:39 ◼   ► the widget thinks the file has been locked

01:03:41 ◼   ► that whole time in the meantime, but it actually hasn't,

01:03:44 ◼   ► so the lock is kind of meaningless

01:03:45 ◼   ► and so you either have lots of bugs

01:03:48 ◼   ► or you have data corruption or whatever else.

01:03:50 ◼   ► So Apple, realizing this would be a problem,

01:03:53 ◼   ► built in this wonderful system of this springboard,

01:03:57 ◼   ► the special springboard crasher that you will get a log

01:04:00 ◼   ► if your app is terminated while it's holding a lock

01:04:04 ◼   ► in the shared container, you get these crazy crash logs

01:04:07 ◼   ► that were the kind I was getting to begin with.

01:04:09 ◼   ► This was the problem I was trying to solve at the beginning

01:04:11 ◼   ► and the reason I did this whole background task thing

01:04:13 ◼   ► that was horrendous.

01:04:14 ◼   ► Finally, the end of the story.

01:04:17 ◼   ► I had that problem for that reason

01:04:19 ◼   ► and I tried to solve it with this background task

01:04:22 ◼   ► on every database query thing, which was a terrible idea

01:04:25 ◼   ► that causes other springboard crashes,

01:04:26 ◼   ► way, way, way worse crash,

01:04:29 ◼   ► And so instead, I decided, you know what?

01:04:32 ◼   ► I'm just not gonna have the database

01:04:33 ◼   ► in the shared container anymore.

01:04:35 ◼   ► I just moved it back into the ask private storage.

01:04:38 ◼   ► And instead of the widget communicating with,

01:04:42 ◼   ► and the watch engine, instead of them communicating

01:04:44 ◼   ► with the parent app by reading

01:04:46 ◼   ► and writing the database directly,

01:04:48 ◼   ► now they just have their own little mini APIs

01:04:50 ◼   ► to the parent app where they are communicating

01:04:52 ◼   ► via dictionaries that are just being written

01:04:55 ◼   ► to little files in the shared container

01:04:57 ◼   ► that are not locked or anything like that.

01:04:59 ◼   ► So I took this elegant but complex solution

01:05:02 ◼   ► of having the database be shared

01:05:04 ◼   ► and made a much, much, much simpler lower tech solution

01:05:07 ◼   ► of just passing around serialized dictionaries

01:05:09 ◼   ► as unlocked files basically.

01:05:12 ◼   ► And now all these problems are gone, it is totally solved.

01:05:16 ◼   ► My crash rate has dropped to barely anything,

01:05:19 ◼   ► relatively speaking, and it is finally done.

01:05:22 ◼   ► And my users were very patient during this time

01:05:24 ◼   ► and I greatly appreciate that,

01:05:25 ◼   ► along with the guy who reported the initial bug

01:05:28 ◼   ► and the Apple people who helped me find it.

01:05:30 ◼   ► So that is the story.

01:05:31 ◼   ► I think this was now,

01:05:33 ◼   ► because this was before the WBC barrier

01:05:37 ◼   ► in the year of Apple scheduling,

01:05:38 ◼   ► I have very little recollection of most of these details,

01:05:41 ◼   ► so I apologize if I got anything wrong.

01:05:43 ◼   ► I apologize for the long-winded nature of this,

01:05:44 ◼   ► but Casey and John made me do it.

01:05:46 ◼   ► - So I have a couple of questions about this.

01:05:49 ◼   ► - Yeah.

01:05:49 ◼   ► - How many database queries were you making?

01:05:53 ◼   ► I never really measured what exactly the threshold

01:05:57 ◼   ► of being a problem was, but I think it's in the order of,

01:06:02 ◼   ► it's in the magnitude of hundreds over a short time

01:06:07 ◼   ► and then maybe approaching thousands

01:06:11 ◼   ► if you use the app for a while.

01:06:13 ◼   ► And it's doing a lot of, 'cause the app syncs back

01:06:15 ◼   ► to the server every few minutes.

01:06:18 ◼   ► As you're listening, every few seconds,

01:06:21 ◼   ► it's writing your progress back to the database.

01:06:23 ◼   ► That way, in case it crashes or your phone crashed

01:06:25 ◼   ► or whatever else, you don't lose too much progress.

01:06:27 ◼   ► So every few seconds, I'm running to disk at least.

01:06:29 ◼   ► And then when you're doing a sync,

01:06:31 ◼   ► there's a lot more queries that go on there.

01:06:33 ◼   ► And if you browse a list, it's loading all that data

01:06:36 ◼   ► as you page through that list.

01:06:37 ◼   ► As you scroll, it's loading all those table cells

01:06:39 ◼   ► for all the episodes, it's loading all that stuff.

01:06:41 ◼   ► So I do a good amount of queries.

01:06:43 ◼   ► - That was my question.

01:06:44 ◼   ► So when you're scrolling through a list,

01:06:45 ◼   ► it's like loading them on the fly at that point

01:06:48 ◼   ► for the data source of whatever it's called,

01:06:49 ◼   ► the data provider for the table view or whatever?

01:06:52 ◼   ► - Yep.

01:06:53 ◼   ► It's not like pulling the whole list and keeping it in memory and then just doling it out to

01:06:58 ◼   ► the GUI data source provider thingy.

01:07:01 ◼   ► You don't just do one big giant query up front to get the whole playlist and then just keep

01:07:05 ◼   ► it in memory and chuck it out as the table cells ask for it.

01:07:08 ◼   ► Well, I do.

01:07:10 ◼   ► And I have some caching built into the database layer as well.

01:07:14 ◼   ► So for instance, I do load the whole list of episodes for a playlist view.

01:07:19 ◼   ► but then as you scroll, I need to know the artwork.

01:07:23 ◼   ► And that comes from the podcast record,

01:07:26 ◼   ► not the episode record.

01:07:27 ◼   ► That's one level up.

01:07:28 ◼   ► And so I have podcasts cached,

01:07:30 ◼   ► I have that in a nice instance cache and everything,

01:07:35 ◼   ► but keep in mind a lot of this stuff falls over

01:07:38 ◼   ► if people have a lot of podcasts.

01:07:40 ◼   ► Basically, if it exceeds what my app reasonably expects

01:07:43 ◼   ► to cache or what will reasonably fit in memory to cache,

01:07:47 ◼   ► there's a lot more queries happening as this stuff is paged in. Also, if things change,

01:07:53 ◼   ► for instance, if I get a change value from the server, it invalidates a lot of these

01:07:59 ◼   ► values that are cached, so I have to then reload those on their next request. If the

01:08:04 ◼   ► list of playable content changes, which basically is like if a new episode comes in or if an

01:08:12 ◼   ► an episode gets deleted, or if some criteria of an episode

01:08:16 ◼   ► changes whether it is a member of a playlist or not,

01:08:19 ◼   ► then that is called, in my app I call that

01:08:22 ◼   ► the playable content change, which is basically

01:08:25 ◼   ► some part of what belongs in the list of episodes somewhere

01:08:29 ◼   ► is no longer valid, and when that happens,

01:08:32 ◼   ► lots of things get invalidated.

01:08:34 ◼   ► And so there's a lot going on.

01:08:35 ◼   ► I mean, a podcast app is a lot more complex

01:08:39 ◼   ► than most people would assume.

01:08:41 ◼   ► It's certainly more complex than I assumed when I started writing one, as I found out.

01:08:47 ◼   ► And they're only getting more complex over time as the market matures and as the scope

01:08:52 ◼   ► of these things, the scope of what people expect, expands.

01:08:55 ◼   ► It'll be interesting to see, you know, like a, I guess you'd probably get it in just

01:09:01 ◼   ► looking at the console log or using instruments or something like the flow of database queries.

01:09:06 ◼   ► Because I, you know, using Overcast, I would think, well, you know, this does some database

01:09:09 ◼   ► queries, but surely I would never guess the size of the number of database queries like over the

01:09:14 ◼   ► course of a 10-minute period that it would literally do thousands, like especially just

01:09:18 ◼   ► during normal use, even accounting for the "hey I'm syncing your play position every few seconds" so

01:09:22 ◼   ► I found that pretty surprising in this whole thing. And also I would say that whatever that

01:09:26 ◼   ► blog post was that you found that was like "just do everything as a background task"

01:09:31 ◼   ► I can't really blame the blog post because any solution like that is like the underlying

01:09:36 ◼   ► assumption is you're not doing hundreds or thousands of database queries. Like you've

01:09:40 ◼   ► got a couple database queries and you do them and you just put them in the background you'll be fine

01:09:44 ◼   ► and you would have been fine if the number was like 20. You know, like a couple a minute, like

01:09:49 ◼   ► five, five database queries every minute. But if it's hundreds or thousands then obviously that's

01:09:54 ◼   ► just not going to work. My question, another question about your not having the database in

01:09:58 ◼   ► the shared container and just having the thing right, pls to the shared container, does that mean

01:10:02 ◼   ► that the main Overcast app has to be running

01:10:06 ◼   ► to notice those things appearing in the shared container

01:10:08 ◼   ► and take action on them because it owns the database now?

01:10:11 ◼   ► Like, how does that work?

01:10:13 ◼   ► - It would mean that, but for a variety of reasons,

01:10:16 ◼   ► it either always is running or becomes running.

01:10:19 ◼   ► So let me go through.

01:10:21 ◼   ► Basically, there's only two extensions

01:10:23 ◼   ► that matter for this purpose.

01:10:25 ◼   ► One of them is the widget, the other one is the watch app.

01:10:28 ◼   ► When watchOS, watchkit one, back for watchOS one,

01:10:33 ◼   ► the extension for the watch was actually a process

01:10:37 ◼   ► running on the phone.

01:10:39 ◼   ► And so it had access to the shared data container

01:10:41 ◼   ► on the phone.

01:10:42 ◼   ► So it was actually surprisingly easy to make watchkit one

01:10:45 ◼   ► apps because you could just share the database,

01:10:48 ◼   ► share all the model files and everything,

01:10:50 ◼   ► and then just have the watchkit read and write

01:10:51 ◼   ► the database directly.

01:10:52 ◼   ► And it kind of took care of that for you.

01:10:54 ◼   ► WatchKit 2 and forward, so watchOS 2 and 3 and also now 4,

01:10:59 ◼   ► moved the extension to run on the watch itself

01:11:03 ◼   ► instead of on the phone.

01:11:04 ◼   ► Because running it on the phone made the app really slow

01:11:07 ◼   ► basically because it had to send all the interface stuff

01:11:08 ◼   ► over Bluetooth and it sucked.

01:11:10 ◼   ► That's one of the reasons why watchOS 1 was so bad for apps

01:11:13 ◼   ► and why the newer watchOS has since then

01:11:15 ◼   ► have been so much better.

01:11:17 ◼   ► Anyway, so when the extension execution moved to the watch,

01:11:23 ◼   ► That meant it could no longer read and write

01:11:24 ◼   ► the container anymore, 'cause the container's on the phone.

01:11:26 ◼   ► So it had its own container on the watch now,

01:11:29 ◼   ► but anyway, it couldn't read and write the database.

01:11:31 ◼   ► So that eliminated the whole reason

01:11:33 ◼   ► I had it there in the first place.

01:11:35 ◼   ► The only reason this came up is that

01:11:36 ◼   ► when I made Overcast 3.0, I made a widget.

01:11:38 ◼   ► And the widget was like the first time

01:11:41 ◼   ► I was really using, or really taking advantage

01:11:43 ◼   ► of the fact that the database was somewhere shared

01:11:45 ◼   ► to begin with.

01:11:46 ◼   ► The watch extension already needed to have

01:11:50 ◼   ► ways to communicate with the app,

01:11:52 ◼   ► to wake it up if need be and the watch connectivity

01:11:56 ◼   ► framework allows for all that.

01:11:57 ◼   ► There's different commands and stuff.

01:11:59 ◼   ► There are ways for the, it basically is already

01:12:02 ◼   ► communicating the way that I said I moved everything to.

01:12:05 ◼   ► It was really only the widget that was causing the problems

01:12:09 ◼   ► because I'd already made this move with the watch app.

01:12:11 ◼   ► Anyway, so the widget, either the phone app

01:12:15 ◼   ► is already running, in which case it notices anyway,

01:12:17 ◼   ► or basically I have a mechanism in the widget,

01:12:21 ◼   ► 'cause the widget has a play button,

01:12:22 ◼   ► and if you tap into the artwork in it,

01:12:24 ◼   ► it'll play that one also.

01:12:26 ◼   ► So the widget needs to have some way to wake up the app.

01:12:29 ◼   ► If it was just one way data transfer,

01:12:31 ◼   ► if the widget was just showing status from the app,

01:12:34 ◼   ► like you might have,

01:12:35 ◼   ► if you made a weather app, for instance,

01:12:37 ◼   ► there'd be no reason for the widget

01:12:40 ◼   ► to communicate back to the parent app

01:12:41 ◼   ► and something like that.

01:12:43 ◼   ► You might just have the widget launch the app,

01:12:45 ◼   ► but you can do that already.

01:12:47 ◼   ► In my app, I had to have some way for the widget

01:12:49 ◼   ► to communicate back to the app,

01:12:51 ◼   ► and the app might not be running.

01:12:53 ◼   ► So what I did for that was the widget basically

01:12:57 ◼   ► sends a command through one of these serialized playlets.

01:13:01 ◼   ► It puts a file in the shared app container that says,

01:13:03 ◼   ► "Hey, do this command."

01:13:05 ◼   ► And if it doesn't get a response within some very short time,

01:13:08 ◼   ► I think it's like a half a second or something,

01:13:10 ◼   ► then I launch a URL scheme that launches the app

01:13:13 ◼   ► in the foreground, and you'll notice that if you use it.

01:13:15 ◼   ► If basically the app was already running,

01:13:17 ◼   ► if you hit that play button in the widget,

01:13:20 ◼   ► it'll start again in the background

01:13:21 ◼   ► 'cause the app was still running.

01:13:23 ◼   ► So basically, this is all a giant pile of hacks.

01:13:26 ◼   ► (laughs)

01:13:28 ◼   ► The short version is the widget tries to send the app

01:13:31 ◼   ► a command and if it doesn't get a response,

01:13:32 ◼   ► it knows it isn't running so it launches it

01:13:34 ◼   ► in the foreground.

01:13:35 ◼   ► - My final question, because every overcast segment

01:13:39 ◼   ► would not be completed without some kind of bug report

01:13:41 ◼   ► slash feature request is--

01:13:43 ◼   ► - Oh my word.

01:13:43 ◼   ► - I'm still--

01:13:44 ◼   ► - This better not be about offline watch playback.

01:13:46 ◼   ► I swear I'm gonna set that feature on fire.

01:13:48 ◼   ► - No, it is not.

01:13:49 ◼   ► I know about your woes with that and I did try it myself and I had my own woes because

01:13:53 ◼   ► the volume is too low and I couldn't hear anything but you know about that, right?

01:13:55 ◼   ► It's getting worse.

01:13:56 ◼   ► It's getting so much worse.

01:13:58 ◼   ► My question is about the mysterious world of the internal state of something on my phone

01:14:05 ◼   ► that decides what it is that will start playing when my phone connects to Bluetooth on my

01:14:10 ◼   ► car.

01:14:11 ◼   ► So very often I will enter my car and I just did it today actually.

01:14:17 ◼   ► Before I entered the car, I unlocked my phone and looked and either I used the multitasking

01:14:24 ◼   ► switcher to go to it or I tapped the overcast icon or it was already in it, but the point

01:14:27 ◼   ► is overcast was the frontmost application.

01:14:30 ◼   ► I had previously been listening to a podcast.

01:14:32 ◼   ► I was not listening now.

01:14:33 ◼   ► I think I double tapped my AirPods and put them away or whatever.

01:14:36 ◼   ► But overcast is the frontmost app.

01:14:38 ◼   ► So I hit the sleep button, I go into my car, it connects to Bluetooth, and instead of overcast

01:14:42 ◼   ► continuing to play the thing I was just listening to, music starts to play.

01:14:46 ◼   ► I have no idea how iOS determines what it should start playing when I get into my car

01:14:54 ◼   ► and it connects to Bluetooth, but maybe 15% of the time it starts playing music instead

01:15:01 ◼   ► of playing...

01:15:02 ◼   ► I mean, sometimes Overcast is long gone.

01:15:04 ◼   ► It kind of makes sense.

01:15:05 ◼   ► Like, the last time I listened to Overcast was last night, and I've since used Twitter

01:15:08 ◼   ► and browsed the web and done all sorts of stuff.

01:15:10 ◼   ► Surely Overcast is not running anymore because I've just thrashed through the memory.

01:15:14 ◼   ► So when I get into my car, maybe it's going to be like, "Oh, I should start playing a

01:15:18 ◼   ► random playlist," right?

01:15:19 ◼   ► But when Overcast is the front-most app and I was just listening to it and it starts playing

01:15:22 ◼   ► music, it makes me realize I have no idea what the hell's going on inside there.

01:15:25 ◼   ► I'm assuming you have very little control over this, as with so many things that seem

01:15:29 ◼   ► like OS-level things, but can you shed some light on what the hell's going on there?

01:15:33 ◼   ► Sure.

01:15:34 ◼   ► I mean, I can tell you right now I have zero control over it, but I will tell you roughly

01:15:36 ◼   ► what I know is happening.

01:15:39 ◼   ► And a lot of this is conjecture.

01:15:40 ◼   ► A lot of this is just figuring it out

01:15:43 ◼   ► over time with experience, and some of this,

01:15:46 ◼   ► some, very little of this has been documented before,

01:15:49 ◼   ► but not most of it.

01:15:50 ◼   ► So the short version is it tries to do whatever you did last

01:15:55 ◼   ► but there are some exceptions to when this will happen.

01:15:58 ◼   ► So first of all, a lot of times, especially recently

01:16:02 ◼   ► with the rise of video clips on Twitter and Facebook,

01:16:06 ◼   ► what you did last might not be what you think you did last

01:16:10 ◼   ► or you might have forgotten, or it might have been something.

01:16:11 ◼   ► So for example, if you are, you know,

01:16:14 ◼   ► if Overcast is paused and you're looking at Twitter

01:16:17 ◼   ► and you view some kind of embedded video clip,

01:16:21 ◼   ► Twitter has just become the most recently audio playing app.

01:16:24 ◼   ► And iOS keeps track of whatever the most recently

01:16:26 ◼   ► audio playing app was for this purpose.

01:16:29 ◼   ► But lots of things can take it.

01:16:31 ◼   ► So anything that plays video, anything that plays audio,

01:16:34 ◼   ► if you view a video in Safari, for instance,

01:16:39 ◼   ► that becomes then the latest,

01:16:41 ◼   ► the last app that played audio.

01:16:43 ◼   ► So it could be lots of different things

01:16:45 ◼   ► that you don't even realize are taking it over,

01:16:46 ◼   ► but if you play that quick video and then it stops,

01:16:51 ◼   ► it doesn't give the last audio app status

01:16:54 ◼   ► back to what it had before.

01:16:56 ◼   ► It just stays on whatever that was.

01:16:59 ◼   ► So much of the time in practice, that's the problem,

01:17:02 ◼   ► is you actually did play audio through something else

01:17:05 ◼   ► and either subconsciously assumed

01:17:08 ◼   ► that it would go back to Overcast after that, or that you just forgot or didn't realize

01:17:13 ◼   ► that that's happening.

01:17:14 ◼   ► So that makes sense, but that's never the case with me because I've never heard it

01:17:18 ◼   ► do a tweet.

01:17:19 ◼   ► And I think I've done what you've described, like played a video on Twitter, played a video

01:17:22 ◼   ► on YouTube, but I've never heard it resume a YouTube video, resume a tweet video, resume

01:17:26 ◼   ► anything.

01:17:27 ◼   ► It always starts playing seemingly random playlists in my music collection.

01:17:31 ◼   ► And I can tell you, I'm not listening to music on my iPhone, like ever.

01:17:35 ◼   ► Like because I have a dedicated iPod connected with a USB cable in my car

01:17:39 ◼   ► Like I don't even listen to my phone music on my phone in my car. I have a separate iPod for that, right?

01:17:44 ◼   ► And it's amazing to me and I know it's doing it was taking a really long time. It's amazing to me

01:17:49 ◼   ► I don't I don't know how it picks the playlist cuz it's not even like the playlist that I was last like if I go into

01:17:55 ◼   ► music app intentionally and play a song and then forget about it that like it just I

01:17:58 ◼   ► Don't know what it's doing

01:18:00 ◼   ► Is it is it?

01:18:01 ◼   ► honoring my shuffle, but anyway, it always plays music

01:18:03 ◼   ► and it takes forever for it to start playing music

01:18:05 ◼   ► because I have never even launched the music app.

01:18:08 ◼   ► - Right, so there's other things at play here.

01:18:10 ◼   ► So first of all, even if it remembers properly

01:18:14 ◼   ► the last app you used was Overcast,

01:18:17 ◼   ► if Overcast has crashed in the background, it won't resume.

01:18:20 ◼   ► But again, I try to minimize background crashes.

01:18:25 ◼   ► There are occasional ones that still happen

01:18:27 ◼   ► that I'm still trying to track down,

01:18:28 ◼   ► but they're very, very, they're relatively rare

01:18:30 ◼   ► compared to other things, so that probably is not

01:18:33 ◼   ► doing what you're doing now, but I can't tell you

01:18:34 ◼   ► for sure it isn't.

01:18:35 ◼   ► So other things that are at play here,

01:18:39 ◼   ► if it tries to resume something like in a car

01:18:44 ◼   ► setting like that, and the last used audio app

01:18:47 ◼   ► does not respond in a certain amount of time,

01:18:50 ◼   ► I think iOS defaults to music.

01:18:54 ◼   ► There's also, this has been a problem for cars

01:18:58 ◼   ► for a while that had iPod support.

01:19:01 ◼   ► There are these iPod control protocols

01:19:03 ◼   ► that work over with USB and Bluetooth, I think.

01:19:07 ◼   ► And if a car has iPod support, what will sometimes happen

01:19:11 ◼   ► is the car will basically invoke that

01:19:14 ◼   ► instead of the generic Bluetooth play thing.

01:19:18 ◼   ► And iOS devices, when prompted for music support,

01:19:21 ◼   ► try to do the right thing.

01:19:22 ◼   ► They try to show their music library to the car.

01:19:24 ◼   ► And you said that the music's kind of random.

01:19:26 ◼   ► So it used to be, whenever it would ask for the iPod library music stuff, it used to be

01:19:31 ◼   ► that most of them would default to playing the first alphabetical song title.

01:19:36 ◼   ► So whatever song you had that began with A in your library, it would just play that every

01:19:41 ◼   ► time.

01:19:42 ◼   ► My car does that with the USB connected iPod touch.

01:19:45 ◼   ► It always goes with the A songs.

01:19:46 ◼   ► Here's the thing.

01:19:47 ◼   ► It still takes me two songs to figure out what it's doing.

01:19:50 ◼   ► I don't know why I haven't identified the first song with the second song.

01:19:54 ◼   ► Sometimes I'm up to the third song because I kind of like the first two songs.

01:19:56 ◼   ► It's like, "Oh, this is good."

01:19:57 ◼   ► Oh, and I do use random play.

01:19:59 ◼   ► It's like, "Oh, conceivably, I could have picked the song 'Rainbow' by the third song."

01:20:02 ◼   ► I'm like, "Wait a second.

01:20:03 ◼   ► These all begin with A." And like, how many years am I going to do that?

01:20:06 ◼   ► I'm dumb.

01:20:08 ◼   ► And just as a note, okay, as this episode has been somewhat about, I'm not perfect.

01:20:13 ◼   ► I make a lot of mistakes.

01:20:15 ◼   ► But as a note to people who implement these kind of audio systems, nobody ever wants to

01:20:20 ◼   ► play songs alphabetically by their titles.

01:20:22 ◼   ► that is never a thing anybody ever wants,

01:20:24 ◼   ► it is never the right idea,

01:20:26 ◼   ► that just never ever ever do that.

01:20:28 ◼   ► And there are so many like cars and things that do that,

01:20:31 ◼   ► sometimes it's the only option in certain like views

01:20:33 ◼   ► or certain parts of the UI,

01:20:35 ◼   ► nobody wants alphabetical song playback, ever.

01:20:39 ◼   ► Just don't do that.

01:20:40 ◼   ► Anyway, it seems like in recent times,

01:20:42 ◼   ► Apple has figured out, okay, nobody wants to hear

01:20:45 ◼   ► All In by Stroke Nine every single morning

01:20:47 ◼   ► when they get in their car,

01:20:48 ◼   ► let's instead play something from Apple Music

01:20:51 ◼   ► they might like.

01:20:52 ◼   ► So if you subscribe to Apple Music, you often get that,

01:20:55 ◼   ► where your car will suddenly start playing

01:20:56 ◼   ► just some kind of radio station or some playlist

01:21:00 ◼   ► off of your phone if you can't figure out what else to do

01:21:02 ◼   ► and the car is requesting music.

01:21:04 ◼   ► So that's a thing too, and that's been, I think,

01:21:06 ◼   ► in the last two iOS versions, I think since iOS 10 at least.

01:21:10 ◼   ► So there's lots of things that your car

01:21:14 ◼   ► and your phone can negotiate and do.

01:21:17 ◼   ► What you quickly realize as you dive into this stuff

01:21:19 ◼   ► is that this is just an incredibly complex system.

01:21:21 ◼   ► And most of it's complex for a reason.

01:21:23 ◼   ► Like, you know, iOS's whole concept of the last used

01:21:27 ◼   ► music playing or audio playing app,

01:21:29 ◼   ► that is complex in itself because of all the things

01:21:30 ◼   ► I mentioned like videos and Twitter and Facebook

01:21:33 ◼   ► and stuff like that, you know, embedded stuff in Safari,

01:21:36 ◼   ► YouTube videos, like it's so complex to even just try

01:21:39 ◼   ► to keep track in a reasonable way of like,

01:21:41 ◼   ► what was the last app that played media?

01:21:44 ◼   ► And so if the user says play, what should I play?

01:21:48 ◼   ► So I don't have a good answer for you,

01:21:50 ◼   ► except it might be one of those things,

01:21:53 ◼   ► or it might be something else.

01:21:55 ◼   ► (laughing)

01:21:56 ◼   ► - Nice. - I mean, like I said,

01:21:58 ◼   ► almost, most of the time it works, 85% of the time,

01:22:00 ◼   ► 90% of the time, it does what I expect it to do,

01:22:02 ◼   ► but it's the times that it doesn't,

01:22:04 ◼   ► like, well, what was your problem this time?

01:22:06 ◼   ► What you said about it trying to get it to play,

01:22:09 ◼   ► but like it giving up after a short period of time

01:22:11 ◼   ► and then going to music rings true for me,

01:22:13 ◼   ► because it always seems to take longer.

01:22:15 ◼   ► I can always tell when, uh-oh,

01:22:16 ◼   ► I'm not gonna hear my podcast,

01:22:17 ◼   ► I'm gonna hear music because everything on my car takes a long time in terms of Bluetooth and booting up and everything

01:22:22 ◼   ► You know and so there this is are counted in seconds. I can say oh, I know it's gonna know what's gonna happen next

01:22:28 ◼   ► It's gonna play music and it does but almost all the time at work. So

01:22:31 ◼   ► Anyway, another thing. I mean, I guess that can be solved by

01:22:34 ◼   ► Honda updating their in-car infotainment system which by the way, they did the new accords have an all-new android-based

01:22:42 ◼   ► presumably

01:22:44 ◼   ► non-disgusting

01:22:46 ◼   ► You know like that. What do I have now is like pre Android iOS like the old world of

01:22:52 ◼   ► car

01:22:54 ◼   ► infotainment systems the bad old world and the new one is just you know

01:22:57 ◼   ► The the modern era so I'm hoping it's better, and I'm hoping they put faster CPUs

01:23:02 ◼   ► And I'm hoping it will take less time to connect to Bluetooth so the very least I can get the wrong thing playing faster

01:23:08 ◼   ► In the meantime I will just you know I mean it's pretty pretty high success rate all things considered

01:23:14 ◼   ► But yeah, it does sound complicated.

01:23:16 ◼   ► It almost makes me wish that there was a way that I could say, "Look, all I ever wanted

01:23:19 ◼   ► to happen, ever, ever, ever when I enter the car is for you to start playing Overcast."

01:23:23 ◼   ► And that's it.

01:23:24 ◼   ► But that doesn't seem like it's in the cards.

01:23:27 ◼   ► So before you get a ton of email, hopefully, you were able to backchannel kind of directly

01:23:35 ◼   ► to Apple and say, "Hey, what's going on here?"

01:23:38 ◼   ► What do you think would have been your course of action if that wasn't available to you?

01:23:43 ◼   ► Like aren't there--

01:23:44 ◼   ► - DTS ticket would be the next thing.

01:23:46 ◼   ► - That's exactly what I thought.

01:23:47 ◼   ► So if you're not Marco Arment,

01:23:50 ◼   ► and you know, yeah, that's fine for Marco,

01:23:51 ◼   ► but you would have opened a DTS ticket,

01:23:53 ◼   ► and hypothetically, that would have been escalated

01:23:56 ◼   ► to an engineer who could have helped you, hopefully maybe.

01:23:59 ◼   ► - Yeah, and the way these works,

01:24:00 ◼   ► and forgive me for not having the details right,

01:24:03 ◼   ► 'cause I've actually never filed one,

01:24:04 ◼   ► but basically every paid Apple developer membership

01:24:08 ◼   ► includes, I believe, two developer technical support,

01:24:11 ◼   ► with DTS, incidents per year.

01:24:14 ◼   ► And you basically, you say you wanna use one,

01:24:18 ◼   ► and you can actually get code level support.

01:24:21 ◼   ► You can actually say, okay look, here's this code,

01:24:24 ◼   ► it is not working here, I can't figure this out,

01:24:26 ◼   ► can you help?

01:24:27 ◼   ► And Apple engineers will actually help you

01:24:31 ◼   ► with that problem.

01:24:31 ◼   ► And so it's a pretty involved process,

01:24:33 ◼   ► that's why they're limited.

01:24:35 ◼   ► So it's two per year, and I think there's some provision

01:24:39 ◼   ► where if what you found is actually an Apple bug,

01:24:41 ◼   ► then I think they don't charge you your ticket for that year.

01:24:45 ◼   ► They don't accumulate, so I would have a lot if they did,

01:24:49 ◼   ► but they don't.

01:24:51 ◼   ► But yeah, so there actually is this,

01:24:54 ◼   ► there is a way to do this,

01:24:55 ◼   ► and while I have no experience with it myself,

01:24:57 ◼   ► I've heard from people who have used them

01:24:59 ◼   ► that they're incredibly helpful,

01:25:01 ◼   ► and so this is something

01:25:02 ◼   ► that all developers have access to, so that's pretty cool.

01:25:05 ◼   ► - That's awesome.

01:25:07 ◼   ► - All right, we good?

01:25:09 ◼   ► - I mean, I could tell some more long,

01:25:10 ◼   ► boring stories if you want.

01:25:12 ◼   ► - Well, we gotta figure out an after show

01:25:13 ◼   ► 'cause I have something brief,

01:25:14 ◼   ► but I don't know if it's gonna be worth it.

01:25:16 ◼   ► - Oh, I know what we should do.

01:25:17 ◼   ► All right, thanks to our three sponsors this week,

01:25:19 ◼   ► Betterment, Eero, and Fracture,

01:25:21 ◼   ► and we will see you next week.

01:25:23 ◼   ► ♪ Now the show is over ♪

01:25:28 ◼   ► ♪ They didn't even mean to begin ♪

01:25:30 ◼   ► ♪ 'Cause it was accidental ♪

01:25:32 ◼   ► ♪ Accidental ♪

01:25:33 ◼   ► ♪ Oh, it was accidental ♪

01:25:35 ◼   ► ♪ Accidental ♪

01:25:36 ◼   ► John didn't do any research, Marco and Casey wouldn't let him

01:25:41 ◼   ► 'Cause it was accidental (it was accidental)

01:25:44 ◼   ► It was accidental (accidental)

01:25:47 ◼   ► And you can find the show notes at ATP.fm

01:25:52 ◼   ► And if you're into Twitter, you can follow them

01:25:56 ◼   ► @c-a-s-e-y-l-i-s-s

01:26:00 ◼   ► So that's Casey Liss, M-A-R-C-O-A-R-M

01:26:05 ◼   ► ♪ Anti-Marco Armin, S-I-R-A-C ♪

01:26:10 ◼   ► ♪ USA, Syracuse, it's accidental ♪

01:26:14 ◼   ► ♪ Accidental ♪

01:26:16 ◼   ► ♪ They didn't mean to ♪

01:26:18 ◼   ► ♪ Accidental ♪

01:26:19 ◼   ► ♪ Accidental ♪

01:26:21 ◼   ► ♪ Tech podcast ♪

01:26:23 ◼   ► ♪ So long ♪

01:26:25 ◼   ► - So John, tell us about your Mongoose Californian.

01:26:29 ◼   ► (John laughing)

01:26:31 ◼   ► You're trying to really clear up the things in there.

01:26:33 ◼   ► - Yeah, we have to set the stage.

01:26:34 ◼   ► We have to set the stage.

01:26:35 ◼   ► So forever and a day ago, at the very top of our show notes has some sections, including

01:26:44 ◼   ► follow-up, and then the topics, right?

01:26:46 ◼   ► So the very, very, very top, like the fifth line down, sitting in this document for maybe

01:26:54 ◼   ► a year now, is my mongoose Californian, parenthesis, 1984, parenthesis, and two links.

01:27:01 ◼   ► And this has been sitting there staring us in the face at the top of this document for

01:27:06 ◼   ► probably a year, maybe two.

01:27:08 ◼   ► Three years.

01:27:09 ◼   ► Three years.

01:27:10 ◼   ► It's been three years.

01:27:11 ◼   ► And now is the time that we are going to finally clear this out.

01:27:15 ◼   ► Good call, Marco.

01:27:16 ◼   ► The thing that we really should be clearing out is like early on, someone, probably Casey,

01:27:20 ◼   ► was trying to put sections in the notes.

01:27:22 ◼   ► We didn't have names for things.

01:27:24 ◼   ► It's, you know, the naming problem with programming again.

01:27:26 ◼   ► Like it hadn't actually come up with, you know, precise names for things.

01:27:31 ◼   ► So there's all sorts of like pre post show, public post show, after show, pre show.

01:27:36 ◼   ► Like we didn't have the names for all the different times that we can talk to each other.

01:27:40 ◼   ► The times when we're live, when we're not live, blah, blah, blah.

01:27:43 ◼   ► So this section got anchored up here with public post show, which now we call the after

01:27:48 ◼   ► show and it's lower in the document.

01:27:50 ◼   ► And so that's how old this was.

01:27:53 ◼   ► So I will be able to delete this entire section when we do this.

01:27:55 ◼   ► The problem with this topic is it was a thing I was angry about.

01:27:59 ◼   ► I was angry about it in 2014, unfortunately.

01:28:02 ◼   ► It's hard for me to muster that anger again.

01:28:04 ◼   ► Oh, oh, really?

01:28:07 ◼   ► Is that so, Jon?

01:28:08 ◼   ► Well, I demand that you do it, because you made me dig up my crash story.

01:28:12 ◼   ► You weren't angry about it.

01:28:13 ◼   ► That was a detective story.

01:28:16 ◼   ► The anger should have subsided.

01:28:18 ◼   ► You defeated the beast.

01:28:19 ◼   ► Now you can look back on it, and it's more like a detective story.

01:28:23 ◼   ► There was this mystery, and I solved it, and it was weird, and it's like a programming

01:28:27 ◼   ► problem.

01:28:28 ◼   ► had you do it in the middle of it like your app is currently out there and crashing and

01:28:32 ◼   ► you still can't figure out what it is and then you talked about it, that would be a

01:28:34 ◼   ► different story.

01:28:35 ◼   ► But anyway, I'll see what I can do to bring myself back to 2014.

01:28:39 ◼   ► This is precipitated by me being on a podcast that we'll link in the show notes, "Storming

01:28:44 ◼   ► Mortal" where I talked about a whole bunch of things including, in the words of a show

01:28:50 ◼   ► that God help me Casey must have seen and please Marco know that it at least exists,

01:28:56 ◼   ► The best Christmas present I ever received and ever would receive.

01:29:00 ◼   ► My 1984 Mongoose, California BMX bike.

01:29:03 ◼   ► Casey, do you know what I was referencing?

01:29:06 ◼   ► Nope.

01:29:07 ◼   ► Not a bit.

01:29:08 ◼   ► Come on, guys.

01:29:09 ◼   ► Come on.

01:29:10 ◼   ► What was it?

01:29:11 ◼   ► Did I get it wrong?

01:29:12 ◼   ► The best Christmas present I ever received and ever would receive?

01:29:17 ◼   ► I'm close.

01:29:18 ◼   ► I'm really close.

01:29:19 ◼   ► You're not close enough for us.

01:29:22 ◼   ► A Christmas story.

01:29:23 ◼   ► They only play it 24 hours a day.

01:29:24 ◼   ► I've only seen it like once or twice.

01:29:26 ◼   ► Oh, what do you mean you've only seen it once or twice?

01:29:29 ◼   ► I think I've seen it all the way through.

01:29:30 ◼   ► They play it literally 24 hours.

01:29:32 ◼   ► I know, I never pay attention.

01:29:34 ◼   ► Yeah, exactly.

01:29:34 ◼   ► It's one of those things that it's always on and so you never pay attention.

01:29:38 ◼   ► If you had said you'll shoot your eye out with that thing, I would have known immediately.

01:29:41 ◼   ► Well, at least there is that.

01:29:42 ◼   ► Anyway, that movie, by the way, is a really, really, really good movie.

01:29:46 ◼   ► Ignore the fact that it's played 24 hours.

01:29:48 ◼   ► They used to not play it 24 hours.

01:29:49 ◼   ► It is a really good movie.

01:29:51 ◼   ► the kid in the movie gets uh... begun for christmas and he says that the

01:29:55 ◼   ► narrator the older version of says the best christmas present he ever see you

01:29:58 ◼   ► never received

01:29:59 ◼   ► that's my that's my uh... bike for variety of reasons

01:30:02 ◼   ► uh... and the reason the reference for a poster recognize there was not the bike

01:30:06 ◼   ► itself just

01:30:07 ◼   ► the best present that you know you know you're at the last half a very

01:30:10 ◼   ► very near the end of the movies at the second or third to last line in the

01:30:14 ◼   ► movies be sent chance

01:30:15 ◼   ► no it's pretty significant line expressing the sentiment uh... anyway i

01:30:19 ◼   ► I should get that quote exactly right so I can repeat it with more confidence in the

01:30:24 ◼   ► future.

01:30:25 ◼   ► Anyway, on this podcast we talked about a whole bunch of things I also talked about

01:30:27 ◼   ► at BMX and around this time I was buying bicycles for my kids of various sizes, like non-training

01:30:35 ◼   ► wheel bikes, like their first real bikes.

01:30:37 ◼   ► And I found that experience incredibly frustrating.

01:30:40 ◼   ► So I grew up in a bike culture, which is like 70s, 80s, suburban New York metro area.

01:30:49 ◼   ► It was a bike culture.

01:30:51 ◼   ► Kids were on bicycles.

01:30:52 ◼   ► Can I tell people that you're a biker?

01:30:54 ◼   ► Different, different thing, different context.

01:30:56 ◼   ► Although we did put playing cards in our spokes with a clothespin so your bike sounded like

01:31:00 ◼   ► motorcycles.

01:31:01 ◼   ► Oh yeah, that really confused anybody ever.

01:31:04 ◼   ► That's right, it was awesome.

01:31:06 ◼   ► So kids were on bikes all the time.

01:31:07 ◼   ► Like if you watch a Steven Spielberg movie

01:31:09 ◼   ► showing kids in the 70s racing

01:31:10 ◼   ► and riding around on bicycles,

01:31:11 ◼   ► if you watch Stranger Things

01:31:12 ◼   ► and kids are riding around on bicycles,

01:31:13 ◼   ► that was a real thing.

01:31:14 ◼   ► That's all we did.

01:31:15 ◼   ► We'd come home from school, we'd get on a bike,

01:31:16 ◼   ► so we'd go all over the place.

01:31:18 ◼   ► Nobody had helmets.

01:31:19 ◼   ► We would fashion jumps out of garbage.

01:31:21 ◼   ► We found in other people's trash on trash day

01:31:23 ◼   ► in the middle of the street.

01:31:25 ◼   ► It's all very dangerous and scary,

01:31:27 ◼   ► and we've never let our kids do it today,

01:31:29 ◼   ► but that's what we did.

01:31:30 ◼   ► And because we were in a bike culture

01:31:32 ◼   ► and bikes were so important,

01:31:34 ◼   ► there's a hierarchy of what kind of bike you had

01:31:36 ◼   ► and all the fancy rich kids had Mongoose or Redline,

01:31:40 ◼   ► which were expensive bikes, and I wanted a Mongoose.

01:31:43 ◼   ► I think I thought Redlines were more expensive.

01:31:45 ◼   ► I don't even think they were, but that's what I thought.

01:31:48 ◼   ► And I wanted one for Christmas,

01:31:49 ◼   ► and I thought there's no way I'm gonna get one

01:31:50 ◼   ► 'cause they're really expensive, and they were,

01:31:52 ◼   ► and my parents totally surprised me with it.

01:31:54 ◼   ► Like the present that you don't think

01:31:55 ◼   ► you're ever gonna get.

01:31:56 ◼   ► It's not even I was like hopeful for it, looking for it.

01:31:58 ◼   ► It was just, it wasn't even, didn't even enter my mind.

01:32:00 ◼   ► It's like being a kid and saying,

01:32:01 ◼   ► "I want a Lamborghini Countach," and you're like,

01:32:03 ◼   ► "Yeah, but like whatever, I'm not gonna get one."

01:32:04 ◼   ► And then you wake up on Christmas morning,

01:32:05 ◼   ► it's in your garage and it's yours, right?

01:32:07 ◼   ► It was like that, basically.

01:32:09 ◼   ► And I rode the hell out of that bicycle.

01:32:12 ◼   ► I rode it everywhere, jumped it off jumps

01:32:14 ◼   ► and dirt bikes and trails, rode up and down highways.

01:32:17 ◼   ► This bike has no gears, by the way, it's a BMX bike.

01:32:19 ◼   ► Like it's just straight up, took really good care of it,

01:32:23 ◼   ► cleaned it and oiled it all the time.

01:32:25 ◼   ► Like all the things you can imagine doing this bike,

01:32:27 ◼   ► it's like my prized possession.

01:32:28 ◼   ► I still have this bicycle, it's still in my house.

01:32:30 ◼   ► - Oh my word, wow, in the attic?

01:32:33 ◼   ► - In the basement.

01:32:34 ◼   ► I tried to get my kids to ride the bike.

01:32:36 ◼   ► It's like the reason I saved it for my kids.