311: Mutually Assured Destruction


00:00:00   Oh god. I mean, on the one side, a little sliver of me does feel ever so slightly bad,

00:00:06   but 99% of me is like, "You f***ers deserve this."

00:00:09   Yeah, we'll get to it.

00:00:11   Oh god, just such, such obnoxious bastards.

00:00:15   Save it, save it, save it. Save it for the show. This is the show we're about to do,

00:00:19   including this topic.

00:00:20   Save it. Save it. Save it. Save it.

00:00:21   Well, because I promise I need to get the obnoxious bits out of my system so I don't

00:00:26   say obnoxious s*** on the show.

00:00:28   Oh god, it gets me so fired up.

00:00:31   Wait, all this time, am I supposed to have been doing that instead of saying, "Oh, I've

00:00:34   knocked your stuff on the show?"

00:00:35   Well, you say it, then you just edit it out. You just edit out the things that you said

00:00:39   that you didn't want to be in.

00:00:40   Well, most of them.

00:00:41   We should dive right in, and I need to apologize for all three of us, but particularly me.

00:00:48   Last episode, we were talking about how the kids these days, the youths, have ways around,

00:00:54   you know, or ways to accomplish what basically Instagram close friends does. And I was real

00:00:59   smug last week because I knew about this, and I knew the name of this. And I forget

00:01:04   exactly what I said, but spoiler alert, it was wrong. So anyway, so I forget what I called

00:01:08   it. Maybe I called it fake Instagram or something, but what I was thinking of and couldn't actually

00:01:13   emit from my body was Finsta, F-I-N-S-T-A, which does stand for fake Instagram. I can't

00:01:19   remember what I called it last week. It doesn't really matter. But yeah, it's Finsta. That

00:01:24   is the kind of like casual Instagram that all the youths are doing. And I don't know

00:01:28   why I didn't remember that. And I was so smug about it because I was like, oh, I know this.

00:01:32   Oh, Friendstagram. Thank you, Brian Mitchell. There's Friendstagram that I said, God, is

00:01:34   that right?

00:01:35   Well, Friendstagram actually makes sense as a name. Finsta and spam account, both are

00:01:40   the opposite of what they mean because like a spam account, it's like where, well, I guess

00:01:44   it depends on what you mean by spam, but I feel like it's your fake Instagram is the

00:01:47   most real because it's the one where you are the least fake. The actual fake one is the

00:01:52   one where you present your beautiful life and spam like implies bad. Yeah. I mean, maybe

00:01:59   you're saying if I have a garbage picture, I put it on my spam account. It makes some

00:02:03   sense and quantity. There's just a lot of it because you don't care as much to your

00:02:06   point. But yeah, I'm with you. Friendstagram does make more sense. Why don't you try that?

00:02:10   You can try to make that happen in case you know, it's, it's, that's going to happen right

00:02:13   after fetch. Anyway, I apologize that we're all old men and we really screwed that up

00:02:17   in this will teach me never to be smug about thinking that I'm not old because I am in

00:02:21   date. I did not apologize for being old. I'm a little confused. So is the, is the finsta

00:02:26   the one that you show everyone or the one that you show only your friends? No, it's

00:02:30   the one you show only your friends. So it doesn't make any sense. But it stands for

00:02:33   fake. Yeah, correct. I, yeah, I'm too old to understand this. Kids are dumb. We got

00:02:38   to move on. We're, we're too old. I just wanted to apologize for being old. John, stop including

00:02:42   me. That's just Marco and Casey are, have a, are self hating old people. I am not. I think

00:02:51   it's just that I don't want to speak for Marco, but I'm still clinging to the thought that

00:02:53   I have some amount of awareness of what the, what the kids are doing these days. And it's

00:02:57   a, no, I don't, I really don't. I just need to embrace it. So anyway, moving on. We talked

00:03:01   last week about this quote unquote stage light effect on Mac books and Mac book pros where

00:03:06   there's like a series of spotlights. It almost looks like coming from the bottom of the display

00:03:12   and anonymous person wrote in and said, I work as a genius at a very busy Apple store.

00:03:16   And my colleagues and I triage and repair literally hundreds of Macs a week when there's

00:03:20   a genuine manufacturing or design fault that affects a significant portion of devices for

00:03:24   a particular model. It is very quickly and unambiguously obvious what it is and where,

00:03:29   you know, the exact symptoms and so on and so forth. I'm paraphrasing here. For instance,

00:03:34   the keyboard issues in the post 2016 Mac book pros were definitely a real problem. And the

00:03:40   eventual, what is it? Repair extension program. That's what, yeah, yeah. Everyone kind of

00:03:45   thought that was coming, blah, blah, blah. Anyway, so this anonymous genius says, you

00:03:50   know, all of us were really honestly baffled about the brouhaha about the stage lighting

00:03:56   thing because they apparently have seen no pattern or increased symptoms from their customers

00:04:02   and they see a lot of customers. And this individual is actually asked around and apparently

00:04:08   nobody really knows what we were talking about last week. Now granted, we were just going

00:04:13   off of the news reports that we had read. So maybe we got hoodwinked by some fake news.

00:04:18   I don't know.

00:04:19   Tim Cynova This type of, we usually don't have access

00:04:21   to this type of information because it's not as if Apple has some sort of giant dashboard

00:04:25   showing the prevalence of repairs that they're, you know, there's not public information,

00:04:29   right? So when we see a story that says there's a problem with Apple product or whatever,

00:04:36   the only things we have to judge it by are like how many people do we personally know

00:04:40   who that happened to? And we know a lot of people who know a lot of people who use Apple

00:04:42   products. So that's something and also the reputation of whatever, you know, website

00:04:47   or is that is publishing the information. In this case, this, I forget where this was.

00:04:52   Maybe it was Apple insider or something. It's very easy for just one person who has a problem

00:04:56   to be angry about it and get it, put it online and make it seem like it's an epidemic. But

00:05:01   the, obviously Apple knows for sure. But they're not talking. So having sort of leaks like

00:05:05   this from anonymous geniuses, I think they carry a lot of weight, assuming they're actually

00:05:10   good. Because, you know, to this person's point, a single store does see hundreds and

00:05:15   hundreds of Macs a week. And so that's a pretty good sample size. And then they may not know

00:05:19   other people who work in other Apple stores and the word kind of gets around like that.

00:05:23   It's not, if there actually is a widespread problem, they are in the best position to

00:05:28   know outside of the people who keep the spreadsheets back at Apple. So we'll see if we hear about

00:05:33   this problem anymore or if it was just a thing that a few people got and were mad about,

00:05:37   it is not an actual epidemic like the keyboard issues.

00:05:41   John, you want to tell me about the silence of Siri and other Internet of Things information?

00:05:45   Yeah, we got a lot of feedback about my little smart outlet and my experiences with Siri.

00:05:52   One thing I was speculating about was I was asking if HomeKit knew what room it was in

00:05:59   and if it was using that information to determine whether it talked back to me. Pretty much

00:06:03   everybody said, yeah, it should know what room it's in. My HomePod actually is in the

00:06:09   same room as the outlets. And in theory, if it's in the same room as the outlets and I

00:06:13   ask the HomePod to turn the lights on or off, it shouldn't talk back to me at all. Some

00:06:19   people had interesting theories about when HomePod decides to talk back to you. I didn't

00:06:24   really test any of these and I'm not sure if they're based on documentation or just

00:06:28   like a notion. One person, I think my favorite one, said, "If you're far away or if you

00:06:34   yell, it will talk back to you to confirm."

00:06:38   Oh, interesting.

00:06:40   I don't know if that's true. Like, far away my voice would be faint, but yelling it would

00:06:45   be loud? Very confusing. Most people agree, though, if you were in a different room. So

00:06:51   if you asked it to turn off or on something that's not in the same room as the HomePod,

00:06:56   it will talk back. All that said, my HomePod has always been in the same room as these

00:07:00   outlets, and it has talked back to me occasionally. Why? I don't know. So there's this mystery

00:07:04   still abound, but in theory, it should be able to do this.

00:07:08   Same thing with the other devices, the Amazon Echo devices. They have room awareness. You

00:07:12   can place things in a room and in general, they try to be silent or less verbose if you're

00:07:18   doing something in the same room. If the thing you're talking to is in the same room as the

00:07:22   devices. So that's good to know. In practice, it hasn't talked back to me for a while now,

00:07:27   so maybe, I don't know, maybe it's just halting a grudge or it's giving me the silent treatment,

00:07:33   but either way, I'm liking it.

00:07:35   Moving on, but in the same vein, [bleep] wrote that he was involved in creating a technology

00:07:42   known as Wireless Accessory Config, or WAC. This is an Apple technology that allows an

00:07:48   app developer to basically load the current Wi-Fi information onto an external device.

00:07:55   This was in the context of John saying that, "Oh, my smart outlet, just kind of by magic,

00:08:01   figured out what my Wi-Fi information was, etc., etc." He wrote this mechanism, or was

00:08:07   involved in this mechanism that does this, and he said, "Apple was sick of the Wi-Fi

00:08:11   network dance that we had to go through for these accessories and decided it was a problem

00:08:16   worth solving on the platform level. It's available to any MFI partner for both AirPlay

00:08:20   and HomeKit accessories, and you can turn on the entitlement for your app in Xcode to

00:08:24   allow it to find and configure these accessories as well," which I thought was pretty cool.

00:08:27   And apparently, it does automate the whole drop off your Wi-Fi, connect to the device's

00:08:34   Wi-Fi, upload the information, and get back on your regular Wi-Fi dance, which is what

00:08:38   I had said that I had to do by hand with a lot of this stuff. I didn't realize, because

00:08:43   I had seen this happen before, where it just works by magic and you don't have to go

00:08:46   to settings and change Wi-Fi access points, whatever. Well, apparently, this is just automating

00:08:53   that whole thing, and that was surprising to me. I guess I should have guessed, but

00:08:58   I didn't realize it was just automating that whole dance.

00:09:01   Yeah, he also provided links to some patents that are related to this type of technology.

00:09:07   It's got a name, it's got an acronym, but as is the case with a lot of stuff in HomeKit,

00:09:11   this is one thing Apple has been smart about with HomeKit, but it's also historically

00:09:15   been a weakness of HomeKit. As far as you're concerned, it's just this one thing called

00:09:20   HomeKit. It has been difficult for devices to get qualified with HomeKit either because

00:09:25   you have to go through this testing and you have to buy a bunch of stuff from Apple and

00:09:28   it can be expensive. They had all these problems in the beginning. If you got a HomeKit device,

00:09:34   it had a good user experience, but there are far more devices from other manufacturers

00:09:38   just because what they had was simpler or cheaper or less of a hassle to get done or

00:09:43   all sorts of other things. Apple has been changing that slowly, but still, the idea

00:09:48   that there's one thing you look for, this works with HomeKit. Under that umbrella, you

00:09:52   get a bunch of stuff like this, the technology whose name nobody probably even knows or cares

00:09:56   about that basically just makes the experience better is one thing that's attracting me

00:10:01   to the HomeKit ecosystem. Despite the fact that the other Internet of Things ecosystems

00:10:05   seem much larger and more diverse, I'm impressed with how much the core experience of HomeKit

00:10:12   is polished. I know people say, "Well, you can't do as much with HomeKit as you can

00:10:14   do with other things," and that's undoubtedly true, but I do like the fact that the main

00:10:21   gameplay loop got – I don't know why I've been playing too much Destiny. The main gameplay

00:10:26   loop in HomeKit seems to work well. I had a rough day in Destiny today.

00:10:34   Do you want to talk about it?

00:10:35   Well, maybe. We have so much stuff. Don't worry about it. Thank God.

00:10:39   Oh, thank God. I mean, I'm sorry for your bad day, John.

00:10:41   I can put it in that. It wasn't as bad as when I lost all my legendary shards, but don't

00:10:47   worry about it.

00:10:48   Well, I am sending good Destiny thoughts your way or something.

00:10:53   We are sponsored this week by Casper. You spend a third of your life sleeping, so you

00:10:59   should be comfortable. Casper's experts work tirelessly to make quality sleep surfaces

00:11:04   that cradle your natural geometry in all the right places. They are designed to cleverly

00:11:09   mimic human curves, providing supportive comfort for all kinds of bodies.

00:11:14   The original Casper mattress combines multiple supportive memory foams for a quality sleep

00:11:19   surface with just the right amount of both sink and bounce, and a breathable design to

00:11:23   help you sleep cool. With over 20,000 reviews and an average of 4.8 stars across Casper,

00:11:29   Amazon and Google, Casper is becoming the Internet's favorite mattress.

00:11:33   And Casper also now offers two additional mattress models, the Wave and the Essential.

00:11:37   The Essential has a streamlined design at a price that won't keep you up at night.

00:11:41   The Wave, and this is the one I sleep on and can strongly recommend, the Wave has a patent-pending

00:11:45   premium support system to mirror the natural shape of your body. They also have a wide

00:11:51   array of other products like pillows and sheets. All of these are designed, developed and assembled

00:11:56   right here in the U.S. and brought to you at affordable prices because they cut out

00:12:00   middlemen and sell directly to you. All of these are risk-free. They have a 100-night

00:12:05   risk-free sleep-on-it trial so you can be sure of your purchase. You have hassle-free

00:12:10   returns if you aren't completely satisfied, free shipping and free returns in the U.S.

00:12:14   and Canada, so really you can be sure of your purchase with this 100-night risk-free sleep-on-it

00:12:19   trial. Get $50 toward select mattresses by visiting

00:12:22   casper.com/atp and using code ATP at checkout. Terms and conditions apply. That's casper.com/atp

00:12:30   with code ATP. Thank you so much to Casper for sponsoring our show.

00:12:34   [Music]

00:12:37   There's been some interesting things going on in the Apple ecosystem over the last, I

00:12:42   don't know, 48 hours. This began, I don't know, like six months ago, a year ago, something

00:12:47   like that, with this thing that Facebook created which was called Onavo, and I'm probably pronouncing

00:12:54   that wrong, but Onavo Protect. And my recollection of this was that it was pitched as one of

00:12:59   those VPNs that, like the ones that have sponsored this show, I believe in the past, but sleazy

00:13:06   instead of the ones that have sponsored, which are not sleazy. And really the point of this

00:13:12   VPN, if you peel back all the layers of the onion, was to be able to sniff all of these

00:13:18   people's network data to see where they're going, what they're doing, who they're interacting

00:13:22   with, etc. And this was called Facebook's Onavo Protect app. And once Apple got wind

00:13:29   of what was really going on here, which again was like six months or a year ago, something

00:13:32   like that, they forced Facebook to remove it from the App Store. And that did happen.

00:13:37   Fast forward a few months, and it seems that Facebook re- like kind of white-labeled and

00:13:43   to some degree rebranded the exact same app and started distributing it via Enterprise

00:13:50   Deployment. Generally speaking, when you deploy to the App Store, you know, you give your

00:13:53   binary to Apple and you've signed it with your developer certificate, and then Apple,

00:13:58   you know, signs it again with their certificate and puts it on the App Store. And then in

00:14:01   theory, anyone can download it. With Enterprise Deployment, it's kind of sort of a deliberate

00:14:07   back door to that. It's a deliberate and blessed back door to that where you can ask Apple,

00:14:12   "Hey, we would like to distribute some things internally." And internally is the key there.

00:14:17   Let's say you work for Acme Widgets Incorporated and you have a company directory. Well, you

00:14:22   don't want your company directory on the App Store, and it's probably useless even to put

00:14:26   it behind like a login, you know, on the App Store. The writer answer, arguably, is to

00:14:32   distribute it internally only to your own people. And so somewhere within your company's

00:14:37   intranet, you would have a link to where you could install the Acme Widget Company directory.

00:14:43   And in order to do that, like I said, you have to go to Apple and say, "Hey, we would

00:14:45   like to distribute stuff internally." And they say, "Okay, sure. Here's a certificate

00:14:49   with which you can use to do that." And then through some magic, if you give the person

00:14:54   a link to the app, to the bundle, the IPA that is the app, they can install it and run

00:15:01   it. Well, Facebook leveraged this in order to basically redistribute this Onavo Protect

00:15:08   app outside of their internal people. So what they did was they signed basically the exact

00:15:14   same app, but with their internal enterprise distribution certificate such that they can

00:15:21   get around the App Store. Well, Apple found out about it and they got angry. But before

00:15:27   we talk about that, does my quasi-summary of enterprise distribution, does that ring

00:15:31   basically true to you guys or would you like to refine anything I just said?

00:15:35   I think I'd add is that they, you know, as you said, like they reskin the app or redistributed

00:15:40   it. A couple of people did like class dumps on it. And so they didn't even rename all

00:15:44   the like ONV prefixed classes and functions. Like it was sort of the, I mean, what could

00:15:52   they have really done? Like, I mean, the bottom line is the functionality of the app was going

00:15:55   to get them controlled no matter what, but they didn't even make the most trivial effort

00:15:59   to hide themselves. All they did was basically change the name of the app and maybe reskin

00:16:03   it. But it was like literally all the same source code underneath. And it's sort of an

00:16:08   amazing pattern of behavior where you get banned like with cause, like you're doing

00:16:14   a thing you're not supposed to do. You get booted out of the App Store and you immediately

00:16:18   say, "We need to find a way to distribute this outside the App Store. How can we do

00:16:22   that?" Like you're not chastened. You don't think we have to find a new way to get the

00:16:26   information we need or whatever. You just immediately switch gears and say, "How can

00:16:30   we work around this ban?" Which is like, I feel like that's the instinct. You see this

00:16:36   sometimes, I can't recall any specific stories, but like when it's a young, usually lone,

00:16:43   inexperienced developer who gets a ban or some other ruling from the App Store that

00:16:49   they think is unfair and their immediate reaction is to find a way to fight back, like to try

00:16:55   to get around the ban, to try to circumvent it, to try to find like some way that they

00:16:59   can find a loophole, right? And it ends up in this back and forth. And each time they

00:17:04   go back and forth, Apple gets angrier and angrier and gets bigger and bigger bans until

00:17:08   they are no longer a developer on the App Store at all. And then they like get a fake

00:17:12   credit card. And like it just escalates in a way that you usually only see when it's

00:17:16   just a petulant 20-something versus Apple only. This is not a petulant 20-something.

00:17:23   It is a very large, very important corporation acting like a petulant 20-year-old.

00:17:27   Which is of no great surprise. So Apple catches wind of this and this happened late Tuesday,

00:17:35   is that right? We're recording this late Wednesday. And Apple caught wind of this and nobody really

00:17:40   knew what they were going to do because Facebook has done a lot of really shiesty and shady

00:17:44   stuff in the past. And it seems from an outsider's perspective that they've gotten some slaps

00:17:49   on the wrist, some maybe harder than others, but mostly has been able to do whatever the

00:17:53   crap they want because, hey, they're Facebook. So this morning, Wednesday morning, Apple

00:17:59   decided to pull, if we understand things correctly, there was a little bit of debate about this,

00:18:04   but it sounds like Apple decided to pull their enterprise distribution certificate. Which

00:18:08   means that this rebadged Facebook or Navo Protect app did in fact go away. Because when

00:18:15   all of these iPhones went to run it, presumably at some point or another, they would phone

00:18:19   home and verify that the certificate is still valid and Apple would tell them, "No, it's

00:18:23   not." And they wouldn't run this app. That in and of itself, okay, good. What I find

00:18:28   fascinating about this though is it seems like they pulled the entire enterprise distribution

00:18:36   certificate, which is I know what I just said. But what that means is all of Facebook's internal

00:18:41   apps, like their Acme widget company directory, in this case the Facebook company directory,

00:18:47   that stopped working. Apparently these poor, poor souls couldn't use their app in which

00:18:52   they order food. They would have to actually get up and walk to the free restaurant and

00:18:59   talk to someone to order their food. See, I'm getting angry again.

00:19:04   - It's cruel and unusual. - Getting angry again.

00:19:05   - And all their internal beta testing of all their apps. This is usually how big companies

00:19:10   do internal beta testing, it's a lot easier than the other methods. And so it's a big

00:19:14   deal. They lost all of their internal apps that they were signing with their enterprise

00:19:22   certificate, which on a number of levels is the result of incredible brazenness and also

00:19:29   stupidity. This is not a decision by Apple to interpret a rule differently or to punish

00:19:37   Facebook because they don't like them. Apple is very clear with enterprise distribution

00:19:42   certificates and the agreement that you agree to when you get one, it's very, very clear

00:19:48   that it is for internal use by your own employees only. It is not for any kind of public distribution.

00:19:55   They couldn't possibly be more clear about that. And so Facebook literally sending this

00:20:01   out to people in the public brazenly because they knew they were Facebook, they're invincible.

00:20:07   They deserve every single bit of this. Every drop of the inconvenience this causes their

00:20:12   employees, every single minute of the wasted time that they're gonna have to spend working

00:20:18   with this problem and then hopefully routing around it in some way for themselves. They

00:20:23   deserve every single bit of this because they were literally blatantly, flagrantly violating

00:20:31   the very clear rules of this. It's not even close, it isn't up for interpretation, it

00:20:35   isn't Apple overreaching. It's a very clear cut and dry, Facebook knew exactly what they

00:20:43   were doing. They knew exactly that they were violating this agreement completely both in

00:20:47   spirit and in letter of the law and they violated it anyway and they got hit for it. I don't

00:20:52   know how anybody at Facebook can possibly complain as a result of this.

00:20:56   It's a double whammy too because remember the reason this was banned from the app store

00:21:00   is because the app itself violates Apple's rules of collection and everything. So this

00:21:04   was already, it got banned from the app store not because Apple was immediate but because

00:21:08   the app was just collecting too much data or violating Apple's rules on how much data

00:21:12   you're allowed to harvest from your users. So this is an app that wasn't allowed on the

00:21:16   app store. The app itself by itself does sneaky gross things which is why Facebook is paying

00:21:22   people to install it because who else has any motivation to install a thing that's gonna

00:21:25   monitor every single thing you do and report it back to Facebook. And then on top of that,

00:21:31   the way they get, the whole thing is we have this app and we're told that it violates the

00:21:35   privacy rules for privacy or whatever. They don't come back and say is there a way that

00:21:39   we can learn what we want to learn about our customers without basically watching everything

00:21:45   that they do and violating their privacy. No, they say let's take the application as

00:21:50   it exists and find a way to distribute it. So it's a double violation. It's doing a bad

00:21:57   thing getting caught and then doing a bad thing by violating a different rule.

00:22:01   - Yeah, on so many levels this is like, it is inexcusable, it is horrendous behavior

00:22:07   by Facebook. They continue to be a completely appalling company. Like everything they do

00:22:15   is appalling and this is no exception to that. - So I do have some sympathy for the employees

00:22:21   who can't order their lunch though because in a large company, yeah, because in a large

00:22:25   company those people who just want to order their lunch probably didn't even know this

00:22:29   thing existed. Now at this point you could say well they didn't know this specific thing

00:22:33   existed but they knew the kind of company they work for which is true which is why I'm

00:22:36   not saying I have tons of sympathy for them but I have a small amount of sympathy. Maybe

00:22:41   not for the lunch ordering thing or whatever but like.

00:22:43   - No look, I will be very upfront with how I feel about this. I know people who work

00:22:48   at Facebook. I know good people who work at Facebook. I think if you work at Facebook

00:22:53   in 2019 you have decided to work for a company that is morally bankrupt and you've decided

00:23:01   that's gonna be okay with you. Now I'm not totally, my hands are not totally clean here.

00:23:06   I use Instagram. I like Instagram. Facebook owns Instagram. I don't feel good about this

00:23:11   right now. There's never been a day where I've considered it more than today. I don't

00:23:16   feel good about this but I think the company is abhorrent. I object to so much of what

00:23:21   they do and have done but I continue to use one of their products. So I recognize my hands

00:23:25   are not clean here but if you work at Facebook you have decided this company is horrible

00:23:32   but I'm going to work there anyway and I think if you make that decision you have to accept

00:23:38   the consequences that come with that.

00:23:41   - So speaking of, I mean this is not the first and only bad thing that Facebook has done

00:23:46   to say the least, right? In and of itself it might not be disqualifying or might not

00:23:53   make someone decide not to work there but there's definitely a pattern of behavior.

00:23:56   But speaking of pattern of behavior, another company that has some similar business models,

00:24:00   Google where they make a lot of their money by knowing information about people who use

00:24:03   their products, Google in fact had a very similar application that would monitor what

00:24:10   you did and report back to Google so they could learn more about their customers and

00:24:14   I believe, I don't know if they were paying people to install it but it was a similar

00:24:17   type of thing. They wanted people to install this and were trying to motivate them somehow

00:24:21   and they distributed this application under the Apple Developer Enterprise Program and

00:24:26   they issued a speedy statement after Facebook got hit with the certificate revocation. This

00:24:32   is what Google says. Their application was called Screenwise Meter. The Screenwise Meter

00:24:37   iOS app should not have operated under Apple's Developer Enterprise Program. This was a mistake

00:24:42   and we apologize. We have disabled this app on iOS devices. Very straightforward. We're

00:24:46   sorry. We're doing it too. We're sorry. We totally stopped. We're not doing it anymore.

00:24:49   Don't look at us. It's not our fault. Leave us alone. I bet Google has a lot of internal

00:24:53   applications signed by their enterprise certificate that they probably don't want to stop working

00:24:58   because they have a lot of employees and they probably have a lot of apps. The other angle

00:25:02   on this by the way is setting aside these companies' weird business models and their

00:25:08   questionable practices and so on and so forth. This came up in a couple of the slides we're

00:25:13   in. This reinforces the idea that Apple controls what gets installed on Apple devices and I'm

00:25:21   sure we're going to talk about this in a second about the appropriateness of Apple's punishment

00:25:25   of Facebook if it goes too severe or not severe enough. But there is a power dynamic between

00:25:32   all these companies here. One thing this can make companies think is, "Well, maybe we should

00:25:40   stop making all of our internal applications iOS applications or maybe we should stop making

00:25:45   iOS versions of our internal applications because it's a corporate risk." If the proper

00:25:51   functioning of our company relies on these applications continuing to work and some other

00:25:55   company that's not us and maybe isn't even friendly to us can at any time cripple our

00:26:00   entire enterprise by turning off all of our internal applications, maybe we shouldn't

00:26:05   make them on iOS anymore. Maybe we should make them on Android so we are masters of

00:26:10   our own destiny. Imagine if Microsoft could turn off all your copies of Office remotely

00:26:14   if for all I know they can these days because everything is subscription. That's one power

00:26:19   dynamic at play here. Who is in the driver's seat? Can Facebook actually "afford" to

00:26:26   not have iOS versions of our applications or would that make its employees revolt and

00:26:30   say, "I don't want to use an Android phone." What is the power balance? In a lot of these

00:26:35   situations, I don't think either party really knows who has the power. I guess we'll slide

00:26:43   into the obvious question here. Could Apple have kicked Facebook off of the App Store?

00:26:49   Who has the power in that situation? On the one hand, Apple certainly can. They control

00:26:53   those in the App Store and say, "Guess what Facebook? I'm treating you like a petulant

00:26:56   20 year old. You no longer have a developer account." Which is certainly what they would

00:26:59   have done to a petulant 20 year old. Have the petulant 20 year old been banned from

00:27:03   the App Store and then used an enterprise certificate or whatever.

00:27:06   Let's be clear. If anyone else did this, if it was anyone other than a giant, very influential

00:27:12   company, the entire developer account would have been banned. Facebook got off relatively

00:27:18   easily by only losing enterprise distribution rights. Even then, we don't know that this

00:27:23   could only be temporary.

00:27:24   And they'll probably get them back eventually through some negotiation because these are

00:27:28   giant stories. The same way that Netflix gets the better subscription rate. It makes everyone

00:27:34   feel good to think that Apple is treating everyone the same, but realistically speaking,

00:27:38   that's not how the world works and it's probably not even how the App Store should work. It

00:27:42   never has worked that way. It's an illusion. It makes some kind of sense.

00:27:46   Yeah, because ultimately Facebook has so much power here. If Apple actually removed the

00:27:52   Facebook app from the iPhone, who do you think is going to get hurt more by that?

00:27:57   That's the question. Both of these parties probably have a notion of who they think is

00:28:00   in the driver's seat. But I think the accepted wisdom has been Apple can't afford to not

00:28:05   have the Facebook app on its phone. That Facebook was in the power position. Because Apple can

00:28:12   huff and puff, but they need Facebook more than Facebook needs them. I think that's what

00:28:17   we've thought for a long time. But at a certain point, the iPhone installed base and the entrenched

00:28:24   love for the iPhone maybe starts to tip that the other way. If you ask the world of iPhone

00:28:30   users that you have a choice, you can either keep your iPhone but not have the Facebook

00:28:34   app or keep the Facebook app but you got to get an Android phone. How would that turn

00:28:39   out? I don't think either party wants to find out. It's like, "Who would win World War III?

00:28:43   What do you think?" It's like, "Let's not find out." Because no one's going to come

00:28:48   out 100%. There's no scenario in which 100% of iPhone users would stick with their iPhone.

00:28:55   It's like World War III. There's not going to be a victor who's unscathed. No one wants

00:28:59   to even try that. But the more I think about it, the more I'm thinking that I no longer

00:29:05   accept that it's a slam dunk that the iPhone can't survive without the Facebook app. Partially

00:29:11   because yes, people can still go to Facebook on the web but partially because of just the

00:29:15   year of bad news of Facebook. I don't know what Facebook's numbers are like or whatever.

00:29:19   All I know is that it has been relentlessly bad, deservedly bad news about Facebook. I

00:29:24   feel like that has to be having some kind of effect.

00:29:26   I think your World War III comparison is apt. I feel like this is kind of a mutually assured

00:29:34   destruction scenario here that neither company can really afford to totally blow off the

00:29:40   other one on their platform. That would be very bad for either company if Apple kicked

00:29:46   Facebook's app off the platform. By the way, again, Facebook owns multiple things. If Apple

00:29:52   actually revoked Facebook's entire developer ability, that would also include WhatsApp

00:29:57   and Instagram. That would have a massive effect. Think about every single news report that

00:30:03   would happen from this. Every single common, every single non-nerd out there would have

00:30:08   the exact same opinion. Apple took away Facebook from my phone. Apple took away Instagram from

00:30:13   my phone. Apple took away WhatsApp from my phone. They would 100% blame Apple for that

00:30:18   and not in a good light. Apple can't afford that. They can't do that. I think ultimately

00:30:27   Facebook holds the power in this relationship because if Apple were to ever remove Facebook's

00:30:32   ability to ship their apps, all of the blowback would fall on Apple, not Facebook, from their

00:30:37   actual customers.

00:30:38   But I feel like at a certain point, because we all accept that as probably the way things

00:30:42   are and I think Apple accepts it, at a certain point, Apple could find itself in a position

00:30:47   where that is way less true than it used to be and they just don't realize it. This happened

00:30:51   with web browsers a couple of times too where the power shifts but nobody noticed because

00:30:56   the conventional wisdom is just so strong, like the conventional wisdom that IE is dominant

00:31:00   and then eventually you wake up one day and it's like, "What the hell happened to IE?"

00:31:05   You don't adjust your thinking.

00:31:07   On the flip side of that, if the conventional wisdom still is true, which it probably is,

00:31:13   if Facebook wants its enterprise certificates back, it's just like World War III, they can

00:31:17   say, "Apple, you should probably give us enterprise certificates back or we're going to pull the

00:31:22   Instagram and Facebook and WhatsApp apps from the iOS App Store," because Facebook can unilaterally

00:31:27   pull its own apps from the iOS App Store. They can use some or all of the power they

00:31:32   have or hold it as a threat to try to get what they want back, which is why in any kind

00:31:36   of flare-up of PR and yanking the certificate or whatever, these two behemoths will probably

00:31:44   talk to each other over the phone or in person and work something out, billionaire to billionaire

00:31:53   and come to an understanding.

00:31:54   In the same way as the Cold War did, no one wants to launch any of the nukes. Everyone's

00:31:59   got enough nukes to destroy each other ten times over. No one thinks they're ever going

00:32:04   to come at anything like that unscathed. They all just, in the end, just want to go back

00:32:07   to making money. So I feel like they will come to some kind of agreement. Facebook will

00:32:11   get its enterprise certificate back. They will not pull their things from the store

00:32:14   and we will just continue to lurch forward.

00:32:16   I don't know offhand how granular a big red button Apple has. So said differently, could

00:32:26   Apple have turned off just this one app?

00:32:28   Yeah, they can.

00:32:29   Yeah, that's what I was gonna say. I suspect so, but I don't know. And on the one side,

00:32:34   I feel like isn't that the rightest, most mature answer is to turn off this one app

00:32:39   and leave everything else alone?

00:32:41   Well, no, because they violated the terms of enterprise distribution.

00:32:45   Yeah, the terms are for the certificate.

00:32:48   And they were stupid enough to use their main enterprise distribution certificate for this

00:32:53   app.

00:32:55   But either way, it seems to me like I don't think that Apple turning off this app, whatever

00:33:03   it's called now, it doesn't matter, turning off this sleazy VPN app...

00:33:06   Oh, Notvo. It's not Ohnava, we swear.

00:33:09   Yeah, exactly right. Fohnava. Instagram. Anyway, turning off the Fohnava app won't do anything.

00:33:18   Like, it won't accomplish anything. Facebook's just gonna go trudging along because from

00:33:22   an outsider's point of view, if there is anyone that has even more hubris than Apple, it's

00:33:29   Facebook. Facebook seems to be pretty self-obsessed. And it seems to an outsider's perspective

00:33:37   that Facebook thinks that Facebook can do no wrong. And so I think that I'm okay with

00:33:44   Apple just completely turning off the enterprise certificate because it seems like to me, it

00:33:50   seems like that punishment is commensurate with the crime committed.

00:33:53   "Crime" isn't the right word, but you know what I mean. It just seems really frickin'

00:33:57   gross. The offense. Yeah, thank you. That's a very good word for it. And it just, I don't

00:34:03   know. I think I'm okay with this. And like one of you said earlier, I think it was John,

00:34:08   you know, some of the slacks that we're in, we're going back and forth about whether or

00:34:11   not this was fair and whether or not this is the right answer and whether or not Apple

00:34:15   should have that kind of control. And I see the argument there that maybe Apple shouldn't

00:34:20   be able to turn off, to some degree, the livelihood of another corporation. But I don't know,

00:34:26   man, it's their swimming pool. And if you're choosing to swim in that pool, shouldn't you

00:34:30   have to abide by the rules? I mean, if it's adult swim, shouldn't the kids have to get

00:34:33   out of the pool? Yeah, I don't think anyone's arguing that.

00:34:35   They were just getting back to the root thing that's been with the App Store the entire

00:34:39   time that we don't like these rules. We gotta choke them down because being in the App Store

00:34:45   is better than not being in the App Store. But the idea that there's a platform that's

00:34:49   so tightly locked down with no sideloading at all, and even the sideloading that is allowed

00:34:54   to enter our certificates is still locked down to, you know, like, I remember before

00:34:59   Test Flight and everything, people were just so angry about the fact that they couldn't

00:35:02   even just make betas and give it to people. Like, that is a developer experience question

00:35:06   that has been with the App Store since its introduction and probably will never leave.

00:35:12   It just comes up again in this context because this is a fairly high-profile example of Apple

00:35:17   flexing those muscles. Like, they have the ability to do this. Everyone knows they always

00:35:22   have it, but it's mostly academic until you see, you know, the, uh, oh god, I can't, I'm

00:35:27   getting so old, people. I'm getting so old and I'm just thinking about my bad day in

00:35:30   Destiny. Now I witness the power of this fully operational App Store. Close. Oh, wow. Wow.

00:35:37   I'm surprised that didn't roll right off your tongue.

00:35:41   We are sponsored this week by Marine Layer. For 15% off your first order, visit MarineLayer.com

00:35:46   at our promo code ATP15 at checkout. Marine Layer makes really nice and incredibly soft

00:35:53   clothing. When they say their stuff is soft, they really mean it. It's not just like, oh,

00:35:58   it feels nice soft. It's like, oh my, how did they even do that kind of soft? It turns

00:36:03   out the softest tees are made from trees. Micro Modal is found in Marine Layer's signature

00:36:08   fabric and it's made from recycled beechwood. So the pulp production for these t-shirts

00:36:13   is self-sufficient, which makes their tees sustainable, eco-friendly and incredibly soft.

00:36:20   They also make in between sizes like between medium and large is Marge or between large

00:36:25   and XL is larger for people who just don't conform to exactly the same little small number

00:36:30   of standards. And the return policy is insanely good. You can return pretty much anything

00:36:35   for up to a year. They really stand by their clothes and there's no risk to you. Plus they

00:36:40   have for free shipping and free returns on all US orders. So I got to say I have some

00:36:45   Marine Layer shirts and they really are remarkably soft. I've had them now for, I mean, they've

00:36:51   gone through probably now five or six washes each at least, and they're still just as soft

00:36:55   as the day I got them. They're really nice. I'm very happy with them. So check it out

00:36:59   today at Marine layer.com and for 15% off your first order, use promo code ATP15 at checkout.

00:37:08   Once again, 15% off your first order of these amazingly soft shirts, visit Marine layer.com

00:37:13   and use promo code ATP15 at checkout. Thank you so much to Marine layer for sponsoring

00:37:18   our show.

00:37:23   Last night I tested this thing that let me spy on Erin. Now granted she was sitting three

00:37:28   feet away from me and I told her exactly what I was doing, but Ooh, this FaceTime thing,

00:37:33   not good, not good at all. So if you start a FaceTime video call with another person

00:37:42   and then you do the little swipe up thing in order to add a person to that call. So

00:37:46   say I was starting to call Marco and then I wanted to add John to the call. Well, instead,

00:37:51   what I do is I call Marco, I swipe up and instead of adding John, I add myself. If you

00:37:57   do that, suddenly Marco's phone with no intervention on Marco's part starts broadcasting Marco's

00:38:05   microphone to me without Marco's permission. Then if Marco thinks to himself, Hmm, I know

00:38:11   how to make this stop. I will hit the side button, whatever they call that, the lock

00:38:14   button, the power button. I don't care. I'll hit the side button to silence this ring.

00:38:18   I think it actually is called the side button now. Oh, is it? Okay. Um, and so I'll silence

00:38:23   this ring and make it go away. Well, guess what? Now I'm looking at Marco's front facing

00:38:27   camera without Marco's permission. And I tried both of these with Aaron. I, and honest to

00:38:33   God, both of these worked like, well, they don't work if you know what I mean. But like

00:38:37   these exploits accomplished what they set out to to accomplish. And Ooh, that did not

00:38:43   make me happy. And it went around the internet fast and it makes sense because people were

00:38:49   turning off FaceTime left and right until Apple distressingly slowly eventually turned

00:38:55   off group FaceTime at the server level, which is the right thing to do. But it took from

00:38:59   what I could tell several hours for that to happen, which I'm not keen on. So that bug

00:39:05   is bad in and of itself. And I'll leave a chance to talk about that. But what's worse

00:39:09   is apparently Apple has known about this for at least a week, which is really gross. But

00:39:14   we'll talk about that in a minute. Gentlemen, thoughts about this bug and what it means.

00:39:19   I mean, look, everybody ships bugs. Like I said, this on Twitter, I'll say it here again.

00:39:22   Like I don't blame Apple for this bug existing. Like everyone ships bugs. What's important

00:39:30   is how you respond when you have accidentally shipped a bug. All right. That's how that

00:39:34   and it's when you have something that is a very severe security bug like this, one of

00:39:39   the key elements is speed of your response. And on one level they did a really good job

00:39:47   actually. Like the bug hit the press, you know, and then like within a couple hours

00:39:52   of hitting the press, all group FaceTime was disabled server side, which effectively neutralizes

00:39:57   the bug and makes it impossible to exploit again. So that alone, like that in itself

00:40:03   was an adequate response. Like that was fine. Yeah, sure, I would have liked if it was less

00:40:07   than a few hours, but it was hell, it was later that night. Like it was fine. Like that's

00:40:11   fine. I give them full marks for that. But there is this other problem, which is a pretty

00:40:16   big problem, which is that this was, this was reported to them like a week earlier through

00:40:21   the official channel, through like security at apple.com or whatever it is, product security.

00:40:25   Like they have an official channel to report security flaws in their products. And it was

00:40:29   reported almost a week earlier and it was fully documented. And apparently the woman

00:40:36   who reported it was told to file a radar. So she did, even though this was a security

00:40:41   incident, this reminds me of the movie Die Hard, which got Mark, have you seen Die Hard

00:40:47   please? Yes, it was my number one Christmas movie. We did it. We did it everybody. I just

00:40:53   showed it to my son recently and there's a scene in Die Hard where I think it's Bruce

00:40:57   Willis gets finally manages to call like emergency services or whatever and it gets on line is

00:41:02   telling them like, you know, the building is under attack, whatever you got to come

00:41:05   send the police or whatever. And the person on the other end of the line is like, this

00:41:09   is a restricted channel. If you have an emergency, please call 911. Essentially saying, this

00:41:13   is not the right way for you to report an incredibly severe security bug. File a radar.

00:41:19   So Bruce Willis was having none of that. He's yelling and screaming. But this person, this

00:41:22   poor person was like, I have to get a, I have to file it with a bug report. Like I'm talking

00:41:27   to you now, you're apple, right? Like, look, this is severe. And like, nope, sorry. Wrong

00:41:31   channel. Like this is not how you report bugs to us. Please do it through the proper channels.

00:41:35   So the person did do it through the proper channels. But I was thinking the same way

00:41:38   of like, if someone comes to you and is telling you there's an emergency, and your only reaction

00:41:43   is to say, I'm not the person you're supposed to tell about emergencies. I'm not going to

00:41:48   take any action on this. I'm just going to redirect you to the right channel. And then

00:41:51   you go to the right channel, which is, you know, this signing up for an account and going

00:41:55   through and filing radar or whatever. And still, it doesn't get to the right people.

00:41:59   It's this is sort of the faceless corporation that we always complain about that there that

00:42:05   we think of a company as this giant entity, but it's really this big group of people with

00:42:11   very different incentives and motivations. There is no one consciousness that is Apple

00:42:17   that you can talk to. So trying to get the cooperation writ large, to be aware of a thing

00:42:22   in the world is very difficult when you are interacting with even the correct one or two

00:42:29   people who are touching the outside world, which is why when things go to the press,

00:42:33   the giant entity that is Apple becomes all of a sudden very conscious. And how does it

00:42:38   work? Maybe the press is seen by people at the top, and the people at the top make everyone

00:42:44   below them aware of it. But if you start it through the right channel, it's dysfunctional

00:42:48   and baffling, but not actually a very rare occurrence that you can present what you would

00:42:58   imagine is incredibly important critical information. And because it is presented alongside tons

00:43:05   and tons of other information that's just garbage, that it gets lost in the shuffle.

00:43:11   You can't distinguish it from the 800 other cranks who emailed you that day and said,

00:43:14   "I have a way to secretly destroy every iPhone by pressing a big red button." Tons of garbage

00:43:20   comes into that same channel, and here is this tiny little gem that is a legit super

00:43:24   serious security bug. In hindsight, we all say, "Why didn't they jump into action on

00:43:29   this one big thing?" And it's like, "Well, you didn't see the 8,000 other radars that

00:43:32   were filed that day, and you didn't have the millions of other things that got into that."

00:43:36   So it's frustrating from the outside, and it's explicable on the inside, but it is a

00:43:39   dysfunction. You want to design an organization such that if some well-meaning person is nice

00:43:48   enough to tell you about a super-duper important problem, that that finds its way in less than

00:43:55   a week to the collective consciousness of the company and that they take action on it.

00:44:00   And you don't have to wait for it to go all over the press and come top-down through the

00:44:03   executive suit and just make all their underlings have a giant fire drill about it.

00:44:10   So first of all, I was about to insult the whole bug reporter radar system, but the funny

00:44:16   thing is that wasn't the right channel for this. The right channel for this was product

00:44:20   security. She went there first, and they told her, you know, a buzz off, basically.

00:44:24   It sounds like some weird bug. You probably don't know how to use computers to file a

00:44:28   radar. Yeah, right. So she went to the proper channel

00:44:31   first, and they told her to go to the incorrect channel. Now, the incorrect channel she was

00:44:35   directed to, which is the bug reporter/radar system, is horrible. The pipeline, whatever

00:44:42   the process is that Apple uses to filter and respond to and close those bugs that get reported

00:44:51   is horrible. Any developer who's ever filed bugs will tell you that most of the time it's

00:44:58   a waste of your time to even bother with it. Most of the time you don't get a response

00:45:02   at all. If you do get a response, it's basically them asking you to provide a simple project

00:45:07   and syst-diagnose files. Both of which take a lot of time to gather and prepare. Both

00:45:12   of which are usually not necessary for lots of the bugs that you report, and it seems

00:45:18   like they request those things before they've actually really read what you wrote. It seems

00:45:24   like Apple has, oh, and then every time there's a new beta released of anything, they will

00:45:31   seemingly go through in bulk to all the open bugs that haven't been responded to yet and

00:45:36   say, let us know if this is still a problem or we'll automatically close it in like a

00:45:40   day. And what this leads to is a whole bunch of legitimate bugs either going back to the

00:45:48   filer and asking for them to spend way more time on it when that wasn't necessarily, like

00:45:53   when Apple didn't really pay enough attention to it to see whether it needed that or not,

00:45:58   or a whole bunch of valid bugs being closed that are still bugs but that the person didn't

00:46:04   get to in time to say, yes, it's still a problem. And so you have Apple over-filtering the inputs

00:46:10   on that system dramatically. It seems, I mean, I don't know how this works internally, but

00:46:15   it sure seems like there are people whose job it is to just close as many bugs as possible

00:46:21   as like a performance metric and not anything about the actual outcomes of those things.

00:46:27   So just, they find, they seem to find ways to just close them in mass, if you're lucky

00:46:32   that they even close them at all. Most of the bugs that I report are just left open

00:46:35   forever with, they never get a single response. But some of them do get these wonderful like

00:46:39   mass closures that happen to them that, you know, the bugs are still there and nothing

00:46:44   ever happened. And this is why when Apple employees tell us on Twitter, file a radar,

00:46:50   it really helps, it really matters. I don't appreciate that very much because filing a

00:46:56   radar from the outside, from our point of view, is a massive waste of our time, 99%

00:47:01   of the time. Like the vast majority of bugs you file are a total waste of your time. And

00:47:06   the way Apple deals with them and the people who file them, the message it sends us from

00:47:12   the parts of it that we see, which apparently is very little of it, the parts of it that

00:47:15   we see tell us we are wasting our time. Whether that's true or not, internally we can't

00:47:21   see. But the part we see is you don't care about us, you are dismissive if you respond

00:47:26   at all, and we are wasting our time doing this. So that's just a little rant about radar.

00:47:32   But the fact is, for this story, the woman who was reporting this bug shouldn't have

00:47:36   even been sent to radar in the first place because this wasn't that kind of bug, it

00:47:39   was a product security thing and she emailed product security.

00:47:42   Steven: Yeah, this makes me so angry that she was told, "Oh, just go file a radar."

00:47:47   That is such a, it's such an--

00:47:49   Steven (interrupting)

00:47:50   It's dismissive and it shows that they weren't really reading things very closely. Exactly

00:47:54   the same problem that happens with the radar bug screeners. It's exactly the same. It's

00:47:57   like you ask us to file bugs, you ask people to report security problems, you ask us to

00:48:01   do this, and those of us who take our time, and this doesn't take a small amount of time

00:48:06   to do this kind of stuff, those of us who take our time to do it more often than not

00:48:10   get dismissed in a way that suggests they didn't even read it.

00:48:13   Steven (interrupting)

00:48:14   Right, this is such a big "f*** you" to this poor lady who took the time to actually

00:48:20   do this the right way. Rather than just find-- and my understanding is, you know, there's

00:48:24   a series of tweets from a person with the Twitter handle @Beastmode, which I am quite

00:48:30   impressed by. Anyway, John H. Meyer, who I guess got in touch with the woman whose child

00:48:37   initially discovered the problem, and she apparently made a video of it, she sent like

00:48:42   a whole write-up to Apple about it, and the response from them was, "Oh, go file a radar."

00:48:47   How off-putting is that? How obnoxious and off-putting is that? It makes my blood boil,

00:48:53   it's so ridiculous. It's just so obnoxious, and it makes me so angry.

00:48:58   John H. Meyer, MD So the more charitable interpretation is not

00:49:00   that there's any malice or dismissiveness there, but that the volume of stuff that comes

00:49:05   in is such that they are overwhelmed with information, and their sort of first-pass

00:49:10   filtering algorithm identifies this as probably just some person who has some kind of bug

00:49:14   they want to report. And this person is the mom who reported this is a lawyer, and if

00:49:21   you look at the letter, it's not lawyer-ese, but like, I mean, maybe if you saw something

00:49:25   that was from a lawyer or looked like it was written by a lawyer, like, maybe that doesn't,

00:49:30   you know, like, when you get a lot of input, when you get a lot of-- Apple sure gets lots

00:49:33   of people telling them all sorts of things. You have to have some sort of way of filtering

00:49:38   it. And in the letter, I think the person asked about like bug bounties, like a lot

00:49:43   of companies pay people to find bugs. And I think one of the benefits of bug bounties,

00:49:48   which Apple has, by the way, I think they just recently started a bunch of them, I think

00:49:50   we were talking about an ATP a while back, is that if you have a program where you're

00:49:55   going to pay people for finding bugs, I think the people at the other end of that fire hose

00:50:03   of incoming information probably have slightly different incentives, because they're, like,

00:50:08   the whole point of that program is to find the super valuable stuff, whereas the product

00:50:12   security one probably gets way more, like, just random people with opinions about Apple

00:50:17   stuff or, like, people who think that their phone is spying on them at night because it

00:50:23   wakes up on a notification call. Like, who knows what kind of stuff they get? But the

00:50:26   bug bounty ones, the only people who even know that that channel exists are people who

00:50:30   are out there looking for bugs or think they found some super big security bug. So everything

00:50:34   that comes in there, in theory, is like the sender thinks it's a super important security

00:50:39   thing, whereas product security probably gets just, you know, who knows what kind of random

00:50:43   stuff. So it's hard to tell from the outside whether it actually is sort of jaded dismissiveness

00:50:48   or if it's we are not good at dealing with the volume and type of information that we

00:50:54   have, or understaffed, we don't have good heuristics for figuring out, we have tier

00:50:58   one, two, and three, I would get this impression like this, tier one, two, and three radar

00:51:01   stuff before it ever even gets to a developer, it has to go through this sort of, kind of

00:51:05   like the App Store used to be and probably still is to some degree, like a mechanical

00:51:09   rules-based thing implemented by people who don't have any awareness and don't need

00:51:14   to, by design, don't need to have any awareness of security or App Store rules or anything

00:51:19   like that, and are just doing sort of a first pass, following whatever rules they've been

00:51:23   told. And eventually you get down to a human who understands this is an application, this

00:51:27   is for the App Store, this is what security is like, whatever. But, you know, as with

00:51:31   so many things from the outside, we can't tell which one of those things it is, so it's

00:51:35   just a black box that does the wrong thing.

00:51:37   It's funny to me too, because all of us have friends within Apple and I'll talk

00:51:42   to them about radar and they'll try to say, "Well, you know, oh, you don't understand

00:51:47   the volume," which I think you're right, Jon, like that's true, and, "Oh, you know,

00:51:50   there's this and there's that and there's this and there's that," and it just makes

00:51:53   us sad that you guys always assume the worst of us. And that also kind of grinds my, I

00:51:57   don't know why I'm so angry tonight, but that kind of grinds my gears too, because

00:52:01   if you don't want us to think the worst of you, then talk to us. Tell us what's going

00:52:08   on. Like, this is your fault. Not my fault. It's not my fault I'm assuming the worst.

00:52:13   This is your fault for being so secretive. You can't have it both ways. You can't

00:52:18   be so secretive about what's going on inside and you can't be so just a black box. You

00:52:24   can't be such a black box and expect us to just assume it's roses and pansies and

00:52:28   daffodils inside. Like, this is a self-created problem.

00:52:31   Right. It's like Apple is saying, like, "You keep mispronouncing my name," and

00:52:35   we're like, "All right, how do you pronounce it?" And they're like, "Ugh, I told you

00:52:37   before."

00:52:38   That's not what it's like at all, actually. They don't tell us, they would never have

00:52:47   told us. I mean, so here's the thing, like, we want more transparency, and Apple, I think,

00:52:51   has been more transparent than it has been in the past.

00:52:53   Agreed, agreed.

00:52:54   But for a large company like this, you don't want complete transparency, because that's

00:52:57   just chaos. Like, the company does have to speak. Like, there's a reason PR exists.

00:53:01   You can't have every employee in the company purporting to speak for the company or even

00:53:05   for their little section of the company, because it's madness, right? But Apple still, when

00:53:10   it comes time to speak with one voice about this type of thing, does probably the smart

00:53:16   thing from a PR perspective, but that nevertheless leaves us in the dark, which is, "We're

00:53:21   not here to explain to you exactly what happened here." Like, in this whole thing, they said,

00:53:27   we had their statement, and you read part of it, like, "Google or Facebook did this

00:53:30   thing with their certificates, it's against the rules, so we revoked their certificate,"

00:53:33   blah, blah, blah. That's what they said. They didn't go into this whole sort of, like, post-mortem

00:53:37   of like, "Oh, how do we miss this?" They didn't mention this person and her son and finding

00:53:42   the security thing a week ago. They didn't even acknowledge that even exists as far as

00:53:46   I know. Like, why would they? There is no, like, that's not as far, it's like, we're

00:53:50   going forward. Like, it's the smart PR thing to do to say, "Here's the situation, here's

00:53:54   what we did," boom. Not to say, "Oh, how do we miss this?" And have some sort of back

00:53:58   and forth with the press about scolding Apple about how they should have found this and

00:54:02   soul searching about how they, like, they'll do that internally, but they will never do

00:54:05   that externally. And so, because of that, which I think is probably mostly the right

00:54:10   thing to do, we are left with just, you know, an unknown, and then we're free to map all

00:54:16   of our past experiences and prejudices onto it and think to the best of the worst about

00:54:20   what's happening. But, I mean, honestly, it doesn't really matter whether it was malice

00:54:27   or, you know, understaffing or whatever. The end result is the same as far as the company's

00:54:33   concerned. They were not appropriately reactive to a very serious, very cleanly well-presented,

00:54:42   helpful thing from the outside. So, it's not like an edge case. They really need to

00:54:46   make systems that, in the best case scenario, where a conscientious, motivated person who's

00:54:51   found a legit bug does everything you tell them to do, does all the right things, and

00:54:55   still can't get you to pay attention for a week.

00:54:57   Kind of makes you wonder what else they're missing.

00:54:59   Yeah, well, those are the, I mean, that's the best case. The worst case scenario is

00:55:03   people find tons of bugs and never tell Apple about them, because why the hell would you

00:55:06   tell Apple? They'd either use them themselves or sell them on the black market. Like, that's,

00:55:09   you know, that's why bug bounties exist, where you try to, like, pay more money than

00:55:13   the bad people are going to pay to find out your bugs.

00:55:16   So maybe we can talk about something a little happier. Just a couple hours before we recorded,

00:55:22   Mark Gurman had a scoop at Bloomberg where he had a bunch of information about the next

00:55:27   iPhone, about iPads, and about iOS 13. And it's hard for me to really get too anxious

00:55:35   about this because, or maybe anxious isn't the right word, excited, I guess, about this,

00:55:40   because it seems like Bloomberg can get some details right and a lot of things wrong. And

00:55:44   I think Marco had summarized this best in saying, you know, he'll get the little specific

00:55:48   things right, but then they'll try to put it in a narrative and it all falls apart at

00:55:51   that point. But there's an article on Bloomberg, we'll link it in the show notes, where Gurman

00:55:59   talks about all of this stuff. And one of the big features, apparently, of the iPhone

00:56:04   vNext, I don't know if it's, you know, XT or whatever, is very much camera related.

00:56:12   And there was no information about any sort of knockoff of Google's seemingly very impressive

00:56:17   night shift. But apparently, Apple's starting to really angle toward, I shouldn't say starting

00:56:22   to, is really angling toward augmented reality. And so what's being said in this article is

00:56:27   that there's going to be a more powerful, longer range 3D camera, which is "designed

00:56:33   to scan the environment to create three-dimensional reconstructions of the real world." Supposedly,

00:56:38   it's going to work up to about 15 feet away from the phone. It uses a laser scanner instead

00:56:43   of dot projection, because the dot projection, you know, goes to crap over long distances.

00:56:48   And it got really, I don't think Gurman's a very good writer. It was very unclear to

00:56:53   me what is or is not getting the third camera, because early in the article, he says there's

00:56:59   a third, more advanced camera. Then later on, he says the third camera is only in the

00:57:04   Max. And then immediately after says, "And other handsets could eventually come with

00:57:07   the upgraded system." So...

00:57:08   - Well, that's not, I don't think that's like, you know, a mistake. I think that's hedging.

00:57:13   Anytime you see a rumor article like this, especially one from a publication that's,

00:57:18   you know, edited like Bloomberg, not fact checked like Bloomberg, but edited, it's generally

00:57:26   a result of they don't actually know for sure. And so they're offering, it's like, "This

00:57:32   could come out this year or in the future." Like whenever they, you know, they write stuff

00:57:35   like that. It's rumor articles that have, that they literally heard a rumor. They don't

00:57:40   really, it's not really firm knowledge or, you know, they don't want to commit. So they

00:57:45   use flexible language to give themselves an out.

00:57:49   - Yeah.

00:57:50   - So that, yeah, basically anything that is not very clearly stated and unambiguously

00:57:55   and unequivocally stated in an article like this, you can assume they're just guessing.

00:57:59   - And on the night sight thing, you said night shift before where you meant night sight.

00:58:02   - Oh yes, thank you.

00:58:03   - There is a thing about that with the multiple cameras in terms of like catching a wider

00:58:07   field of view and more pixels, whatever. But for something like a night sight type feature,

00:58:13   that is probably, I mean, there is a component of that that deals with the cameras, but it's

00:58:17   very much a software thing. And the software is not going to leak from these same channels.

00:58:21   Like that's the thing that you have the most difficulty finding out because that is just

00:58:24   kept in Apple, you know, in Apple proper. It's not, there's no manufacturing or parts

00:58:29   or any other places where that can leak. So we tend not to know as much about the software

00:58:34   and what we do know comes much later. So it doesn't surprise me there's no mention of

00:58:37   a predominantly software only feature. All they have to go on is probably a bunch of

00:58:41   parts leaks or drawings or whatever of phones with a bunch of different cameras on the back

00:58:46   and information about them. So I am still optimistic that Apple will copy that feature

00:58:53   to some degree and they'll do it in software and we'll find out about it like, you know,

00:58:59   the day before the keynote or the keynote.

00:59:00   Yeah, we'll see. I sure hope so. But it was interesting. They also had a couple other

00:59:06   tidbits. The third camera, which may or may not only be on the max, has a larger field

00:59:11   of view and a wider range of zoom. It will also, and this is now a direct quote, it will

00:59:15   also capture more pixels so Apple software could, for example, automatically repair a

00:59:19   photo or video to fit in a subject that may have been accidentally cut from the initial

00:59:23   shot. Finally, enhanced live photos, which they say is basically just doubling the time

00:59:27   from three seconds to six seconds. There will be updates for the XS, XS Max, and XR. And

00:59:35   then this is another thing that seemed contradictory to me, but maybe my reading comprehension

00:59:39   is bad. They said that the laser powered 3D thing could debut in an upgrade to the iPad

00:59:44   Pro in spring 2020, even though they started talking about it in the iPhone, whatever.

00:59:49   They also apparently are testing, Apple is testing an iPhone with USB-C, which I think

00:59:54   I'm pretty excited by. I think I would like to go all USB-C now.

00:59:57   Well, see again, that's not information. They are testing it. Well, of course they're testing

01:00:02   it. Like that doesn't mean anything. Chances are by now, I would assume based on like the

01:00:07   iPhone production scale and everything, they have probably already decided whether the

01:00:11   2019 iPhones will be USB-C or Lightning. And so by Grumman saying that they're still testing

01:00:17   it, like that just means he doesn't know yet.

01:00:20   The iPad thing actually makes some amount of sense to me because this is kind of a coincidence

01:00:24   that yesterday in a different cycle, I forget about the context that we were talking about.

01:00:28   I forget that we were talking about it. I made some joke about just wait until you get

01:00:32   your Apple AR glasses. You'll just walk around the house and look at all the rooms and it

01:00:35   will make a 3D map of your house. I think it was you Casey asking about like CAD application

01:00:38   or talk about linear, making floor plans and stuff. And so here we are with this rumor

01:00:43   a day later about basically a way to get a 3D image of the stuff around you at a longer

01:00:50   distance at basically room distance, like 15 feet-ish or whatever, rather than face

01:00:54   ID distance, which is a foot or two.

01:00:58   On an iPad, you're like, well, iPads, people don't carry around iPads as cameras for the

01:01:02   most part. They usually have their phones with them, despite the fact that we see people

01:01:04   taking pictures of iPads. But if you wanted to do some kind of AR thing and scan your

01:01:09   whole room and then manipulate it, an iPad is a perfect device. You're in your house

01:01:12   already. You've got your iPad. Hold it up, show it all the walls of your house, and then

01:01:16   you have a nice big screen on which you can move your furniture around in AR or whatever.

01:01:21   It makes sense as an iPad thing. Not that it gives me any more stake in any of these

01:01:27   rumors because they take every side of everything, even on the USB-C thing. They're testing a

01:01:31   phone with the USB-C, but they might not include it in this phone.

01:01:34   So you've offered no information. USB-C could be there, but it might not. Great, thanks.

01:01:38   I mean, VGA could be there. It might not.

01:01:41   Right. Well, it's not that bad. But anyway, the iPad Pro thing, I would actually be excited

01:01:47   about an iPad Pro that had the laser camera thing on it because the iPad, in many contexts,

01:01:53   is a better device for AR because it has a bigger screen for you to, you know, a bigger

01:01:58   window into the world to see through than the tiny little porthole that is your phone.

01:02:04   Even if you have one of the really big phones. So I think that would be neat. Although it

01:02:07   coming out ahead of the phone is a little bit weird. Like, that's what they were saying.

01:02:11   Like, that they would ship new iPad Pros with this new camera. It seemed like it would steal

01:02:17   a little bit of the iPhone's thunder because they would demo this camera. "Look, it's got

01:02:20   lasers and it does all these things. Isn't it amazing?" And then when the new phones

01:02:23   came out, they're like, "Yeah, it's got that camera too." It's kind of underwhelming. So

01:02:27   I don't know how they would deal with timing of that, but I fully endorse an iPad with

01:02:33   this 3D scanning laser thing on the back of it.

01:02:35   Yeah, me too. I was just confused because it seemed like early in the article they were

01:02:39   saying, "It's coming to the phone next!" and "Oh wait, it's coming to the iPad next." Alright,

01:02:43   weird. Whatever.

01:02:44   Look, they just got a bucket of parts, Casey. They just take a bucket of parts, you dump

01:02:48   it on the table, like, "Can you make an article out of this?" Like, "Yeah, I can give it a

01:02:51   shot."

01:02:52   Why not? They also had some other interesting news. Apparently there's going to be an updated

01:02:56   lower-cost iPad with a 10-inch screen and lightning. And what would have been of interest

01:03:01   to me, although now I think I've converted myself to the large-ish iPad world, a new

01:03:06   cheaper iPad Mini. And I was a devout Mini fan up until I got this brand new iPad a couple

01:03:11   of months back, and now I really love this iPad. But a new iPad Mini sounds very cool.

01:03:17   I would love to see that come back.

01:03:18   And then finally, iOS 13 apparently will have dark mode. It will have improvements to CarPlay.

01:03:23   I actually am a fan of CarPlay. I think CarPlay's pretty good. It's not fantastic, but it's

01:03:27   pretty good. And I'm curious to see what "improvements" means. Apparently there's going to be a new

01:03:34   iPad-specific home screen. So I don't know if that means -- I think Steve Trout and Smith

01:03:37   kind of implied this. I'm not sure if that means that instead of Springboard, there would

01:03:42   be some equivalent to Springboard that is not the exact same, you know, some different

01:03:46   home screen app, if you will. I think that could be very cool, because certainly the

01:03:50   home screen on the iPad seems like a place that it could be treated very differently

01:03:54   and a lot more done with it. And then a couple of direct quotes. "Ability to tab through

01:03:59   multiple versions of a single app like pages in a web browser." That's not something I

01:04:04   feel like I need right now, but I bet you once I have it, I'll say, "Oh, this is amazing.

01:04:07   I can't believe I didn't have it before."

01:04:09   What do you think that means, before we move on from that? Like, again, with the vague

01:04:13   writing. "Ability to tab through multiple versions of a single app like pages in a web

01:04:18   browser." There's a lot going on in that sentence, right? So some people may read that and think,

01:04:23   "Do they mean, like, Command-Tab?" Because you don't use Command-Tab on a Mac to go through

01:04:28   multiple versions of a single app like pages in a web browser. You do Command-Tilde for

01:04:34   multiple windows, and it's a different keystroke, usually Command-Shift-square-bracket or whatever,

01:04:38   to go through multiple tabs. Or do they mean tab through multiple versions of a single

01:04:42   app as in tabbed windows? Where there's a really hard way if you wanted to have multiple

01:04:47   versions of an application running, like, the tabbing would be the sort of the windowing

01:04:52   interface, kind of like Safari tabs. It's so confusing. It's basically a sentence that

01:04:57   says almost nothing other than the fact that there would be some new way to handle multiple

01:05:01   versions of something that there are not currently multiple versions of. I don't know. How do

01:05:06   you guys interpret tab? Do they mean the tab key as an analogy to what you do on a Mac,

01:05:10   or do they mean tabs as in Safari tabs?

01:05:12   Steven McLaughlin I think tabs as in Safari tabs. I think the

01:05:16   idea here is that if you wanted to have multiple copies of Google Docs running at the same

01:05:21   time, you could do that. I'm not sure the exact mechanism, but something like a tabbed

01:05:26   Safari experience. Or actually in Safari on the iPad, you can create two side-by-side

01:05:33   Safari quote-unquote "windows," if you will, and they're each their own tab, and you're

01:05:37   looking at two tabs or two windows simultaneously. I think something like that—god, this terminology

01:05:42   is terrible—but it's something like that, I think.

01:05:45   Eric Meyer The vocabulary reflects the mess that is the

01:05:47   actual UI.

01:05:48   Steven McLaughlin Yeah, well put. But I think it's something

01:05:51   where basically you could have two copies of the same app open at the same time.

01:05:54   Eric Meyer Just like Windows.

01:05:56   Steven McLaughlin Yeah. I don't know, Marco, any thoughts

01:06:01   on that?

01:06:02   Eric Meyer No.

01:06:03   Steven McLaughlin All right. And then finally, improvements

01:06:04   to file management, which would be great.

01:06:05   Eric Meyer Which is as vague as you can get without saying

01:06:07   nothing, literally nothing.

01:06:08   Steven McLaughlin Yeah. But everyone wants it. I was just thinking

01:06:11   today about how I feel like if I had file system access in a couple places, my iPad

01:06:16   could really potentially replace my computer for anything but—

01:06:19   Eric Meyer Would you like that file system access to be

01:06:20   improved, Casey?

01:06:21   Steven McLaughlin Yeah, it would be great. I would love that.

01:06:24   Eric Meyer The most exciting thing I think here, like buried

01:06:27   at the bottom of this article—I read this whole article, and I'm like, the very end

01:06:31   has something I think is the most significant. I told this story before. I'll keep telling

01:06:37   it again for the people who've been listening to every single episode of the show or weren't

01:06:40   alive in 2010 or whatever the hell. When the iPad came out, before it came out, we all

01:06:45   knew they were going to come up with a tablet, and we were all talking about it. A big topic

01:06:48   of discussion was what will this Apple tablet look like when you turn it on, right? We knew

01:06:58   what the phones looked like. The phones looked like a bunch of icons on a grid that had springboard

01:07:02   on them. But when you have an iPad, what will that look like? The leading contender was

01:07:07   not it would look like a phone, but the icons would be more spread out. I'm not going to

01:07:12   say it wasn't in contention because it was brought up, but it was so distant. Everyone

01:07:16   was thinking about what could it look like. It's not going to be the finder, obviously.

01:07:20   It's not running Mac OS. It's going to run iPhone OS or a variant of iPhone OS. But with

01:07:26   all that screen space, what could you do? When they came out, they said, "We're not

01:07:30   doing anything. It's exactly like the phone but spread out." It's like, "All right, well,

01:07:35   it's the first version of this." In practice, it works fine. It's an interface everyone

01:07:41   is familiar with. Here we are, what is it, eight years later, nine years later? Just

01:07:48   now we're getting rumors, hopefully, substantiated in some way, that now they're going to take

01:07:54   a crack at doing something other than having just a big grid of icons spread out a little

01:07:59   bit more. It's so far overdue. I'm not even a super iPad power user. Every time I hear

01:08:06   the umpteenth seven-hour dissertation of how people are arranging their icons into folders

01:08:10   with spacers and color-coding them and stuff, I'm like, "Oh, there's so little flexibility

01:08:16   in such an important interface." It's nice that folders were an addition, so that's good.

01:08:21   It's nice that people can find ways to make these two Lego pieces fit together in different

01:08:27   patterns because you've got the icons and you've got the folders, and we can do creative

01:08:30   things with them. But that's a pretty big screen, especially on the really big iPads.

01:08:35   There's so much more you could do to let people build a more efficient place to go back to

01:08:42   when they're not in an application. I don't know how far Apple will go with this, but

01:08:47   I will be very happy when they do something. Unlike my Mac Pro, they're going to get it

01:08:52   in under the decade line and say, "See? We eventually addressed what was a burning question

01:08:58   about an Apple tablet in less than a decade."

01:09:00   We are sponsored this week by Aftershokz bone conduction headphones. If you haven't tried

01:09:06   bone conduction headphones, I urge you to give them a shot because they work differently

01:09:10   from any other kind of headphone, and it really can be game-changing. The main difference

01:09:14   is that there's not anything sticking on or in or around your ears or blocking up your

01:09:19   ears. They have these little transducers that rest next to your ears that send microvibrations

01:09:25   that you don't feel through your cheekbones and that your eardrum picks them up to make

01:09:28   sound. But all that doesn't matter. What matters is there's nothing covering up your ear or

01:09:32   sticking in your ear or anything like that. So comfort on these for me is way better than

01:09:36   things like earbuds because there's nothing in my ear to make my ears hurt. And the main

01:09:40   thing is you hear ambient sound in addition to the podcast you're listening to or the

01:09:45   phone call you're taking on these. And so what that means is you can be walking or jogging

01:09:50   outside and you can hear if a car is coming. You can be using these around the house and

01:09:54   you can hear if someone's calling your name or if someone knocks on the door or if the

01:09:58   phone rings. It's amazing the freedom this opens up for both convenience and for safety

01:10:04   and just niceness. Like when I'm walking around outside I don't want to be blocked out from

01:10:08   the rest of the world. I want to hear the rest of the world but I also want to hear

01:10:10   a podcast and these are just great for that. They also have you know just nice convenient

01:10:14   features. They are small. They are weightless. They only weigh 1.06 ounces. These are amazing

01:10:21   lightweight headphones. They're of course wireless because you wouldn't buy non wireless

01:10:25   headphones these days. So they're wireless. They're weightless. They have good reception.

01:10:29   Bluetooth 4.2. Multi point pairing. All this stuff you know a built in EQ. Good battery

01:10:34   life. Hassle free two year warranty if you need it. So they're really nice headphones

01:10:38   to use. I highly recommend Aftershokz Bone Conduction headphones. So the flagship model

01:10:43   is the Trex Air. This has an MSRP of $179 but listeners can snag a pair for $50 off

01:10:49   by visiting ATP.Aftershokz.com and using promo code ATPBUNDLE. That's ATP.Aftershokz.com

01:10:57   code ATPBUNDLE for $50 off the weightless wireless Trex Air bundle. Thank you so much

01:11:03   to Aftershokz for sponsoring our show. Oliver DeJean writes, "Currently I have a MacBook

01:11:11   Pro 13 inch mid 2014. With my previous MacBooks I've always updated the hard disk drive or

01:11:15   RAM at some point. For this you had to remove the lower case and I always use this opportunity

01:11:19   to remove dust with some compressed air and with a brush for the fans. I cannot tell you

01:11:23   if it caused the MacBooks fans to spin up more rarely or not. However I thought it couldn't

01:11:27   hurt. On my current MacBook the memory is soldered in place and I also had no reason

01:11:30   to replace the SSD yet. So until now I haven't removed the lower case on this machine. After

01:11:35   four and a half years of usage I'm wondering if it would still make sense to open the lower

01:11:38   case and remove some dust. What do you think?" I don't think I've removed any part of any

01:11:43   of my MacBooks for years. At most I would just blow some compressed air in the vents

01:11:48   and call it a day. But Marco, what would you do? I would say if you're not having actual

01:11:54   problems with the fans and cooling, like if it's not broken, don't fix it. Because you

01:12:01   run two big risks here if you do this. I mean number one, even if you just blast compressed

01:12:06   air into the vents, you run the risk of shoving some dust where it wasn't before that could

01:12:12   possibly cause problems. And if you take the bottom case off, not only do you run that

01:12:18   same risk, but you also run the risk of damaging something in the process of taking it off

01:12:23   or putting it back on. There's a, as you learn from the stage or anything, there's a lot

01:12:27   of very, very small, very sensitive, very weak components that are in these. And for

01:12:34   somebody who has not been formally trained on how to service each model like some of

01:12:38   the Apple geniuses in service places, and for people who don't do this for a living

01:12:43   and maybe this is your first time or tenth time or fifth time opening up a laptop ever,

01:12:48   the risk I think is too high unless you're having a problem. And even if you are having

01:12:53   a problem, I think my first step would be bring it to Apple, like have them do it if

01:12:58   it's still under warranty. Just because the risks of slightly or less than slightly breaking

01:13:06   something I think are too high.

01:13:07   - Yeah, I mostly agree, like especially with laptops and these tiny little devices, I never

01:13:11   like the idea of them ever being opened by anyone, even a professional who's opened 100

01:13:15   of these this week. It's never the same. No matter how good you are, it's just never the

01:13:21   same. That said, we don't know Oliver's life. If you know that you live in a house with

01:13:27   75 cats and a smoker--

01:13:29   - I think you have other problems at this point.

01:13:31   - Right, and if you know your devices are constantly filled with stuff, it really depends.

01:13:37   I wouldn't say, if you know that's a problem, if you know all your devices eventually get

01:13:41   choked with cat hair and dander, you should think about doing something. But if it's just

01:13:49   like sort of a hang up, like you lead an otherwise normal life and your devices do not fail at

01:13:55   an incredibly higher rate and they aren't all choked with dust or sawdust or whatever,

01:13:59   like who knows what's going on. If you don't live in an exceptional scenario and you just

01:14:03   feel like if you look at your MacBook and you look at the events of any other person's

01:14:07   MacBook, they're more or less the same, yeah, don't open it, don't mess with it, don't blow

01:14:12   things into it. Hopefully you can assess this and know whether you have a problem or not.

01:14:18   But I can tell you that I have never opened up any of my little devices to remove dust.

01:14:22   That said, my Power Mac G5 and my Power Mac G3 and my Mac Pro all have a side that opens

01:14:28   up really easily and occasionally. When I'm in there for other reasons, I will blow out

01:14:32   dust. But they are cavernous and they are meant to be opened by people. There are no

01:14:36   screws, it's just a little handle and that's a totally different ballgame. And speaking

01:14:40   of the delicate parts in there, I was reminded of a sad thing I just heard today of someone

01:14:45   who remained anonymous to protect the innocent, who opened up his brand new Mac Mini to replace

01:14:52   the RAM and accidentally cracked some tiny little connector that connects the Wi-Fi antenna

01:14:57   because they're super delicate and if you pull it off the wrong way, it just goes "jink".

01:15:01   And so now all the time and money this person was saving by doing the RAM themselves has

01:15:05   gone down the drain as they have to go crawling back to Apple and say "yeah, I opened it up

01:15:09   myself and I pulled a little bit too hard on this tiny delicate plastic thing and now

01:15:13   please tell me how much money I'm going to have to pay to get this fixed".

01:15:16   Alright, moving on. Jeremy Kennison writes "I'm planning to get an iMac, not an iMac

01:15:21   Pro, for the family to use. No video editing, no coding or high intensity use. Honestly,

01:15:26   the only hardcore thing would be playing Firewatch, which by the way, if you haven't played it,

01:15:29   you should. Is there any difference that I should care about between the 3.4 GHz, 3.5

01:15:33   GHz or 3.8 GHz if they're all core i5 or whatever?" says Jeremy. "I feel like storage and RAM

01:15:38   are more important than the processor for a typical non-coding, non-developer user. In

01:15:43   my mind, this is not something I should worry about or need to pay extra for". My personal

01:15:47   opinion is, when you're buying a new computer, it's exactly what Jeremy said. Start with

01:15:51   RAM, get as much as you can afford, then storage, get as much as you can afford and then if

01:15:56   you need to worry about the processor, so be it. Obviously that's not true for everyone,

01:16:00   you don't need DMLs, but for an average rule of thumb, that's my two cents. We started

01:16:03   with Marco last time. John, what are your thoughts on this? In rare circumstances, there

01:16:08   can be differences between CPUs that are all out of proportion with a difference in price.

01:16:14   If it's just the case where it's like, oh, slightly higher clock speed or whatever and

01:16:17   there's no real differences, then yeah, you should know what you want to do with your

01:16:23   computer and if stuff that you do with your computer will not see an appreciable difference,

01:16:26   don't pay 300 bucks for the faster CPU. But every once in a while, not recently, but every

01:16:30   once in a while, there's like the "good CPU" and the one that's $100 less is way worse

01:16:36   and the one that's $300 more is just slightly better. There is sometimes discontinuities

01:16:40   in the line, so it's good to be aware. If you're not following the CPU world or whatever,

01:16:45   you can just look at benchmarks and say, look, here's how much faster this one is and here

01:16:49   is how much more money it costs and if you know you're never going to see or feel that

01:16:54   speed, it can actually be better to get the slower one, especially in a portable device

01:16:57   because it can use less power. So I think Jeremy correctly assessed his needs and should

01:17:04   not be chasing the best CPU. And to Casey's point, the thing that you will regret about

01:17:10   your computer much more than if it feels a little bit slower after a year is when you

01:17:14   run out of space or when, I mean, it's less now because you're not spending disks, when

01:17:19   you run out of RAM or like it ends up swapping. Again, with SCCs, it's much better than it

01:17:23   used to be, but yeah, RAM and storage over CPU if you're not doing anything where you

01:17:29   need the absolute fastest CPU.

01:17:31   Yeah, you pretty much covered it. The one thing to get to be aware of is that the CPU

01:17:37   performance within a family does not scale linearly with the advertised clock speeds.

01:17:43   There's lots of complexity these days in real world performance. One of the biggest things

01:17:47   is turbo boost, which is, you know, like the CPUs are not always operating at the speed

01:17:52   they advertise. They're often operating at higher speeds. And when you have, you know,

01:17:58   three different clock speeds within one CPU family that's like the same generation of

01:18:03   processor, it's, you know, everything else is the same. They're all branded i7 or they're

01:18:06   all branded i5 or it's a mix of i5s and i7s or whatever. Like the actual difference between

01:18:12   different models is smaller than you might think because of all these complicating factors.

01:18:17   One of the easiest things to do, as John said, is like, you know, you can go do research.

01:18:20   Like for me, my favorite place to do research is the Geekbench browser. If you go to browse.primatelabs.com

01:18:26   or something like that, we'll put the link in the show notes, you can look up every Mac

01:18:30   model with every processor going back a number of years. And so you can see right there,

01:18:36   you can see, all right, you look at the single core and the multi core and you can see, all

01:18:39   right, here is the 2018, you know, MacBook Pro, whatever, whatever it is, right? And

01:18:43   you can see, all right, here's the top CPU here, middle CPU here, bottom CPU here. And

01:18:49   if you look at what the actual variance between these is, most of the time, with a few exceptions,

01:18:53   like if the high end model has more cores, like is the case with the Mac mini and the,

01:18:59   and some of the MacBook Pros right now, if the high end model has more cores than the

01:19:04   low end model, it will have a significant advantage in multi core benchmarks, you know,

01:19:08   on the order of like, you know, 20 to 50%, maybe, you know, it depends on, you know,

01:19:12   what the difference is. But usually within a family, usually you don't have that choice.

01:19:18   Usually it's just like, here's three different clock speeds. And if you look at the actual

01:19:21   differences on Geekbench, you'll see that usually the range between the low end processor

01:19:26   and the high end processor is like 15%, maybe it's like in that ballpark. And so it's probably

01:19:33   not worth you spending an extra three or $400 to get 15% more performance. Most people are

01:19:41   not well served to spend the amounts of money it takes to get such a small percentage performance

01:19:46   increase. Whereas, as John and Casey said, these days SSD size, storage size, is the

01:19:54   premium. And so if you're looking, if you have a limited budget and you're looking for

01:19:57   which aspect of this computer should I upgrade or focus on, I would focus on SSD size because

01:20:03   even the lowest end CPU in almost every Mac family is totally fine and not that different

01:20:10   from the highest end one.

01:20:12   All right, and finally, Andrew writes, "What tweet management approach do you have? I ask

01:20:16   because Marco's tweet count seems to be decreasing since Andrew named Marco. Why don't we start

01:20:21   with you?"

01:20:22   Well, I don't really have much of an approach necessarily. It's not like a methodology or

01:20:26   a formal thing where I'm like, you know, having standups every morning in the parking lot

01:20:30   and whiteboarding my action items. So it's not really a formal thing. I've been tweeting

01:20:36   less because Twitter sucks. That's about it. Like every time, like it's like, you know

01:20:42   how if you eat a certain type of food and you vomit after eating it, you tend not to

01:20:49   want to eat that type of food for a while if ever. Imagine that feeling in a smaller

01:20:55   way every time I go to Twitter. It's like, wow. Every time I use Twitter, I feel a little

01:21:02   bit worse about using Twitter more in the future. And so that feeling is accumulating

01:21:07   over time.

01:21:08   I wonder if your tweet count actually is decreasing though. Like I don't know. I'd like to see

01:21:12   some charts because my impression is you've been tweeting more than usual lately. Or maybe

01:21:16   it's just they're more like bursts. I don't know. I would have to see some numbers on

01:21:20   it, but I'm wondering how this person came to the conclusion that your tweet count, they

01:21:24   said seems to be decreasing.

01:21:26   Well, did you delete a bunch of tweets when that was trendy like a month or two ago?

01:21:30   Yes. So the actual tweet count probably does go up and down a lot because I have one of

01:21:36   those things that just deletes tweets older than X days. But I don't think it didn't sound

01:21:40   like this is what this person was talking about.

01:21:42   Yeah. I read it the way you read it the first time. And then I came back to that, you know,

01:21:47   week or not a week, I guess a few days later. And I thought what he meant was, did you delete

01:21:52   a bunch of tweets?

01:21:54   Oh, yes.

01:21:56   But to kind of come back to your interpretation of this, Marco, I found that I am paying less

01:22:02   and less attention to Twitter over time. I'm still addicted, but less so. So that's good

01:22:07   if you can consider there a gray area in this conversation. I find that I'm tweeting less

01:22:12   and less because more and more frequently, even when I tweet innocuous things, I get

01:22:17   angry tweets in return. And I don't typically pay attention to my timeline anymore. I think

01:22:24   I've said this either on here or on analog, but I have a list of roughly 30 or 40 people.

01:22:29   It's a private list that I do pay attention to. And I am a completionist on that list.

01:22:34   But my main timeline, I only look at if I get really bored. And so I just have been

01:22:39   slowly backing away from Twitter. Slowly.

01:22:44   I mean, the thing is, like, if you go to a place where you're hanging out with your friends,

01:22:49   and suppose it's like a bar that you all go to with your friends, right? I love these

01:22:52   kinds of metaphors on Twitter because they're almost as overused as car metaphors on computers.

01:22:56   So it's like hanging out at a bar with your friends. And you keep going there because

01:23:01   all your friends are there, and it used to be a really good bar. And every time you go

01:23:05   there, someone walks up to you and slaps you in the face and walks away. And you're like,

01:23:10   "Well, that kind of sucks, but my friends are just..." And then a few minutes later,

01:23:13   someone comes up and just farts, and it just stinks up the whole place. And then they walk

01:23:16   away. And they're like, "Oh, God, this, okay, this is kind of unpleasant." And then a few

01:23:19   minutes later, somebody else walks up and just starts yelling at you, like a crazy person.

01:23:24   Like, "Why are you here? What are you doing? You're doing everything wrong." It's like,

01:23:27   "Okay." And eventually, you're going to be like, "Hey, guys, you want to go somewhere

01:23:30   else?" And that's kind of what I've been doing.

01:23:33   John, any thoughts?

01:23:34   Another vague question. Tweet management approach? Speaking of managing Twitter in general, Casey,

01:23:41   I have a recommendation for you. Take that list of 40 people where you read the timeline

01:23:46   and make that your new list of followers and unfollow everybody else.

01:23:49   It's funny you said that. I really mean this. It's funny you said that. And I have thought

01:23:53   about doing exactly that a lot. The reason I haven't yet is because there are a lot of

01:23:59   people who I really, really, really, really like in the real world that I don't want to

01:24:06   upset by unfollowing them.

01:24:08   That's not how—you should change your mental model of Twitter. Come to my mental model

01:24:11   of Twitter, which is, me following you doesn't mean I like you. Me not following you doesn't

01:24:15   mean I don't like you. I follow people whose tweets I want to see. That is the whole criteria.

01:24:21   Just because I don't want to see your tweets doesn't mean I don't like you. Maybe you're

01:24:24   awesome. Maybe you tweet about a thing that I love, but you tweet about it just slightly

01:24:27   too much. Like, "I love dogs, but maybe you tweet about dogs too much. I won't follow

01:24:31   you." Don't take it personally. I just don't want to see all your dog tweets. It's fine.

01:24:35   Anyway, everyone can decide how they want to follow.

01:24:38   Tweet management approach? I guess we did—it's broad brush. Marco talked about his deleting

01:24:43   tweets with some automated thing and why he's not liking Twitter too much, and you have

01:24:48   your list and then your other timeline. Everyone knows my approach. He's listened to the show

01:24:52   for a long time. I'm a completionist. I trim my follow list so I can read everybody's tweets.

01:24:57   Pretty much nobody else in the world does that because it's not the way most people

01:25:00   use Twitter, which makes me think, again, how could Andrew know what Marco's tweet

01:25:06   count is? I get this a lot from other people who reply to me, and it's so clear from all

01:25:10   the replies that I get. People don't have time to read everybody's tweets, so they'll

01:25:13   ask about something that I tweeted about an hour ago. Why don't they know I tweeted about

01:25:16   an hour ago? Why would they? It was an hour ago. They're not like their Twitter completionists.

01:25:24   I tweet when I feel like it. I read Twitter when I feel like it. The only thing that I've

01:25:29   been doing more in the past few years than I have earlier is I'm getting much quicker

01:25:35   with the block button. Blocking and muting are the two biggest new tools in the past

01:25:45   three years that I've used. I don't need to use them that much, but that's the only thing

01:25:51   that I've added to my normal practice of keeping my follow list small. No one slaps me in the

01:26:00   face or farts when I go on Twitter for the most part. I have the discipline not to respond

01:26:06   to most people, which really helps, but the people I do respond to are nice and I have

01:26:09   fun, and sometimes I do see pictures of dogs and they're cute. Except for we rate dogs

01:26:14   who rejected my dog picture a year ago and then post one that looks exactly the same.

01:26:18   Not that we're bitter. So angry about that. Thanks to our sponsors this week, Aftershocks,

01:26:23   Marine Layer, and Casper. We'll see you next week.

01:26:53   Bye.

01:27:00   [MUSIC]

01:27:28   We rate dogs is all, "Oh, if you have a cute picture of your dog, just DM it to me. Maybe

01:27:32   I'll include it on my thing." "Great, I've got a cute dog. Check out this cute picture."

01:27:36   Nothing. Silence. No response for a year. A year later, we rate dogs is like, "Look at

01:27:41   this dog that blends into the surface that it's on." There's been a million of those

01:27:46   pictures from we rate dogs. "I sent you a perfectly good dog blending into the surface

01:27:51   that it's on picture a year ago and you never posted it. Are you trying to say my dog isn't

01:27:54   cute or isn't blending with its surroundings?" Nothing. Nothing from we rate dogs.

01:27:59   Very, very sad times. I'm sorry, John.

01:28:01   It is. My dog is cute. Everyone should see my dog.

01:28:03   Marco, you shipped something this week. You shipped Instant Search and I'm assuming that

01:28:11   you used RxSwift for that. Is that correct?

01:28:13   Oh, yeah. I always use PrescriptionSwift for everything. You should have called it InstaSearch.

01:28:20   That's true. Oh, that's a missed opportunity.

01:28:21   See, I can't use Insta anymore because I came out with Instapaper and then a couple of years

01:28:27   later, Instagram came out. And now, if I would use InstaSomething, everyone would think I

01:28:34   was just aping Instagram. Just call it FinstaSearch. Wow. Fake InstaSearch.

01:28:41   Anyway, I haven't actually taken the time to play with this because I'm a terrible person.

01:28:47   But you said in your blog post about it, which we'll link in the show notes, you said, "This

01:28:52   was a lot of fun to build and I'm proud of it," which means now I need you to tell me

01:28:58   why was it fun to build and tell me, not that you shouldn't be, why you're proud of it.

01:29:02   Well, all right. So what Instant Search is, for a long time, I've had this fantasy for

01:29:08   Overcast of, "Wouldn't it be cool to move the most common search items, like the most

01:29:16   commonly searched podcasts, into a local cache that just is bundled with the app and will

01:29:22   therefore be instantly searchable so that when you're typing in the name of something

01:29:27   that you want to add, like the name of a new podcast you just heard about you want to go

01:29:30   at it, it just shows up immediately." And I've tried lots of different ways to do this.

01:29:38   And the main problem I kept coming back to, I basically tried this every six months for

01:29:42   like the last few years, and the main problem I kept coming back to was the index that it

01:29:47   would create, the search index it would create, would just be too big. And so I would shelf

01:29:52   the idea, like, "Okay, it isn't worth the size. I'll shelf it for now and I'll come

01:29:56   back later." What I started doing, maybe about a year ago, I forget exactly, is I started

01:30:03   recording with new subscriptions for new podcasts what search queries led to that podcast being

01:30:12   subscribed to. And so that gives me, when I combine that with popularity rankings on

01:30:18   Overcast, just how many people subscribed to each one, that combination is able to give

01:30:23   way more relevant results than straight up title relevance matching with no waiting.

01:30:31   Imagine Google without page rank, that's kind of crappy, it's just a keyword search

01:30:35   at that point. That's how most podcast apps are able to search, is they just kind of do

01:30:40   like a dumb text relevance score, like there's algorithms like BM25 and stuff like that,

01:30:47   there's algorithms where if you search for accidental tech and our podcast shows up on

01:30:52   the list because it includes those keywords and you might rank it because it was part

01:30:56   of the title, which is more important than the body, and it was two thirds of the title

01:31:00   of this show, and some other show it might only be two of ten words in the title, and

01:31:06   the words are next to each other, so you might rank it highly. There's all these different

01:31:08   algorithms that search engines use to rank to see what's most important, that you can

01:31:13   use without any knowledge of the result set of how important they are. All you know is

01:31:20   how well does it match the text of the query. So that's what I was doing for years, and

01:31:24   it's been mediocre, it's been fine. Anyway, so I started in more recent times adding popularity

01:31:29   rankings to that, and episode recency, so that way if you search for the talk show,

01:31:36   you get John Gruber's more recent talk show podcast, not the one that was on Five by Five

01:31:42   like five years ago and hasn't been updated since. So there's all sorts of matching things

01:31:47   you gotta do to kind of account for that. So you have, I factor in recency of episodes,

01:31:51   I factor in popularity on Overcast, so that way if you launch a podcast called This American

01:31:56   Lives, you know, and you shouldn't get every search for This American Life because you

01:32:02   kind of didn't earn that, like people who search for that probably meant the first,

01:32:05   probably meant the big public radio show, not yours, right? So you gotta account for

01:32:10   stuff like that, and so the more factors you can put in that are based on popularity or

01:32:16   like the merits of this as verified by human beings, the more relevant your search can

01:32:21   be. So I started building in this thing about like whatever people typed, and then whatever

01:32:27   they actually subscribed to, weight that more highly. So over time, if you figure like that

01:32:33   will eventually make, like if you type in a search query and it brings up three results

01:32:38   and the one that everybody really wants always shows up third, over time that will actually

01:32:43   rise in the list up to first place. Because if everyone types in This American Life and

01:32:48   above it is This American Lives and This American Love and then This American Life, below that

01:32:52   like eventually people will type This American something, subscribe to This American Life,

01:32:58   and eventually that will rise up to the top until it is the first result in that list.

01:33:02   So because I had this increased relevance building algorithm, I was able to then make

01:33:07   a downloadable local search index using that combined with my other popularity and stuff

01:33:12   like that, that was way, way, way smaller, but that actually had significant relevance

01:33:19   to most searches. And the index of that takes about three megs. And I could make it smaller

01:33:26   if I made it slower, but I include in that three meg index everything that is needed

01:33:33   to show it in the search result list, like the artwork URL, the title, the description,

01:33:37   the authors, like I include all that in the database even though I could fetch it, but

01:33:41   it just shows up faster. And so the result of this is I've been able to achieve my dream.

01:33:48   You can go to the search box and you can type in a single letter. Effectively immediately

01:33:54   you will get three results right there on top. You type in letter number two and you

01:33:58   get a different three results, or maybe one of the same ones, but like letter number two

01:34:01   you refine your search even further, and chances are by letter number two or three if you're

01:34:07   searching for a popular show it's probably already on screen.

01:34:10   You're bearing the lead here. The most important feature which you demonstrated in your movie

01:34:13   is now finally you can go to Overcast. Go to the search feature and type the letters

01:34:19   ATP. And the number one hit is this podcast, which does not contain any words with the

01:34:26   sequence of letters ATP in them. Nevertheless, it will be the number one hit.

01:34:30   You've been able to search for ATP by the phrase ATP in Overcast for about six months

01:34:35   now because about six months ago I started indexing keywords in domain names as part

01:34:43   of search results. And our domain name as you heard in the song is ATP.fm, and so I

01:34:48   separate tokenized based on the dot, I discard the TLD, and so our domain name's keywords

01:34:52   are ATP. And so it would show up in relevance for a long time if you type in ATP for about

01:34:58   six months, and over time that query leads to result algorithm made it slowly rise in

01:35:05   that list, and now it's on top when you search for ATP.

01:35:08   Was it number one before? I don't know. I don't remember.

01:35:11   Well, anyway, it's the number one result now as it should be, and honestly Marco should

01:35:15   have added special case code for this a long time ago.

01:35:18   So Overcast, people always accuse me of this. Overcast has no special case code for ATP

01:35:24   with one very, very, very small exception. I do have a flag in the database. So normally

01:35:31   when a show, like you have a show like This American Life where it only usually includes

01:35:37   the one most recent episode in the feed, and then any old episodes you can't download

01:35:41   fresh anymore, like you have to like go buy their app or whatever to get them, like they

01:35:45   remove it from the public feed. So Overcast has a feature that's been there for a while

01:35:49   where it keeps track of what episodes it knows about from a feed that are no longer in the

01:35:55   feed. And for people who are newly arriving at that podcast, it doesn't show the ones

01:36:00   that are no longer in the feed because the assumption is you probably can't download

01:36:03   them anymore. ATP is hosted on Squarespace. Squarespace has a limit of 100 items in RSS

01:36:10   feeds. So because Overcast have been running for long enough to catch ATP from episode

01:36:15   one in that feed, Overcast knows about all the episodes because they were at one time

01:36:20   in the feed that it saw. It just has many of them marked as no longer in the feed. I

01:36:25   do have a special flag in the backend that I can set for a podcast that says ignore what's

01:36:32   no longer in the feed and just show everything I know about that has ever been in the feed.

01:36:38   And so I have that set for ATP and a handful of other shows where I know this is a problem.

01:36:42   And therefore, Overcast can see all the back catalog of ATP even though the RSS feed only

01:36:48   ever included the last 100 episodes. But other than that, that is the only special case handling

01:36:53   of ATP in Overcast and that's not applied only to ATP.

01:36:57   Cote- So what made this so fun to write other than just accomplishing this thing that you've

01:37:01   wanted to accomplish for so long?

01:37:03   Jared- It was a fun technical challenge for me because it involved a lot of low level

01:37:08   stuff. I love low level optimization code. I love like take this enclosed problem that

01:37:14   like it's not, it doesn't really involve anything like frustrating. Like there's no weird concurrency

01:37:22   stuff. There's no weird like UI conditions and testing in weird states. There's no time

01:37:28   based anything. There's no network based anything. Like it's just a very like here's a bunch

01:37:34   of data that you need to operate on in low level ways and make it really fast and keep

01:37:40   it really small. And I love problems like that. So it was a very fun problem to work

01:37:44   on of like how do I make the index in the first place? How do I decide what goes into

01:37:48   the index? How do I rank what's in the index? And then how do I make the index as small

01:37:52   as possible? And then how do I make the app read and display contents from the index as

01:37:57   quickly as possible? And all of that was, oh and then how do I make the servers serve

01:38:02   the index as quickly as possible through the CDN and everything so the servers don't all

01:38:06   get hammered when all the apps come down with the update? There's all sorts of fun little

01:38:10   challenges that were involved here that I just, I just love that kind of work. I get

01:38:14   a lot of joy out of it. That's why I've been doing this. And then the result of it is sheer

01:38:19   delight. Like the result of seeing instant search because most search boxes don't work

01:38:25   this way. Like most search boxes in any other podcast app and even in many other kinds of

01:38:30   apps in general you're searching a catalog of things like the app store. You know like

01:38:35   you type in a letter and you don't immediately get results. You might after a few letters

01:38:41   get one of those like suggested search things where it just shows like the text below the

01:38:45   box and you can tap on it to actually run that search. You might get that but I don't

01:38:50   know of anyone else who does search like this where you're actually searching what is technically

01:38:53   remote content but it shows up instantly thanks to this local cache. So I really I'm very

01:38:59   proud of it. I think it's a very delightful feature and as I wrote up my blog post it's

01:39:04   an important feature because search is incredibly important to podcast apps and I think overall

01:39:11   we as podcast app makers have not done a great job prioritizing search for the importance

01:39:16   that it actually has. Like it is a really important feature and I kind of learned this

01:39:20   like I had over winter break. I had some server issues. Mostly it wasn't a problem that people

01:39:28   couldn't sync their playback. The biggest problem was you couldn't easily search to

01:39:33   add new podcasts and that made me realize quite how important search is because the

01:39:37   first thing you do when you install a new podcast app is use the search box to add the

01:39:41   shows you want to hear and what people do frequently throughout you know their usage

01:39:46   of the app is oh they hear about a new show I want to go add it. Go you type into the

01:39:49   search box. Speaking of that by the way, that use case I run across all the time. Someone

01:39:55   mentions a cool podcast and I want to go add it to overcast. I wish search was more available

01:40:02   in the UI like more prominent in the UI. Now that you've improved it so vastly I would

01:40:05   love it if like everywhere I could get access to from wherever I am and it was context aware

01:40:11   like if I'm on the playlist and I want to go find an episode and stick it in the playlist

01:40:15   that I'm currently looking on I don't want to leave the playlist and go back to the home

01:40:18   screen and do a search whatever I want to do it right there. So I think now that you've

01:40:21   improved search you should whatever your next UI redesign is think about how to feature

01:40:26   it more prominently and make it more useful and more context because now you know even

01:40:31   the old search I wanted to use frequently and now that it's fast I would love to just

01:40:34   use it immediately.

01:40:35   Yeah thanks that's actually that's a good idea. I mean I it's it's a little tricky in

01:40:40   that the the combination of the search box and the way iOS standard controllers display

01:40:47   search results by overlaying the table view with a new table view that has just the search

01:40:52   results in it and they kind of like it dims the background and moves the search box up

01:40:56   the way that's all done is through this thing called UI search controller. UI search controller

01:41:00   has been notoriously buggy and hard to work with over time. The current version is no

01:41:05   better it is very buggy and hard to work with it especially is buggy when dealing with navigation

01:41:12   pushes and pops where you have multiple levels of navigation like overcast has and where

01:41:16   you might you might want a search controller in a deeper level of it than the root level

01:41:21   or you might want one in both but or one or the other and overcast also has a bug since

01:41:27   5.0 and when I added search which is search controllers just totally break when used in

01:41:34   combination with pull to refresh. Apple has UI refresh control it is a built-in pull to

01:41:40   refresh thing and if you use that with Apple's built-in search controller to display search

01:41:45   box as a navigation bars they just conflict and you have animation bugs all over the place

01:41:49   this is I know this is a bug in overcast the reason why I haven't fixed it yet is because

01:41:53   it's merely a minor cosmetic bug if you pull to refresh and then you get the search box

01:41:56   kind of stuck on screen for a second it's a cosmetic bug and it is to fix it will involve

01:42:02   me writing my own UI search controller or and or UI refresh control and that has not

01:42:09   been a pressing enough need yet to justify the amount of work that it will be. Tell me

01:42:14   about the video player for this thing that appears to be a gif on your website but actually

01:42:19   has play/pause controls which are emoji by the way and I vastly approve of them. So that's

01:42:24   yeah so when marketing this feature it was important to me to show the video like so

01:42:29   a couple weeks ago I did a tweet from the overcast account saying like coming soon this

01:42:33   like fast search thing and the response to it was massive so I knew people really like

01:42:38   this this really resonated with people's and so I wanted to you know really market it right

01:42:42   with the blog post and so it had to be a video. As far as I could tell I didn't look too much

01:42:47   into it honestly but as far as I could tell animated gifs there's no good way for browsers

01:42:52   to pause their playback some of them will allow you to like click them to stop them

01:42:56   or hit the stop button to stop them but like I couldn't find a way to like offer a play/pause

01:43:00   button in the actual content of the page and give the users that control like reliably

01:43:05   and so I didn't want this big animated thing to be forcibly animating for everyone as they

01:43:11   try to read the text below it necessarily like I wanted it to be animating upon upon

01:43:16   loading the page but to be stoppable. I also like the gif quality was okay but it was like

01:43:22   two megs and I thought well let me see if I can get better quality with you know h264,

01:43:27   h265, webm whatever else I can get right so I I use this opportunity to revisit what can

01:43:34   the html5 video tag do these days and that is an html5 video it is literally it is three

01:43:41   versions of the file it's h265 which plays in most modern safaris and that is the smallest

01:43:47   of the files it also is h264 which plays in almost everything else and webm which plays

01:43:53   in google's whatever bs that plays webm that I think covers all modern browsers that have

01:43:58   a chance to do this I in order to at first I thought like oh let me let me use just the

01:44:03   built-in controls it doesn't make it super simple but the building controls at least

01:44:07   in safari overlaid the search box in the video and I thought that's that's not great that's

01:44:12   kind of ruining the point here so I did no controls autoplay play in line and there's

01:44:19   no sound so I don't worry about that and I thought well I'll just I'll make it so that

01:44:24   if you click on it or tap on it playback stops and if you click on it again it'll start again

01:44:30   people will figure that out but then I thought no people won't figure that out I can't I

01:44:34   can't depend on that so it does that if you click on it it will stop playing and it'll

01:44:39   click on it again it'll resume but I also wanted like some kind of very very small play

01:44:45   button and I tried different things I tried first like a link that just said play slash

01:44:49   pause or pause or whatever and it didn't it was ugly and big and eventually I'm just like

01:44:53   let me just make a little emoji thing I searched found the emoji play and pause buttons and

01:44:58   I wrote the smallest video player that I thought I possibly could it uses very you can view

01:45:04   source it's right there it's like you know a couple lines of JavaScript and and and that's

01:45:09   it and so it's very very simple and I'm actually kind of proud of like how easy and simple

01:45:15   it ended up being and it was kind of a fun little thing to just like play with HTML I

01:45:18   even did a I thought like what is the equivalent of an alt tag for video so that people using

01:45:26   screen readers can know what this is I learned this thing called web vtt and it's basically

01:45:33   a caption and transcription format like a subtitle kind of format and you can use web

01:45:39   vtt formatted caption files to use them either as subtitles for the video or as something

01:45:47   called text descriptions now semantically subtitles are not for this purpose like the

01:45:54   purpose of a subtitle is not for a visually impaired user to learn what if it what's in

01:45:59   a video it's for a hearing impaired user to be able to read what is being spoken in a

01:46:06   video now my video had no sound all I wanted was the equivalent of a video alt tag of like

01:46:12   to describe in text the image that is showing in the video and so there's a thing for that

01:46:19   called web vtt and and you can use the track sub element of HTML 5 video and you can give

01:46:24   it the the role like the semantic role of text description or something like that it's

01:46:28   all it's all in the spec somewhere and so I did all that and I transcribed like I have

01:46:33   I transcribed a video and it's it's great except no browser actually supports that yet

01:46:37   turns out so nice if I label it as subtitles then the built-in controls which I've disabled

01:46:44   because they were covering up the search bar the building controls would be able to display

01:46:48   it as subtitles if you turn them on but they're not subtitles and that kind of sucks and as

01:46:54   far as I can tell the built-in screen reader on iOS and Mac OS voiceover just skips right

01:47:00   over the video video element no matter what I did whether it was subtitles or text descriptions

01:47:03   or whatever like voiceover just skips right over the whole video does it like you it pretends

01:47:07   it is even there so I don't as far as I can tell this that was not a good use of time

01:47:13   to do this but hopefully in the future that will be useful to somebody I don't know and

01:47:18   at least I learned how to do it so if browsers ever do support that then that'll be nice

01:47:21   but it was it was kind of like a fun little thing to like make this little experiment

01:47:25   of this totally self-contained little video player on in this blog post and just see like

01:47:29   how easy can HTML5 make this for me and how minimal can I make it while still being functional

01:47:34   I dig it it's cool when do you fall back to doing a search across the internet against

01:47:43   your servers like at what point do you say the local cache is not enough and you go go

01:47:47   and ask overcast servers for help I always will run a remote search once you've typed

01:47:53   your third character so and one thing too like the instant search cache results never

01:47:57   move once they're on screen when the remote results from my servers come in they will

01:48:02   only ever go below them because what I what I didn't want is for you to type in something

01:48:08   see the instant search results go to tap it and then as you're going to tap it it changes

01:48:12   that from under you to something else and you tap the wrong thing accidentally like

01:48:15   that's exactly how the iOS app search works for me all the time yeah exactly that's a

01:48:19   terrible experience so I didn't want to do that so it's basically like the results are

01:48:24   stable whereas like the only way to make the contents of the list change is for you to

01:48:28   enter more text or different text in the text box so if you have typed a query and then

01:48:34   you go to tap something without typing any more letters whatever you go to tap on will

01:48:38   still be there when you get there I dig it it's cool so Casey you published a video I

01:48:44   did I win the competition between me hooray to not to my surprise yeah so I published

01:48:52   the CRV video what is that yesterday morning and it has been up for a while unsurprisingly

01:48:59   it hasn't made a tremendous splash I didn't promote it quite as aggressively as I have

01:49:02   in the past and I think the subject matter is slightly less interesting than some of

01:49:07   the stuff I've had in the past but the video is out I'm mostly happy with how it's turned

01:49:13   out and we talked about this a lot last episode so I don't really need to relive the motivations

01:49:19   and trials and tribulations of the video but all told I am pleased that it's out the door

01:49:27   I am going to start on the next video probably tomorrow and in general I'm I'm happy with

01:49:34   it there are definitely problems with it the one thing that slipped through the cracks

01:49:38   and I'll beat probably John if not both of you on this is that at one point hand to God

01:49:43   I don't know how this happened I don't know if I did it and then never caught it I did

01:49:47   I did notice it at one point and then slipped my mind after that but at one point when I'm

01:49:52   discussing how chatty the car is I have the video flipped in such a way it looks like

01:49:58   it's a right hand drive car I don't know how that slipped through slipped my mind to fix

01:50:02   that I don't even know how the video got that way but somehow that happened and somehow

01:50:08   the one time I noticed it I didn't fix it immediately and then forgot about it and that

01:50:12   really really ticked me off it was a self-created problem there are a couple audio pickups here

01:50:17   and there but mostly I don't think they're nearly as egregious as they had been in the

01:50:21   past and mostly I'm happy with it cool that's easy I mean I'm I don't we don't need to belabor

01:50:28   it any more than that I'm assuming John is about to beat me up with a whole bunch of

01:50:31   complaints and criticisms but but if if we're done at this point then I'd say go watch the

01:50:37   first video mash that subscribe and hit the bell or whatever I'm supposed to say and and

01:50:42   we're good John oh this is a video from the past so we're gonna talk about I want to talk

01:50:47   about his future videos when you release one that's newer than the one you released before

01:50:52   then we could talk about it yeah it's fine although I your your opinion of the CRV is

01:50:56   warped by your contact with these luxury cars of course but like it is you could say what

01:51:01   you want to say it is you're right and like I recorded as I stated on the video that wasn't

01:51:06   a lie I recorded what I thought was going to be darn near all the footage for the car

01:51:13   where I basically said it's a piece of trash and then it occurred to me it's really not

01:51:17   it's actually a very nice car it's just that it's not it's not as nice as the car that

01:51:22   cost twice as much which is weird right no no no you make my XC90 is nicer than this

01:51:28   is it is it nicer it is nicer it only costs two times as much well there's that I hope

01:51:34   it's nicer you're absolutely right and how do you feel about two CRVs versus the XC90

01:51:41   no you're absolutely right and that's why I that's that's why I ended up re-shooting

01:51:47   a lot of it and not to say I did it perfectly but my initial cuts at this were like oh this

01:51:52   is a piece of crap and it doesn't do anything nice and that's not true actually it's really

01:51:56   genuinely not true it's actually a very very nice car and it is an SUV I mean you can you

01:52:01   can slam it for that but like you have to compare it to like it's the comparables as

01:52:04   they say in the real estate business like there are other SUVs that cost about the same

01:52:07   amount of money and how does you know like how does this compare to them is when you're

01:52:12   you know running one of your the fancy cars like the quadrifolio or whatever you're comparing

01:52:18   it to other cars of the same type in the same price class and that's the valid comparison

01:52:23   but you're going to be constantly be comparing cars to cars that cost twice as much you're

01:52:27   always going to be disappointed yeah and you're absolutely right and I think I I should have

01:52:32   been more cognizant of that when I was when I was filming but you know you live you learn

01:52:36   but yeah so I'm gonna hopefully be able to make heads or tails of this Tesla footage

01:52:41   I shot in November I think within the next couple of working days I'm gonna know whether

01:52:46   or not I can do anything with it I certainly hope to I plan to it's not gonna be as long

01:52:53   a video as this was it's not gonna be as involved a video as this was but I hope to be able

01:52:57   to put something short out based on it even if I end up like I'm really worried about

01:53:01   the audio because I don't think I had my lavalier game dialed in correctly so even if I end

01:53:07   up just you know doing a voiceover with a bit of b-roll that I might just try to do

01:53:11   that we'll see once I go digging through this footage but I've set up a schedule for the

01:53:17   next several months of what I want to review and when I want to do it and so I think if

01:53:21   I don't do the Tesla video for February then I'm gonna do like a here's here's my my golf

01:53:27   are after having lived with it for several months and what do I like what do I not like

01:53:32   and so that's the plan for February and then I think the XC90 in March and and then we'll

01:53:39   see where it goes from there yeah go check it out youtube.caseylist.com it's only 10

01:53:44   minutes could you not get you don't you have a vanity or youtube.com/caseylist I haven't

01:53:50   done it yet I need to do that but I haven't done it yet youtube.com/saracusagethetameineverenouncemarco

01:53:58   yep speaking of which I should have recorded my day of woe and destiny because I would

01:54:03   have made a great video ABR always be recording

01:54:06   [BEEPING]