183: I Filed a Radar

0:00:00it is so hot in this room and so he mad

0:00:04the humidity has been brutal here it's

0:00:07only 77 degrees and i'm dying in this

0:00:09room because it's so humid and of course

0:00:11the product 90 degrees in his room

0:00:13I feel great air-conditioning's on I

0:00:15haven't put my slippers on mid-show

0:00:16cause my feet were getting a little bit

0:00:17cold

0:00:19we should probably actually get started

0:00:21huh

0:00:22no that's making on the show none of it

0:00:24not come on no yeah so here you know how

0:00:27I'm gonna make you put some of that in

0:00:28the show

0:00:28watch this hey speaking of cars we

0:00:31should probably do some follow-up

0:00:32starting with more reasons for qnx akc

0:00:35do we have any follow-up he just said

0:00:38that so Casey do we have any follow-up

0:00:43who he should probably start with the

0:00:45follow-up Marco that your you make an

0:00:46excellent point and we had an individual

0:00:49right in with more reasons for using qnx

0:00:51and these actually were the to my ears

0:00:55and eyes the best reasons I've heard yet

0:00:58and what if we basically boiled down to

0:01:00was incredibly fast boot times were in

0:01:04poor generally important for cars

0:01:06because you don't want to say have to

0:01:08wait until you're five minutes down the

0:01:09road before your air conditioning or

0:01:11radio turns on or something like that

0:01:12and the other reason that this

0:01:14individual gave was you know if you

0:01:16think about it the instrument clusters

0:01:18on a lot of cars are often driven by

0:01:20some sort of display or perhaps driven

0:01:22by signal that's coming off some

0:01:24connects derived computer and those need

0:01:28to be real time I mean that data needs

0:01:30to show you exactly how fast you're

0:01:32going right freaking now and so having a

0:01:34real time I was like you index makes

0:01:36that sort of thing a lot easier by

0:01:39Tesla's system is not based in qns i'm

0:01:41pretty sure it's just some kind of linux

0:01:42and you can tell like you can tell you

0:01:45did because like you know something like

0:01:47again like that last one like most of

0:01:49the time it works fine but sometimes

0:01:52like we were of state and and I heard we

0:01:56were driving around found out directions

0:01:57and the map is getting a little bit

0:01:59wonky like it was starting to like flick

0:02:01out and slow down a little bit and like

0:02:02not update quickly enough and that's

0:02:04that's great to like you know when

0:02:05you're following a map to enter my turn

0:02:07and just doesn't update so you're

0:02:08looking at like the last turn from 30

0:02:11seconds ago instead that's fun and then

0:02:13there's one point where it made the turn

0:02:17signal clicking sound when the turntable

0:02:20wasn't on a what there was in one of the

0:02:24like I described last episode that I've

0:02:26had to reboot it three times so far over

0:02:28the three or four months whatever it is

0:02:29one of the times that i had to reboot it

0:02:32immediately before hand the turn signal

0:02:35was not clicking like it would turn on

0:02:36and the light with the indicator blinks

0:02:38in the dash but you wouldn't have a

0:02:39click noise so obviously the clicking

0:02:41noise is made by the center console

0:02:43computer like that it's adding it to the

0:02:45sound system and all likelihood anyway

0:02:48so so the sims one of the symptoms I

0:02:51knew something was wacky this time was

0:02:53the turn signal noise started clicking

0:02:55when a certain terms of was not on i was

0:02:58driving straight down a road ahead and

0:02:59text it like that was it

0:03:01it's so that i rebooted again DX

0:03:03goodness so yes I i would apprehend

0:03:07anyone it's not booted the turn signal

0:03:09does not like one is rebooting you can

0:03:11turn and ended the indicator on the on

0:03:12the dashboard did blink when it was when

0:03:15it when the computer is booting like it

0:03:16was still blinking and so I assumed this

0:03:18signal is still on i hope the light was

0:03:20still blinking outside the car but

0:03:22there's no sound again it's like one of

0:03:24those things where like when you tie

0:03:26these things to software that you know

0:03:28II you are at the whim of the software's

0:03:30stability responsiveness up time you

0:03:35know it's similar to like one of the

0:03:37things that that I didn't like so much

0:03:39for the Apple watch is so many at so

0:03:43many interactions are tied to force

0:03:44touch and when you when you push really

0:03:47hard on the watch face too

0:03:51and you expect that button click even

0:03:53though it's not really a button even

0:03:54though it's all fake and simulated you

0:03:55know you expect that better click back

0:03:58immediately as if it were a button and

0:04:00the same thing with all the force touch

0:04:01trackpad like they have the same the

0:04:03same requirement like it has to respond

0:04:05like the physical object that it is

0:04:07mimicking right has to be quick it has

0:04:09to be immediate

0:04:10there can't be any delay on the watch

0:04:12you know the watch is very slow hardware

0:04:14wise and sometimes that sometimes the

0:04:15software gets coming up a little bit and

0:04:17so you push and sometimes delay before

0:04:19actually clicks back at you and it

0:04:22totally breaks the illusion that makes

0:04:23it feel broken or cheap or wrong or you

0:04:26know

0:04:27whatever the case may be and that's

0:04:28actually one concern I have with the

0:04:31force touch home button that is rumored

0:04:33to be in the next iphone that like you

0:04:35know if you get if if you push that home

0:04:38button and the iphone software is like a

0:04:40little bit you know a little bit

0:04:42overwhelmed or buggy or whatever if you

0:04:44push that and you don't get a click feel

0:04:46back first few seconds or at all like

0:04:49that's gonna feel really broken really

0:04:51quickly so anyway

0:04:53similar thing with like this car stuff

0:04:54like you know back to this follow-up

0:04:55saying the QX is really good for boot

0:04:58time and response times of interactions

0:05:00like I totally get that and I totally

0:05:03respect that because when you're tying

0:05:05things to controls of physical object

0:05:08that makes a big difference my complaint

0:05:10about the ancient iphone 6 that I'm

0:05:11using is that I have a physical home

0:05:13button and I press it and it always you

0:05:15know goes in just like you said I feel

0:05:17that physical feedback immediately but

0:05:18you know it doesn't happen immediately

0:05:19springboard doesn't appear immediately

0:05:22that's what does it by clicking and then

0:05:24look at my phone like what are you what

0:05:25are you doing well I push the button

0:05:27I'm know i know i push it because it

0:05:28went in and then i went out and then

0:05:30I'll ok now the animation starting i'm

0:05:32getting picking my old age I don't know

0:05:34getting picky

0:05:35what you should really do to make it

0:05:37seem like your phone is even older

0:05:39instead of saying my old iphone 6 you

0:05:41should say my first iphone which i'm

0:05:43still using that's true yeah I don't

0:05:46think my cell phone sex is getting

0:05:47slower but i'm just getting less and

0:05:49less patient for like when it doesn't

0:05:51respond i mean and i know it's not just

0:05:52being slowly my ipad pro then I points7

0:05:55driver was pretty darn fast i still feel

0:05:57like when i hit the home button on that

0:05:58sometimes I have to wait I want them to

0:06:00be instant just wait for the new touch

0:06:02ID sensor it's pretty damn fast

0:06:05yeah I know I that use my wife so I know

0:06:07that the touch ID sensor scans I'm just

0:06:09saying like I didn't feel like sometimes

0:06:11it's not responding to me as immediately

0:06:13as it has I feel like it should

0:06:15who knows anyway the other thing this

0:06:18person to say about you know why why do

0:06:20cars have crappy hardware we read some

0:06:23other feedback from someone who works on

0:06:25these systems talking about the how the

0:06:28car manufacturers want to save money and

0:06:29pinch pennies and I said that was silly

0:06:30couple of the factors that were also

0:06:33mentioned many other people are really

0:06:34very best

0:06:34person is that this a long lead time on

0:06:36hardware that goes into cars that

0:06:38development cycles are really long so

0:06:40they have to pick hardware that's

0:06:41available like three years before the

0:06:43carbon comes out and you know maybe it's

0:06:45even longer so maybe like four or five

0:06:46years but I'm coming over to pick

0:06:48something that is currently like 3 45

0:06:50years old by the time the ships in the

0:06:51car and finally stuff in cars has to

0:06:54operate extremes of temperature if you

0:06:57ever left your phone on the dashboard or

0:06:59even in a sealed cubby inside your car

0:07:01like not in the Sun but just tried

0:07:03taking your phone and putting it in the

0:07:04glove box or in the little console thing

0:07:06anywhere inside your car on a hot day

0:07:08when you come back

0:07:09chances are good after being away for

0:07:11several hours in the middle of the day

0:07:12your device will refuse to function

0:07:14because the temperature is too high and

0:07:16the the chips in the car obviously can't

0:07:18do that they have to continue working

0:07:20even when they're like 250 degrees or

0:07:22whatever and they have to continue

0:07:23working on its negative 40 these are all

0:07:26real temperatures not crazy Celsius

0:07:28yeah haha similar like like if you ever

0:07:31heard of like how slow the computers are

0:07:34on satellites or the space shuttle

0:07:35because like you know like you your

0:07:37stories you know about like how they're

0:07:39still running something that's about as

0:07:40fast as like a 486 or something like

0:07:42that and like the reason one of the

0:07:43reasons why they have to use such slow

0:07:46types of hardware is that they have to

0:07:48operate in extreme conditions that would

0:07:50just kill any you know like the chip

0:07:52than an iphone and so it's similar but

0:07:54less severe in a car

0:07:55yeah that's their excuse but like still

0:07:57it's a little ridiculous that you know

0:07:59that the value is the washing machine

0:08:00ships they're using like they're slower

0:08:02than the ones in space probes the space

0:08:04probes have powerpc 60 ones that are

0:08:05like rad hardened in them and they're

0:08:08using chips that are like from a gameboy

0:08:10advance so it's see moresee that are

0:08:13made a gameboy advance fast and that

0:08:14anyway bottom line we have the

0:08:16technology we can make the hardware in

0:08:18cars better and i think we probably want

0:08:20to shorten up the the cycle time on that

0:08:23too i know it sort of all developed as

0:08:25it as one big unit but you just can't

0:08:27afford to do that if we have if a

0:08:29manufacturer can show you in just a year

0:08:30off of the cycle time and they can have

0:08:32a year newer electronics that can make a

0:08:34big difference i think especially like

0:08:36that as the as the infotainment systems

0:08:38and even the instrument clusters and all

0:08:39that stuff becomes a more important

0:08:41differentiator for cars as a software

0:08:44becomes like all of the products

0:08:45software becomes more

0:08:46important part of the whole the whole

0:08:48product mixture the company that can

0:08:50figure that software out is gonna have a

0:08:51big advantage

0:08:52alright tell us about audio sync in

0:08:55quartz crystals which we've gotten a lot

0:08:57of feedback about market you want to

0:08:59cover this

0:09:00yeah sure so basically so you know the

0:09:02last couple of days we discussed a while

0:09:05back i discuss my my podcast tool to

0:09:07automatically sync up tracks that were

0:09:09recorded locally for each person

0:09:11multi-person recording and then you know

0:09:13sync it up to the master track because

0:09:16if you try to import those tracks

0:09:18manually you think it up the difference

0:09:20in the in like the sense of time that

0:09:24each person's audio interface had it

0:09:26caused the problem called drift where if

0:09:29you sink of our recordings like right up

0:09:30front the beginning of the show an hour

0:09:32in we might be out of sync by like half

0:09:35a second or a second my microphone

0:09:37interface the little you know converter

0:09:39in it that samples the audio at 44100

0:09:42times per second has a very very very

0:09:44slightly different interpretation of

0:09:47what that means like have of how long a

0:09:49second is or how how many times it has

0:09:51to be then the one in cases of John's

0:09:53computers and over time that very very

0:09:55tiny error can add up to quite a lot and

0:09:58so this caused the problem of drift

0:09:59we've gotten lots of feedback about why

0:10:02this happens last episode I speculated

0:10:05that you're making these these

0:10:07components physical components and any

0:10:10little tiny bit of imprecision and

0:10:13making them when you when you are taking

0:10:1544,000 samples per second over two hours

0:10:19a very small variance in the clock

0:10:23performance of two different physical

0:10:25devices will cause them to to disagree

0:10:28into drift over time like this

0:10:29so my speculation was basically like

0:10:31it's not really possible to make

0:10:32something that is that is cheap and a

0:10:34computer like this that that is more

0:10:36accurate than that and we have to deal

0:10:38with it and we had lots of people right

0:10:39in from the pro audio world and from

0:10:42some people in from like the scientific

0:10:44equipment world which is pretty cool

0:10:45have pretty cool listeners also

0:10:47confirming what I said last episode that

0:10:49that pro audio gear actually people are

0:10:52going to say also pro filmed here does

0:10:54similar things rather than trying to

0:10:56like sync up a bunch of different audio

0:10:58tracks afterwards and software from

0:11:00different audio recording sources that

0:11:02might be on a film set or a recording

0:11:04studio they use the concept of a master

0:11:07clock and they have like one device

0:11:09whether it's a clock generator or just a

0:11:11device that content has an internal

0:11:12clock and progear has usually clock in

0:11:15and clock out ports on the back of it

0:11:17and so you'd like to actually physically

0:11:19wire all to each other and they like

0:11:22coordinate the clock based on one master

0:11:23source rather than each device keeping

0:11:25its own clock and therefore reducing

0:11:26this drift was a lot of people wrote in

0:11:28saying that was true and then the best

0:11:30feedback we got I I very casually

0:11:32mentioned last episode that i had just

0:11:34anecdotally found that laptops are

0:11:37generally generally have more drift then

0:11:40like a mac pro or a desktop or like an

0:11:42imac and it turns out there's a basis

0:11:44for this the quartz crystals that

0:11:46vibrate at particular frequencies that

0:11:49create these clocks these devices that

0:11:51have to be so accurate they are very

0:11:53dependent on stable temperature and if

0:11:56you if they don't have a stable

0:11:57temperature that are operating or even

0:11:59if it's just two different operating two

0:12:00different temperatures that can cause

0:12:03these these very very slight differences

0:12:05and laptops usually their internal

0:12:08components usually operate hotter than

0:12:11desktops and also fluctuate more and you

0:12:14know desktops have much bigger heatsink

0:12:16mechanisms they have usually larger fans

0:12:18are spinning faster and pushing more air

0:12:20of them so desktop to be cooled better

0:12:22and so I think that alone might explain

0:12:25that little thing that yes this is this

0:12:28is a this is worse on laptops or when

0:12:31you have like one desktop on one end and

0:12:34the other person other end is using a

0:12:35laptop that might have more drift

0:12:37between them then if both people reason

0:12:38desktops or able people using laptops

0:12:40and this is related to what I the the

0:12:43reason i was i put the original feedback

0:12:45and follow-up from like this seems weird

0:12:48to me tell me we can make quartz

0:12:49crystals that I can because I remember

0:12:51him you know of

0:12:53world of watches which i'm not that

0:12:54involved in but they do you know the

0:12:55fancy mechanical Rogers obviously are

0:12:57terrible keeping time because i have a

0:12:58bunch of gears and stuff but like you

0:13:00know and Tencent quartz watch keeps

0:13:02amazing time for you know for an entire

0:13:04year maybe we'll lose a second it's like

0:13:06so how do you make a stupid plastic

0:13:08quartz watch that keeps amazing time is

0:13:10very accurate over the course of an

0:13:11entire year but we can make a quartz

0:13:12crystal for our computer that is equally

0:13:15accurate and there are a bunch of

0:13:16factors that have to go into that having

0:13:18to do with the specifics of the quality

0:13:20of the crystals or even for tents and

0:13:22watch or whatever but one of our readers

0:13:24a reader's whatever is road and tell us

0:13:27that the advantage of watch has is that

0:13:30is essentially kept at an even

0:13:31temperature by being next to your skin

0:13:33so you are essentially temperature

0:13:35regulating the watch the entire time

0:13:37you're wearing it and that's the best

0:13:38thing for a quartz crystal actually in

0:13:40scenarios where they really want them to

0:13:42be stable they put them in a little

0:13:43every what they called it but they put

0:13:45him in a little device that keeps it a

0:13:46stable temperature so having a watch and

0:13:49wearing it all the time keeps it

0:13:51accurate much more elaborate than you

0:13:52know a laptop that you're the

0:13:54temperatures are going up and down and

0:13:55you're putting it to sleep and you're

0:13:56playing a game and it's just it's all

0:13:57over the map so this is all very

0:14:00explicable and now we should feel every

0:14:03time you learn anything about the actual

0:14:04analog physical world of components

0:14:07inside your computer makes you feel

0:14:08scared and so now you can know that uh

0:14:09thing the regulating the clocks or

0:14:11pieces of crap to and change all the

0:14:13time you do it

0:14:15yeah like whenever you learn about the

0:14:16analog component world lucky you start

0:14:18like we we live in like this digital

0:14:20world where we think everything is just

0:14:21like you know one is just on 20 is just

0:14:24off and of course when you get into the

0:14:26analog level of like the physical

0:14:28implementation of these chips and these

0:14:30components it isn't that simple and

0:14:32everything is kind of like tolerances

0:14:34and variances an approximation and

0:14:37tricks and like it's kind of amazing all

0:14:40the stuff we have works at all as

0:14:41consistently as it does

0:14:42yep i'm even just a clock like you learn

0:14:44from like CPU design and even is

0:14:46propagating the clock signal around the

0:14:47died of a large cpu is kind of it's kind

0:14:49of sketchy at best of all sorts of

0:14:51phase-locked loops and other things to

0:14:53make sure that that actually propagates

0:14:55everywhere and it's the same everywhere

0:14:56and you have delays and yeah so anyway I

0:14:59think we are now all completely

0:15:01satisfied

0:15:02that we understand why Marcus tool is

0:15:04necessary in life

0:15:05yep and speaking of in your quest to

0:15:09avoid doing actual work you actually

0:15:11filed radar

0:15:13yes I did I'm very proud of you yeah do

0:15:16you want to tell us about this okay so

0:15:18last episode in the after-show I got my

0:15:22time in the Sun or whatever the metaphor

0:15:26is and but whatever floats your bubble

0:15:29and so I got I got to finally explain

0:15:33the details of the mp3 file format and

0:15:36the the efforts i was trying to do with

0:15:38variable bit rate or vbr encoding and

0:15:41why podcasts couldn't practically use

0:15:44vbr encoding which basically boils down

0:15:46to the method to seek a vbr file that

0:15:49you don't have also like if you if you

0:15:52are playing a stream and and the user

0:15:54jumps ahead to a timestamp its way for

0:15:56the stream you haven't downloaded like

0:15:58the part of the file between those two

0:16:00points

0:16:00you can't really know exactly what bite

0:16:02position to jump to to get to know

0:16:05timestamp one hour 30 without using

0:16:08these look-up tables at the beginning of

0:16:09the VBR file that just don't have very

0:16:11much precision in mode in the current

0:16:13standards are in the old standard

0:16:15they're very precise and so over the

0:16:17course of it to our podcast you only

0:16:19have in in the most common jump table

0:16:21for i only have a hundred entries so you

0:16:23have you know minute precision at best

0:16:26you know it might even be less than a

0:16:28precision I've learned a lot since then

0:16:30I did a lot of experimentation I talked

0:16:32to a bunch of people and I learned a lot

0:16:35about this so I was recommending the use

0:16:37of this there's a 93 tag its abbreviated

0:16:42MLT 4mp location lookup table and this

0:16:46is basically a 93 tag version of the VBR

0:16:50offset jumped able to tell you you know

0:16:52which bites map to which time stamp so

0:16:54you can jump between the file easily

0:16:55without having to have down the whole

0:16:57thing just you know skander manually the

0:16:59benefit of the id3 tag is that it can be

0:17:01any size like the the VBR jump table

0:17:04thing is restricted to the size of like

0:17:06the one mp3 frame that they shoved it

0:17:09into for compatibility reasons so it has

0:17:11to be like basically below kilobyte or

0:17:13so the id3 tag

0:17:15can be anything you want to be so the

0:17:17id3 tag version this would be great

0:17:19because you can you can basically have

0:17:21arbitrary precision as much space as

0:17:22you're willing to devote to this and

0:17:24like for me like i was able to encode

0:17:26one second precision of the time stamps

0:17:28with something like 16 kilobytes of a

0:17:31total jump table so for like a 45-minute

0:17:34by podcast 16k for some table is fine

0:17:37and by having vbr you're saving like 20

0:17:40megs on the file size in a case of it's

0:17:43totally worth the savings to embezzle

0:17:45table so I also heard from Devon govorit

0:17:48after recording last episode i

0:17:50discovered a github open source project

0:17:53called audio cogs and and there's a

0:17:56group people who make JavaScript

0:17:58implementations of decoders for mp3 AAC

0:18:02a couple of the audio formats and a

0:18:04whole audio player all written in

0:18:06JavaScript that can act like decode and

0:18:08play these audio formats entirely in

0:18:11JavaScript and that support MLT tag

0:18:14these property BR seek methods and

0:18:17everything it's it's and if they didn't

0:18:18it's an easy way for me to add support

0:18:20to this like that i could then have my

0:18:23sites web player switch to this and then

0:18:25i would i would eliminate his problem

0:18:27anyway so I heard from one of the

0:18:29authors of this devin govett we went

0:18:31back and forth a few times

0:18:33Devon informed me that the front offer

0:18:35vbr I tag which is an alternative

0:18:38version of the that stupid 100 entry

0:18:41jump table is basically more precise

0:18:44version that it still has to fit within

0:18:451 mp3 frame but instead of being a

0:18:48hundred bites it can be like 1.3

0:18:51kilobytes so that's like that's better

0:18:54you know that's you have more space more

0:18:56precision medicine experimentation with

0:18:58that couple problems came up though

0:19:00first discovered in the experimentation

0:19:02going back and forth Devon that not only

0:19:04does apples decoder not support the MLT

0:19:10or the this VRI tag that has the more

0:19:12precise version even that that stupid

0:19:15little hundred bite version in the in

0:19:17the King tag that we talked about last

0:19:18episode

0:19:19apple doesn't use it there decoder

0:19:22completely ignores it

0:19:23it does read these tags to get the

0:19:26duration of the files so it's it's

0:19:29parsing them because the duration is

0:19:30also one of the fields in these tags so

0:19:33it reads them for the duration and if

0:19:34you add them in a hex editor and you put

0:19:36in duration you want their will show up

0:19:37like a quick look at everything as a

0:19:38duration so we know it's reading them

0:19:39but it completely ignores any of the

0:19:42entries that are in these seek tables

0:19:44that tell it which bite maps two weeks

0:19:46time stamp in any of these forms that

0:19:48ignores all of them

0:19:49this is bad so I basically wrote up a

0:19:52bug report and I i emailed some people

0:19:54inside Apple to say like you know like

0:19:55hey here's here's the problem having and

0:19:57you know if if you guys would support

0:20:00really any of these formats it except

0:20:03for maybe that maybe the stupid hundred

0:20:04white one but if if you support the

0:20:06front for bri frame or the awesome MLT

0:20:10id3 tag either of those would provide

0:20:13useful precision for to our podcast to

0:20:16be able to seek reliably within within 1

0:20:18to 10 seconds of the desired point as

0:20:20opposed to the stupid hundred white one

0:20:22which is like a minute off and their

0:20:24their estimation which is which could be

0:20:26any amount off really and is frequently

0:20:27like 32 34 36 seconds off anyways female

0:20:30this around and I made a blog post

0:20:32basically make as much noise as possible

0:20:34about this issue because i've i've

0:20:37realized I'm sorry for the massive

0:20:38diversion here I i like talking about

0:20:40like Randy apple stuff on a podcast

0:20:43where doesn't get me in trouble and

0:20:44where you guys can take me a little bit

0:20:46and rebut me it's better for me to

0:20:48reserve my blog for issues of maybe

0:20:53greater importance or more boring topics

0:20:56or whatever else

0:20:58my blog is a great way to spread a

0:21:00message to people who don't always

0:21:03follow me and that is often the problem

0:21:06like when i get myself into hot water

0:21:08it's often because a whole bunch people

0:21:10who don't follow me and don't really get

0:21:12my context are reading something that I

0:21:15didn't write very well and where I

0:21:17assume people would get my contacts and

0:21:19understand me and they'd be the ones

0:21:20reading it and that gets me into trouble

0:21:23but something like this like we're

0:21:25requesting you know that Apple implement

0:21:27an esoteric standard of the mp3 file

0:21:30format that's 20 years old that is

0:21:32really boring and doesn't spread into

0:21:34business insider or CNBC and there's not

0:21:37much embedded really controversial

0:21:38although hacker news found some things

0:21:40but most people would not find this

0:21:42controversial

0:21:42so this is i think is a very good use of

0:21:45my blog and kind of its helping inform

0:21:47me like how i should use my various

0:21:50outlets looking for anyway I found a

0:21:52giant bug report with example files an

0:21:54example of why why their way of just

0:21:57ignoring you seek tables for long BRB

0:21:59threes was bad what you know what's bad

0:22:02how you can fix it that's it that's

0:22:04there's nothing has happened yet on this

0:22:05front however i have gotten rumblings

0:22:08here and there that this that this bug

0:22:11report has traveled inside of Apple but

0:22:13that's all I know and I don't know

0:22:14anything else that's going on with it

0:22:16it is probably too late even if Apple

0:22:18decided I was right and they want to do

0:22:20this it is almost certainly too late to

0:22:22get into iOS 10 or mac OS Sierra but I

0:22:27would just love for this to happen

0:22:28sometime soon and the main reason why I

0:22:31can't do the audio cogs and the average

0:22:36is like that the JavaScript version that

0:22:38is only about a minute ago the html5

0:22:40audio element can fetch a file from

0:22:42basically anywhere but for javascript

0:22:45file you were into all these cores

0:22:46issues and you also you can't cross from

0:22:50https to http and all this is like all

0:22:53these restrictions of like you know for

0:22:55various web security purposes all these

0:22:57restrictions on what

0:22:59javascript is allowed to fetch from and

0:23:02for me to be able to play podcast from

0:23:04arbitrary podcast hosts i basically can

0:23:07use this thing I could run a proxy but

0:23:09if i do that then the podcasters don't

0:23:12get unique hit information really so

0:23:15that's no good

0:23:16so it's there's a whole bunch of crappy

0:23:18reasons why I'd in practice I can't use

0:23:21the JavaScript approach just decoded the

0:23:23whole file in java script and basically

0:23:25wearing been working on decoders so I

0:23:27i'm relying on Apple I have to if if if

0:23:30podcast vbr ever going to happen

0:23:32Apple has built in 30 coders that's it

0:23:34well but hold on so a couple of

0:23:37questions here first of all let's assume

0:23:39you did desired proxy it because this

0:23:42was that important you for the web

0:23:44client

0:23:45why couldn't you proxy each request

0:23:47individually I mean it's a it's a

0:23:49credible bandwidth but is there any

0:23:51other reason why you couldn't do that so

0:23:53at least they're all of these hits our

0:23:54unique just

0:23:55coming from you and not the actual

0:23:56person that's asking for him

0:23:58the main problem with with that is that

0:24:01even if you are connected the proxy such

0:24:03that every inbound request equal one

0:24:06back and outbound request which would be

0:24:08silly in general but in this case may

0:24:10make sense right so even if you did that

0:24:12the requested still appear to be all

0:24:15coming from one IP address

0:24:17sure so and most know it's kind of a

0:24:19hotly debated topic and the podcast of

0:24:21like what counts as a download and you

0:24:24know because there's it's as as with

0:24:26most things it's complicated ahead so

0:24:28you can't just say oh well let me hit

0:24:29you get noticed work that way because

0:24:31that's it

0:24:32yeah it doesn't matter listen

0:24:34necessarily so anyway most podcast host

0:24:37have their own idea of what what damage

0:24:40account as and most of them involve the

0:24:42IP address of the source in some way so

0:24:45it might be like you know a unique

0:24:47downloads like one IP like the IP has to

0:24:50be unique within a certain time interval

0:24:52for it to count as unique download or

0:24:54something like that like their there's

0:24:55all these different tricks that people

0:24:56do but basically if all the request came

0:24:59from my single IP that was running the

0:25:01proxy it would not get people's statute

0:25:04under count that being said i actually

0:25:08already wrote and run one such proxy i

0:25:11already have this and I designed exactly

0:25:14this way and that has exactly this

0:25:15problem and the only reason that I that

0:25:19it's not really a problem that hasn't

0:25:20got me into any hot water with anybody

0:25:22is because it's hardly ever used and I

0:25:25designed it because overcast has a

0:25:27twitter card on any share link i have

0:25:29twitter cards and so that it so if you

0:25:31view that with overcast link on twitter

0:25:33website or in any other client support

0:25:36the cards which is very few of them in

0:25:37practice i show a whole little overcast

0:25:39embedded player and it works it's great

0:25:41so you know you can stream into

0:25:42timestamps it works perfectly like the

0:25:44website but tiny twigs cars require that

0:25:47all assets loaded through them must be

0:25:49served over https so i actually have

0:25:53kind of like a little little function

0:25:55like a mapping that many big podcast

0:25:57host including libsyn our hosts many of

0:26:00them have like just like a simple like

0:26:03way to you can transform their they're

0:26:04not they're insecure urls in to https

0:26:08urls like

0:26:09simple string replacement on certain

0:26:10things and everything so like some hosts

0:26:13i can just redirect with I could just

0:26:14redirect that's fine but not all hosts

0:26:17and so that so I actually run this proxy

0:26:20that does this that follows redirects

0:26:22and when it you know when it went to

0:26:23host it knows about it can just keep it

0:26:25continued along when it's not it does

0:26:27that proxy of one-to-one connection

0:26:29mapping exactly the way you think and

0:26:31the only reason it isn't a problem for

0:26:32either my bandwidth costs or people in

0:26:34anatomy is that the cards get pretty low

0:26:37usage relatively speaking i mean that

0:26:39not a lot of people play podcast that

0:26:41way as far as i know so that's the only

0:26:43reason that works and to answer glasses

0:26:46underscore in the chat room who said

0:26:49send them something like the

0:26:50x-forwarded-for header that I do send

0:26:54x-forwarded-for that is a header that

0:26:55that proxy is used to tell what they're

0:26:58fetching from the the IP of the person

0:27:01fetching it from them so it's kind of a

0:27:03way to look forward

0:27:04they're the source IP problem is you

0:27:06can't trust that so if you have a

0:27:08podcast metrics thing that is trying to

0:27:10measure like unique IPS in an honest

0:27:12fashion you can't really trust if

0:27:15somebody sends you next 44 header and

0:27:17you get a whole bunch of you get a whole

0:27:18bunch of requests from one real IP but

0:27:21that IP so i'm actually 40 these

0:27:23requests from these other 10,000 ip's

0:27:25you really can't trust that for purposes

0:27:27of stats that you tell sponsors our

0:27:30first sponsor this week is betterment

0:27:32betterment is the largest independent

0:27:34automated investing service out there

0:27:36you probably hearing a lot about

0:27:37betterment and techcrunch the wall

0:27:39street journal and other major news

0:27:40outlets because betterment makes it

0:27:42easier more straightforward and less

0:27:43expensive to invest better men have

0:27:46built on smarter cutting-edge technology

0:27:48to bring you sophisticated investing and

0:27:50financial advice all about much lower

0:27:52cost than more traditional financial

0:27:54services more than a hundred and

0:27:55seventy-five thousand customers choose

0:27:57veterans advanced advice algorithms and

0:27:59beautiful user interface to manage over

0:28:01five billion of their customers dollars

0:28:03this is the perfect time to start with

0:28:05betterment and get started saving for

0:28:06retirement or other financial goals ATP

0:28:09listeners and get up to six months of no

0:28:11fees now they're regular fees are

0:28:13already very low way lower than what you

0:28:15think for most financial advisors up to

0:28:17six months of no fees but going to

0:28:19betterment calm / ATP that's betterment

0:28:22com

0:28:23/ ATP betterment investing made better

0:28:26ah so you had said before

0:28:31well if Apple just implements this then

0:28:34then you can start using it on the web

0:28:37because the html5 audio tag along

0:28:39Google's supported at cetera but what

0:28:42about your beloved windows users or you

0:28:45know next year is going to be the year

0:28:47of linux on the desktop so what about

0:28:48them

0:28:49first of all we don't have that many of

0:28:51them honestly like not a lot of people

0:28:53video use windows and and use my app or

0:28:58follow my sharing generated by my app or

0:29:00anything else and i would love to change

0:29:01that because there are other users out

0:29:03there i would love for more windows

0:29:05people to see the share links from

0:29:06overcast because that would mean that

0:29:08the share links are being spread far and

0:29:09wide and that's the whole point of the

0:29:10share links but reality is different

0:29:13reality is that they they really don't

0:29:15get much windows you look at all so

0:29:17that's that's problem number one or

0:29:18rather dodge number one where i can

0:29:20dodge this issue but same reason why I

0:29:24never test my websites in IE because I'm

0:29:26my audience my customer bases tend to

0:29:28have such incredibly low IE or whatever

0:29:31is called now what does it call now reg

0:29:33reg reg anyway the the main reason why i

0:29:42don't really need to care that much

0:29:44about this problem for other browsers

0:29:46other than like things that use apples

0:29:47built-in decoders and my own app is

0:29:51because it's only a problem for seeking

0:29:54to time stamps that haven't been

0:29:56downloaded that's it in it in all other

0:29:59ways vbr mp3 to work great already

0:30:01mentioned the same problem even if i

0:30:03added support for it that it doesn't

0:30:04matter because you have to wait for

0:30:06everyone to upgrade to an operating

0:30:08system that includes that support that

0:30:09the apple just added in practice yes but

0:30:12the amount of time i have to wait isn't

0:30:14that much especially for podcast

0:30:17listeners to a tech show like that

0:30:18that's a pretty it's a pretty upgrad

0:30:21group like the it's I we don't have to

0:30:24wait that long

0:30:25it even on the mac like I is probably

0:30:27like their upgrade curves pretty good

0:30:29but the mac upgrade girls has not been

0:30:30great like you're a year and be like

0:30:32fifty percent adoption which is better

0:30:33than windows but still

0:30:35for people who would see it again like I

0:30:38don't have to worry about using my app I

0:30:40control my app I can update my app I

0:30:42have to worry about people who are going

0:30:43to see the share links you know right

0:30:45now the showings don't have big

0:30:46audiences that they they get you some

0:30:48and some people see them and that's

0:30:50great i want to make them bigger because

0:30:52I think podcast sharing could use help

0:30:53and you know I mean everybody in

0:30:55podcasting things pocket sharing can use

0:30:56help but like I Here I am I actually

0:30:58like I actually have something that does

0:31:00help and I want I want to get bigger and

0:31:02better but the reality is even if you

0:31:05know if Apple had support suppose it

0:31:07makes it into the next major versions of

0:31:08OS x and iOS whatever you know not the

0:31:10ones coming out in a few weeks whatever

0:31:12but you know next year or in the spring

0:31:14or whatever else it will probably be

0:31:15about a year after that before i can

0:31:19reliably use it because at that point

0:31:21you know it's gonna be way more than

0:31:23half because first of all like the

0:31:25people who are seeing share links from

0:31:28social media are going to be way more

0:31:30likely be on mobile than on a desktop so

0:31:32the slower adoption curve on Mac OS is

0:31:35not going to be as big of a problem is

0:31:36going to be offset by the large

0:31:38proportion of those viewers who are

0:31:40going to be on ir on iOS and and you

0:31:43know android if if anybody can tolerate

0:31:45me who's using android which is big a

0:31:47big ask

0:31:49so you know I think iOS is I was we

0:31:52updated faster especially as Apple has

0:31:54figured out how to make people upgrade

0:31:55by adding emoji and messages tricks so

0:31:58again like I think give it a year after

0:32:01they add support and i think almost

0:32:04anybody can responsibly do this and it

0:32:06would be fine podcast listening if

0:32:09anybody doesn't know Rob walks the CEO

0:32:12of libsyn look listen as a massive

0:32:14podcast host that they've been around

0:32:16forever and they host a ton of podcast

0:32:19including this one and if you listen to

0:32:21this

0:32:22they probably host a lot of other

0:32:23podcast you listen to i'm pretty sure

0:32:25all the really FM shows are hosted their

0:32:27it so it's a very big podcast host and

0:32:31they do a podcast forget what it's

0:32:35called

0:32:36I'll put in the show notes they do a

0:32:38podcast where it's like tips for

0:32:40podcasters and then something like once

0:32:42a month the CEO of listen goes on there

0:32:44and give stats of like podcast user

0:32:48agents like what clients what platforms

0:32:50are downloading podcasts and listen is a

0:32:52pretty good source of this because they

0:32:55host so many podcasts across so many

0:32:57markets that isn't just like technically

0:32:59any sense i can give you on like how

0:33:01people use overcast is like how Apple

0:33:04nerds who know who I am mostly use

0:33:06overcast like the top podcast and

0:33:08overcast are mostly text shows they're

0:33:10not dead but if you look at the top

0:33:11podcast in the world is mostly not text

0:33:14shows so obviously overcast you so just

0:33:16not representative of all podcasters out

0:33:18there but listen stats are really close

0:33:21I think I i would say listen stats are

0:33:23probably the best representation we have

0:33:24outside of apple and there are talking

0:33:26of like you know how podcasters behave

0:33:29in mass what the overall market looks

0:33:31like and the stats they give I listen

0:33:35and take notes every month and so he'll

0:33:38make a split up here so the most recent

0:33:40ones except for june seventy-seven

0:33:42percent of listens on mobile devices and

0:33:45every month that number increases so

0:33:48already very mobile heavy iOS to Android

0:33:51is a little little over three two one

0:33:53and that that reaches going down and

0:33:55read is becoming more popular now

0:33:57because people are finally building in

0:33:59like android vendors like Samsung

0:34:01finally building in podcast clients that

0:34:03they weren't for a long time and so

0:34:05that's that's adding a lot to the

0:34:06Android side but so basically but you

0:34:09still have liked more than three times

0:34:11as many people listening iOS and Android

0:34:14and and only like twenty two percent of

0:34:18listeners listening on computers so this

0:34:21is a very mobile heavy market and a very

0:34:23iOS heavy market so that's why I think

0:34:26it's fairly responsible for a lot of

0:34:29podcasters especially if you are in like

0:34:31the apple tech world like we are we are

0:34:33audience is going to be even more skewed

0:34:35towards recent apple platforms i think

0:34:38one year after that support to this it's

0:34:40totally safe to do and I might even do

0:34:42it sooner that you feel optimistic about

0:34:45them actually adding support or do you

0:34:46feel like it's going to be two years

0:34:47before this bug is closed as behaves

0:34:49correctly

0:34:50when I filed this bug in my bug reporter

0:34:54it was next to two other bugs that have

0:34:57been open forever my i'm lucky if i get

0:35:01a response that I don't find that many

0:35:03bugs with apple because the the the

0:35:06effort on my side to feedback and

0:35:10potential benefit i will get from that

0:35:11ratio is so terrible like that you know

0:35:14this took the better part of two days to

0:35:16do things to do research properly to

0:35:18like figure out that Apple didn't even

0:35:19support these tags at all you know I

0:35:21thought they were supporting least the

0:35:22crappy one they supported none of them

0:35:23to anyone to to modify my these test

0:35:27files to like you know right these tags

0:35:28properly everything took a lot of work

0:35:30to write a big turtle bug report it give

0:35:32proper test coverage and everything it

0:35:34takes a lot of time a lot of work and

0:35:36usually I file bugs and they go nowhere

0:35:39if i'm lucky they might be closed as a

0:35:42duplicate and when they're closed a

0:35:44duplicate I believe I then lose any

0:35:46visibility on that I forget what the

0:35:49Destiny change the reason anyway so the

0:35:52feedback loop is terrible for filing

0:35:54bugs for Apple I you know people who

0:35:56people who do it are good people they

0:35:58probably floss you know that they're

0:36:00probably very good people

0:36:01I i have a hard time doing it most of

0:36:04the time because again like I i see what

0:36:06happens with the few but i do file and

0:36:08they just they sit around forever and

0:36:10they don't even get close

0:36:11it's not it and and the few that do that

0:36:13closed are often closed in a way that i

0:36:16consider invalid like you know that

0:36:20those they'll be still do something like

0:36:21wearing clothes unless you tell us you

0:36:24know really soon that this is still

0:36:25happening on the news built of iOS four

0:36:27a bug that I filed like you know six

0:36:29months ago that is easily testable like

0:36:30that they they clearly have a wide

0:36:34variety of quality of people who go

0:36:36through the bug reports and many of them

0:36:38clearly just want to close as many as

0:36:40possible that actually do any work like

0:36:41the incentives are wrong there are you

0:36:43saying you don't floss

0:36:45so anyway uh I using file bugs because

0:36:50I've had a very I've had a poor history

0:36:52of of any response from apple on bugs

0:36:55but this time i find it because i

0:36:56figured basically nothing else I can do

0:36:58I can't work around this like I i

0:37:00mentioned the JavaScript and I tried

0:37:02doing the javascript thing but you know

0:37:03cause of all the cores restrictions it

0:37:05and it makes it pretty much impossible

0:37:07to use for an arbitrary set of hosts

0:37:11that you don't control most of which

0:37:12don't send course permission headers

0:37:14already so that's a non-starter

0:37:18you know i could write my own decoder in

0:37:20my app that can calculate these offsets

0:37:22but that's only one app and so for

0:37:25something to make something like even

0:37:26even our podcast that you would think

0:37:29you know what percentage of people who

0:37:31download our podcast you can listen and

0:37:33overcast and a lot of people would

0:37:35probably guess it's a pretty high ratio

0:37:36and it is compared to overcast global

0:37:38ratio I haven't looked recently but i

0:37:40think it's something like sixty percent

0:37:41that's still forty percent of our

0:37:43listeners who don't listen overcast who

0:37:46wouldn't have who would need

0:37:47compatibility in this way you know for

0:37:49for seeking and streams i basically the

0:37:52reason find this bug is that it's my

0:37:53last hope to get BTR mp3's to be a thing

0:37:55I think that'd be great if they were

0:37:57thing I think it's ridiculous that they

0:37:58aren't of thing yet but if there's any

0:38:00hope of them being a thing it's

0:38:02basically on Apple to do because if

0:38:05apple doesn't do it nobody else can and

0:38:06nobody else will

0:38:08did you make a strong recommendation of

0:38:09implement a specific one of these

0:38:11because I if you just said oh here all

0:38:12these things you don't support i can

0:38:13imagine i'm supporting like that

0:38:15terrible one that only it gives you a

0:38:16hundred points and then being like done

0:38:18closed fixed

0:38:19yeah you know that my report actually

0:38:21reads a lot like the blog post i wrote

0:38:22the book report first and edit to be a

0:38:24blog post but basically it's it's a you

0:38:27know similar format and I I recommended

0:38:30you know I i gave three options and I

0:38:32said you know you don't post any of

0:38:34these right now this old one is not is

0:38:35not precise enough don't use this time I

0:38:37and so I said basically the MLT tag will

0:38:40be the ideal one to do because it can be

0:38:42arbitrary length arbitrary-precision so

0:38:45that's the ideal one if you only pick

0:38:47one pick that one

0:38:48the farm for bri tag that's like 1.3

0:38:52kilobytes worth of stuff like that's a

0:38:53good second choice like if you if you do

0:38:55that i'll be happy but the best one to

0:38:57do will be the MLT tag

0:38:59I wonder if they don't support it not

0:39:01out of laziness because like you said

0:39:02they are actually parsing the durations

0:39:04out of them but for kind of for the same

0:39:05reason related to something else set

0:39:07another why why webhost can't trust

0:39:09x-forwarded-for maybe they're afraid is

0:39:11going to people with garbage data and

0:39:13then they're afraid all I can expose

0:39:14you know exploits in there like a buffer

0:39:17of lower something because they put

0:39:18crazy offsets in there and then trigger

0:39:20but who knows i don't know that type of

0:39:22thing where it is or they just don't

0:39:24trust the the encoders to put good data

0:39:26in there they don't want to try to read

0:39:27garbage data these are solvable problems

0:39:30like you know make your your thing not

0:39:31have silly buffer overflows and sanity

0:39:35check the the offsets in the map and

0:39:37make sure they seem reasonable before

0:39:39blindly following may be disregarded it

0:39:42looks like line noise but I do wonder if

0:39:45someone did have to write the code pull

0:39:46these durations out of there

0:39:47why do they do that and like while I'm

0:39:49in there why don't just actually parse

0:39:50this whole format and just implemented

0:39:53you know so I mean your concerns are

0:39:55that are totally valid and I can I can I

0:39:56can completely understand how like it

0:39:59wouldn't surprise me at all if if there

0:40:01was an engineering meeting at Apple some

0:40:03ridiculous amount of time ago like

0:40:04somebody was like yeah you know what we

0:40:06we could parse these things out but the

0:40:09data might be wrong and if we just do

0:40:11this like you know percentage of time to

0:40:13byte offset thing that we kind of like

0:40:15an approximation like that'll be close

0:40:18enough and it will be consistent and so

0:40:21if there is you know there's a school of

0:40:23thought that says like you should know

0:40:24you you should override the the stupid

0:40:26mp3 encoder because who knows what

0:40:28garbage and you're gonna get there and

0:40:29you could just do your thing but and and

0:40:32you know just make it so that it works

0:40:33it works its kind of clothes for short

0:40:35songs and then it doesn't matter anymore

0:40:38that's good enough for scrubber work

0:40:40like that they want the scrubber like

0:40:41when you're removing scar around anyway

0:40:43it's probably like an inch on some

0:40:44webpages you can't tell if it's like

0:40:46what pixels on forever but for your

0:40:47specific use case which is no no I'm not

0:40:49had someone's not dragging the scrubber

0:40:51here i'm putting it off set in the URL

0:40:52down to the second i want that precision

0:40:54out of it

0:40:55hey YouTube does it you should do it too

0:40:57exactly yeah like so you know that there

0:41:00is an argument to be made there but you

0:41:02know a they're already reading the

0:41:04duration of these tax even if they have

0:41:06the whole file they still read the

0:41:08duration from that header and so if you

0:41:09put garbage generation as it did during

0:41:11testing it says alright yeah sure this

0:41:12file isn't you know nine minutes long as

0:41:14that of an hour like it's happy to use

0:41:16that value so it's already trusting it

0:41:18on some level also like mp3 encoder

0:41:21don't change that much like they it's

0:41:23there's very few that are actually in

0:41:24active use today and they all right good

0:41:27data like this isn't nice

0:41:28b7 anymore like we we have solved the

0:41:30mp3 encoder problem mp3 encoder is work

0:41:33and they work well and there are next no

0:41:35one else besides me is going on with a

0:41:38hex editor messing with these values

0:41:39like that here I can't like total

0:41:41garbage on a regular basis so I i think

0:41:45what's what's more likely to have

0:41:47happened is that maybe these decoders

0:41:49were written a very long time ago maybe

0:41:51around that time frame when the world of

0:41:53mp3 was still very much in flux and you

0:41:55had crappy encoders in like the late

0:41:58nineties and maybe they just haven't

0:41:59revisited since then because there

0:42:01wasn't a reason to that is the move that

0:42:03is the way more likely explanation for

0:42:04this is like this is very old code that

0:42:07no one had any justification of touch

0:42:08for a long time it works it well enough

0:42:10so fine and that's why I think

0:42:13ultimately it probably won't get done

0:42:15because I would be surprised if anybody

0:42:17was really motivated to devote you know

0:42:20a limited people have limited time so

0:42:23who's going to devote time to this out

0:42:25of there like engineering time budget

0:42:27inside Apple when they might have more

0:42:30pressing things to worry about with like

0:42:31supporting the new bluetooth headphones

0:42:33the next iphone you know I think that if

0:42:35anything else baby somebody working on

0:42:37is probably more important than this to

0:42:39apples overall corporate goals and SP

0:42:42just supposed to be an AAC anyway it's

0:42:44the future just use a different standard

0:42:45format and those uh the the offset

0:42:48things that can be all the front of the

0:42:49file

0:42:50yeah I love that magnetars thread that

0:42:54just use a different format does not

0:42:56solve all your problems that you know is

0:42:58the best yeah all things you think you

0:43:00can for example doesn't get your own

0:43:01doesn't solve all your problems and then

0:43:03I guess make every encoder on the planet

0:43:05understand your new container format

0:43:06will be fine

0:43:07yeah this has been a lot of time on this

0:43:09topic I'm sorry to everybody who doesn't

0:43:11care

0:43:11you probably not long ago but we

0:43:14actually got a decent response on other

0:43:15people actually enjoy hearing hearing

0:43:17about all this crap I'm very surprised

0:43:18by that actually but i guess our

0:43:20listeners are are both cooler and

0:43:22geekier than I then I would have assumed

0:43:24they were definitely cooler here i'm not

0:43:28sure we also sponsor this week by

0:43:31tracker tracker makes losing things a

0:43:34thing of the past go to the tracker com

0:43:36right now that the tracker com enter

0:43:38promo code ATP for thirty percent off

0:43:40your entire order now

0:43:42packers a coin-sized device that locates

0:43:44misplaced keys wallets bad computers

0:43:47anything you want it's a little thing

0:43:48you attached to anything in seconds use

0:43:50pear tracker to your phone and you

0:43:53attach anything I communicate over

0:43:54bluetooth and you can find precise

0:43:56location with a tap of a button so

0:43:58here's how this works

0:43:59the tractor itself is quite shipping it

0:44:00is it has on it a button and so when you

0:44:03push that button you can locate your

0:44:06phone so it works that direction

0:44:08you can also launch the app on your

0:44:10phone to locate whatever you have

0:44:12attached the tracker too so it's great

0:44:15for anything you lose your keys it's

0:44:17your keys or a big one obviously you

0:44:18know your backpack maybe like your purse

0:44:21bags computers whatever else things you

0:44:23have to find an object you have to find

0:44:25in in your home or are around around the

0:44:27workplace whatever else attached tracker

0:44:29to it you can find it very very easily

0:44:30with your phone you can also have said

0:44:32use a tracker in Reverse push the button

0:44:34trackr and your phone makes a beep sound

0:44:36even if the speaker is switched to

0:44:39silent mode so pretty great with over

0:44:411.5 million devices tracker has the

0:44:43largest crowd gps network in the world

0:44:45you're lost item actually shows up on a

0:44:47map even miles away

0:44:49never lose anything again with trackir

0:44:51listeners of our show getting special

0:44:53just kind of thirty percent off their

0:44:54entire order go to the that's th e

0:44:57tracker dot com the tracker com enter

0:45:01promo code ATP the hardest thing you'll

0:45:03ever find is the website then you'll be

0:45:05done the tracker com / ATP for thirty

0:45:08percent off your entire order thanks a

0:45:10lot to track for sponsoring our show

0:45:15so John if the cell signal at your house

0:45:17is so crappy why not and why not enable

0:45:19Wi-Fi calling and solve all your

0:45:21problems

0:45:21the first reason I didn't enable this is

0:45:23I didn't think my carrier supported it

0:45:25but I was just looking at the wrong

0:45:26place in settings for some reason I had

0:45:28in my head that just ended only a tease

0:45:30apart this is not verizon but not

0:45:32verizon sports too

0:45:33of course I've never gotten where the

0:45:34setting is but anyway it's in there

0:45:35somewhere you can find it but the second

0:45:39thing is when I turned it on and like

0:45:40okay alternate wanna give it a try it

0:45:43makes you enter your physical address

0:45:45because once you're on Wi-Fi calling if

0:45:48you call 911 911 can immediately tell

0:45:50where you are because you're you're

0:45:52coming from the internet

0:45:53essentially they can get a location for

0:45:55you which is kinda weird because you can

0:45:57get locations from Wi-Fi base station

0:45:59and everything a lot of things use a and

0:46:01use those that big map of base station

0:46:04mac addresses physical locations and

0:46:05stuff to help a location awareness but

0:46:08anyway does what iOS says please enter a

0:46:11physical location so when you call 911

0:46:13if you can't if you're too injured to

0:46:15speak into the phone or whatever

0:46:17don't worry about that will come to this

0:46:18address but of course that means that if

0:46:21you go someplace else and you're on

0:46:23Wi-Fi so you're working your own Wi-Fi

0:46:25and you call 911 and don't get a chance

0:46:27to tell them where you are they gonna go

0:46:28to your house

0:46:29that's my understanding anyway of what

0:46:31this message telling me after this

0:46:32address here if you call 911 this is

0:46:34where people go to but I didn't enable

0:46:37it because I'm like well then i have to

0:46:40remember to turn off Wi-Fi calling when

0:46:42i leave my house because then if i call

0:46:45911 people go to my house instead of

0:46:46where I am so that's one reason why i

0:46:50didn't believe in on the second reason

0:46:51is this is the stupidest reason that the

0:46:54reason i think is only vaguely so I but

0:46:56this reason is really done but

0:46:58nevertheless is a reason everyone has

0:47:00their reasons this is mine

0:47:01it changes the thing that appears in the

0:47:03status bar to make some ugly thing that

0:47:05says like VZW whatever it doesn't say

0:47:08like the horizon

0:47:10fit Wi-Fi fan symbol it says different

0:47:12words like in all caps and looks away i

0:47:14don't like it that's a reason like I can

0:47:18understand even though i think it's a

0:47:19little ridiculous the whole address

0:47:21thing fine whatever but because you

0:47:24don't want to look at VZW in your status

0:47:27bar that's your reason you you try to

0:47:29turn it on and see what it does to your

0:47:30status bar i use AT&T like a gentleman

0:47:33and so it says AT&T Wi-Fi but it gets

0:47:35rid of the fan thing isn't it or maybe

0:47:36know anyway it changes what's the status

0:47:38bar I didn't like it

0:47:39location 1 is that is the larger reason

0:47:42but I was kind of glad that I had a

0:47:44legit reason why did it look at that I

0:47:45could status bar anymore

0:47:47it would be nice if like I was intense

0:47:48movie would be and is enhancing would be

0:47:50only used Wi-Fi calling when connected

0:47:52to this base station i would like that

0:47:54setting don't mean because then I could

0:47:56say use Wi-Fi calling when connected to

0:47:57my home base station but never anyplace

0:47:59else and they don't have to worry about

0:48:00this then i get my home address to my

0:48:02home base station but no and anyway

0:48:04after I made the decision I made a

0:48:07couple of successful voice coil calls

0:48:09from my house where people could

0:48:10actually hear my voice so maybe a

0:48:12verizon loves me for coming back to them

0:48:15and not using this filthy Internet

0:48:16calling ok alright well I don't even

0:48:22know what to make of that that he asked

0:48:24that's moving on Graham Spencer's as

0:48:29indicated that Google and Facebook and

0:48:31others also do a charity match on bug

0:48:33bounties so just FYI and did it before

0:48:37Apple like it

0:48:38most people didn't write into a to gloat

0:48:40about how all these other companies that

0:48:42have been horrible but they totally did

0:48:43anybody actually topics tonight or is

0:48:47this all follow-up we're getting there

0:48:49that there's topics down there believe

0:48:50me don't worry we gotta bring another

0:48:52podcast like two days I don't know that

0:48:54that's true to its somebody had to go

0:48:56all deep on the mp3 stuff but actually

0:48:58it's very entertaining and I enjoyed it

0:49:00so I shouldn't give you a hard time any

0:49:02imagine how much dead space we have and

0:49:04it like the topic and everything this

0:49:05week if I didn't do it

0:49:06it's fair you don't know what's lurking

0:49:08down there and topics there's all sorts

0:49:09of stuff is there more about evil

0:49:11yes there is actually really yes I'm so

0:49:14excited

0:49:18alright well hopefully will run long

0:49:20enough that we won't get there I mean

0:49:21anyway let me tell you about my imac

0:49:23yeah it has it it has been rebooted I

0:49:27did it on purpose

0:49:29it ran 21 days on the stock ram without

0:49:33a UPS and without a GPS I've gotten in

0:49:35contact with max lwo dude

0:49:39owc whatever they call themselves and

0:49:41have requested and i'm told that I'm

0:49:44receiving a RMA return merchandise

0:49:46authorization so I think that stands for

0:49:48you i think what they're doing is

0:49:50sending me a box to send the ran back in

0:49:51and they said they will replace it

0:49:52posthaste

0:49:53I have plugged into the UPS i have

0:49:57pulled the UPS power and seen the imax

0:50:01day on for at least long enough to make

0:50:03me feel better about myself i did not

0:50:05leave it disconnected long enough to see

0:50:08whether or not the technology would be

0:50:10smart enough to shut itself down but i

0:50:12believe i have those settings squared

0:50:14away in this technology so it should

0:50:15shut itself down gracefully the imac

0:50:17won't but you know I'm used to that this

0:50:19point anyway hail so anyway I i am

0:50:23running still on the eight gigs of stock

0:50:25ram until I receive my new new batch of

0:50:29owc ram i will try that if that doesn't

0:50:32work then i will probably very politely

0:50:35but very sternly ask for my money back

0:50:36and get crucial ram which the entire

0:50:38Internet has written to tell me is the

0:50:40only ram i should ever really buy and so

0:50:42I think crucial buyers any different ram

0:50:44the notary set forget someone at one

0:50:46point sent me a long email telling me

0:50:48about the different bins of who buys the

0:50:50good chips versus the cheaper chips and

0:50:51I think Chris was in the same thing with

0:50:53that bc but anyway I've bought OBC ran

0:50:55for years and like i said i think the

0:50:57last time i had owc thing go bad

0:51:00the chip because my computer is ancient

0:51:01the chip the the the ram the dam was

0:51:04like four years old five years old maybe

0:51:07was six years old like whatever it is I

0:51:09just I just assume like well it's so old

0:51:11whatever warranty or whatever they had

0:51:13it must be completely gone by now but i

0:51:15just called not acceptable son you know

0:51:17that's it

0:51:17like I don't know if they have like a

0:51:19forever lifetime of this Ram everyone's

0:51:21bad they'll just replace it forever and

0:51:22ever and I know it's a hassle to return

0:51:23if you have to keep returning it's a

0:51:25pain but I've had them last for years

0:51:27and years and then go bad years later

0:51:29and then

0:51:29send me a new one it's things like

0:51:31miraculously like try doing that with

0:51:34the hard drive for example him a hard

0:51:35drive died after six years can I get a

0:51:37new one haha nothing like almost all ram

0:51:40has a lifetime guarantee like almost all

0:51:42around that you would buy separately

0:51:43from computer like yourself

0:51:44they almost all have lifetime guarantees

0:51:46but it's you know it's up to the

0:51:48retailer or the manufacturer to make

0:51:50that a good or bad experience if you

0:51:52actually have to clean it and I too have

0:51:54had only good experiences with obviously

0:51:57rama i don't use anymore now I just by

0:51:59my computer with my friend from apple

0:52:00because I'd rather not deal with anymore

0:52:01and the odbc ram when i was using it

0:52:05because up until like two years ago was

0:52:07always great i had to I had to return it

0:52:09one time and that was you know one time

0:52:12in something like eight years of using

0:52:14it i think it's pretty good yeah and I

0:52:17mean they've been really good about it

0:52:18so far I didn't expressly reach out to

0:52:20them they reached out to me at twitter

0:52:21i'm not entirely clear how they caught

0:52:23wind of the fact that I was having

0:52:24issues but they just basically said hey

0:52:27can you send us an email and we'll talk

0:52:29about it and I said hey here's the

0:52:31situation and I spelled it out and said

0:52:35you know it did it had been rebooting

0:52:36every week look at this it's now running

0:52:38for 20-some days at this point I feel

0:52:41like it's the ram so they said yeah

0:52:44you're probably right i will get you in

0:52:47RNA and we will replace it immediately

0:52:49I don't know how they heard about this I

0:52:51just I just told 80,000 people about it

0:52:53for a month and somebody told them well

0:52:56but what I mean is I don't recall having

0:52:58seen anyone like mention me and them on

0:53:01Twitter maybe that did happen i just

0:53:03missed it but it seemed like they came

0:53:06out of the woodwork as opposed to

0:53:08somebody tagging them to to kind of wave

0:53:10the flag in their face and say hey

0:53:11should pay attention over here so i'm

0:53:13not sure how that came to be but you

0:53:15know hey I'm happy that they reached out

0:53:17i'm happy that they seem to be more than

0:53:20happy to replace the ram because I mean

0:53:22like you were saying earlier i bought

0:53:24this room in january i looked up my

0:53:26order number and i forget when I was in

0:53:27January but it's still january i mean

0:53:29that was somewhat long ago and I guess

0:53:32you know most things have a year

0:53:33warranty but I don't know it it it's

0:53:36nice of them to not fight me on it

0:53:38because I'm used to most retailers being

0:53:40like well are you sure it's the rain

0:53:43manish take it to apple for stats etc

0:53:45and they were like no no send it back

0:53:47we'll get your ram so I said you didn't

0:53:49have to wait for someone to hear it into

0:53:51it you could have just called them on

0:53:52the phone all you gotta do is tell them

0:53:53you have bad ram and just like last time

0:53:55I call from like I said it was years

0:53:56after i bought it i said i've got a bad

0:53:59dream it's okay read the serial number

0:54:00ok we'll send you a box like that was it

0:54:02there was not even give them a reason it

0:54:04was dead

0:54:05yeah and i truly hand on heart was

0:54:06planning on calling sometime this week

0:54:08and then right as i decided no it's been

0:54:11long enough i'm happy while not happy

0:54:13that i am satisfied with my my

0:54:15scientific experiment that around that

0:54:18time when I'd finally concluded that was

0:54:19the case that's when they reached out

0:54:21were like hey what's happening so that

0:54:23was my plan

0:54:24Marco I am deeply sorry but I have to

0:54:28ask john tell me about tivo lifetime

0:54:30service please respond to this week by

0:54:34automatic the small adapter that turns

0:54:36your clunker into a smarter connected

0:54:37car go to automatic calm / ATP for more

0:54:40info

0:54:41automatic has just launched the

0:54:42automatic pro their new unlimited 3g car

0:54:46adapter with no monthly fees and no

0:54:48subscription no service charges you just

0:54:50pay once up front for the automatic pro

0:54:52and you don't pay anything per month to

0:54:55use their services and their abilities

0:54:56now the automatic pro includes always

0:54:59their 3g connectivity to let you know

0:55:02where your vehicle is parked at any time

0:55:04and let you track your vehicle in case

0:55:06of theft and of course is build on what

0:55:08automatic has been doing for years

0:55:10it works with if this then that for

0:55:11endless customization taking your car

0:55:13the rest of your life you can do things

0:55:15like automatically turn on your nest

0:55:17thermostats when you're nearing your

0:55:19home and of course tons of other

0:55:21abilities that you can do using it this

0:55:22than that and the automatic API in

0:55:25addition of course automatic provides

0:55:26the built-in functions they always have

0:55:28things like gas mileage tracking

0:55:30reaching your goals 484 fuel economy

0:55:32tracking your expenses for businesses so

0:55:35think it image of apps such as

0:55:36concurrent Expensify to be able to track

0:55:39costs associated with your driving there

0:55:41are so much you can do with automatic

0:55:43and of course now with automatic pro it

0:55:45has always on 3g connectivity which

0:55:47makes so many of these things even

0:55:48better and possible automatic can also

0:55:51call for help in case of a crash it

0:55:53detects severe accidents and if you

0:55:55aren't able to respond

0:55:56a train responder will call for help

0:55:58because you can't no automatic is

0:56:00normally a hundred and twenty nine

0:56:02ninety five foot and use our exclusive

0:56:04offer code ATP you'll save twenty bucks

0:56:06so it makes it about a hundred and ten

0:56:07dollars great deal

0:56:09no monthly fees after that you again you

0:56:10buy it up front and that's it you don't

0:56:12pay per month to use these services to

0:56:14go to automatic calm / ATP for more

0:56:17information and use offer code ATP to

0:56:20save twenty dollars off the regular

0:56:21purchase price thanks a lot to automatic

0:56:24the smart car adapter for sponsoring our

0:56:26show once again

0:56:27ah John go ahead and tell us about 20

0:56:32your saved by real-time follow-up

0:56:35someone just tweeted at me that from the

0:56:38verizon Wi-Fi calling fact when using

0:56:40Wi-Fi calling 911 calls will always try

0:56:43cellular service in the local market

0:56:44first even when the device is an

0:56:46airplane mode or cellular services off

0:56:48so I don't know

0:56:50now i'm back to just the status bar I

0:56:51suppose which is a silly reason I'll

0:56:54think about it i'll think about it i

0:56:56don't like getting phone calls at home

0:56:57anyway and i have lined up to take a

0:57:01screenshot of that status bar we can

0:57:02mark against which the show are to its

0:57:06you can just see what a horrid actually

0:57:08is alright tivo the news today about

0:57:11tivo not really that big news but uh

0:57:14apparently tivo is discontinuing a

0:57:17service for the tivo series 10 series

0:57:21one that was one introduced in 1999 and

0:57:26what their what they mean by

0:57:27discontinuing is it will no longer get

0:57:29any guide updates or anything so you can

0:57:32it'll you know you can still watch shows

0:57:34that are recorded on it but you can't

0:57:35recording new ones because I don't know

0:57:37anything is on and it doesn't work like

0:57:39a VCR we can do manually anyway and it's

0:57:43because they're changing the format of

0:57:44the guy data probably modernizing and in

0:57:46some way that we would be horrified to

0:57:48learn about if we knew exactly what they

0:57:50were feeding because of these teams

0:57:52here's one here's one was an analog

0:57:54device that could have recorded video

0:57:56from analog sources it didn't it was the

0:57:58pre-digital and of course standard def

0:58:00is all analog I didn't know that you

0:58:03could still use achieve a serious one I

0:58:06thought they were already non-functional

0:58:08entirely

0:58:09apparently can't get like semaphore

0:58:11flags like how does that even work

0:58:13if you have a TV series 1 and your

0:58:15sixteen-year-old hard drive or maybe a

0:58:17hard drive you replace it because you

0:58:18into third party hard drive replacement

0:58:19or whatever we're still using it to

0:58:21recall to add to record an expander does

0:58:23TV there terminating your service this

0:58:27is even if you bought quote-unquote

0:58:28lifetime service that's what the story

0:58:30about this the trees like oh I want

0:58:31lifetime service but now you know I'm

0:58:35still alive and the box is still alive

0:58:37but somehow the services ending so the

0:58:39tweet was 20 series one lifetime and

0:58:41scare quotes service last about 16 years

0:58:43and for this inconvenience tivo is

0:58:46offering your $75 prepaid Visa card so

0:58:49you get 75 box of you if you held on and

0:58:51kept using your turbo series 1 for this

0:58:52long and I say there are 3500 of the

0:58:55series one stone use which is a small

0:58:56number but you know anyway I can't

0:59:00believe that there wasn't some fine

0:59:02print written to the lifetime service

0:59:03that says after 16 years we reserve the

0:59:06right to turn off the service maybe you

0:59:07should buy a new DVR but I salute the

0:59:09people are out there still using it

0:59:11I never expected lifetime survive life

0:59:14controls an almighty boxes but I

0:59:15replaced the boxes long before i get to

0:59:17the lifetime service thing we used to be

0:59:19better i letting you transferred

0:59:20lifetime service from some nominal fee

0:59:22or free to another device to encourage

0:59:23you to upgrade or whatever but those

0:59:25days are mostly gone the deal something

0:59:27getting worse

0:59:28anyway I just thought this was funny

0:59:30that they're actually while they're

0:59:31still using it a series 1 DVR I don't

0:59:33think I have any piece of electronics in

0:59:36my house that is 16 years old and still

0:59:38working including smoke detectors

0:59:40telephones microwave maybe I think my as

0:59:45discussed generation reconcile with

0:59:46oversized microwave maybe more than 16

0:59:48years old but I should look at the date

0:59:50on the back of I should see exactly how

0:59:51old that it's better than that no DVR so

0:59:55that's for sure

0:59:55fair enough we had spoken a few episodes

0:59:59ago

0:59:59ago

1:00:00though about Twitter verification

1:00:02because all three of us and verified and

1:00:05a friend of the show brand-new has been

1:00:08verified now in the last week or so and

1:00:12apparently it's made everything in men

1:00:15sleep better

1:00:17Brianna tweeted a couple days ago to

1:00:19record this fact i have not seen a

1:00:21single death threat or rape threats

1:00:22since being verified and getting the

1:00:24quote quality filter quote everyone

1:00:26should have this funny how that works

1:00:29yeah that's exactly what I mean like

1:00:31like we we talked last time like I joked

1:00:33like we have computers now like the any

1:00:35could be any computational overhead of

1:00:36offering this to everybody like the idea

1:00:38that's totally worth it and we sense

1:00:41heard a little bit more discussion from

1:00:42various other places about it it seems

1:00:44like the the gist of the problem here is

1:00:46not that turn this on one be

1:00:48computationally expensive for everybody

1:00:50it's it's that Twitter it fundamentally

1:00:53twitter is very very tied to the idea

1:00:57internally politically that this should

1:01:00be this kind of open platform and that

1:01:02there should be no filters by default

1:01:04and that's a really nice Theory so is

1:01:07communism but in practice these things

1:01:09don't look quite that well you know like

1:01:11it is really ideal and that would be

1:01:13nice and it worked that way for awhile

1:01:15and that's cool but you know the reality

1:01:18is the harassment problem is very big

1:01:20and very real and that wonderful world

1:01:23where anybody could be anybody in it and

1:01:25they would see it by default i think

1:01:28that's a nice idea like blog comments

1:01:30like that's a nice idea but you know in

1:01:32practice it doesn't work in practice

1:01:34lots of problems with that so I maintain

1:01:37that this should know that the the

1:01:40quality filter feature that is available

1:01:42to verify Twitter accounts not only

1:01:44should be available to everybody but i'd

1:01:45even say it should be on by default we

1:01:48have not agree with that yeah I don't

1:01:50know why you wouldn't have it on by

1:01:51default i guess the idea that maybe the

1:01:53filter is bad and things will

1:01:55accidentally get filtered out but like

1:01:56it's not like you

1:01:58it's like gmail has spam filtering off

1:01:59by default of course right yeah

1:02:01sometimes things end up in your spam

1:02:04folder when they shouldn't be in there

1:02:06but like no one would say the correct

1:02:08default is off

1:02:09yeah I don't get it finally in follow-up

1:02:13and over an hour in a few people have

1:02:17taken to task for saying in the last

1:02:19episode of me was one reason episodes

1:02:22that I was not satisfied with apples

1:02:25diversity numbers particularly with

1:02:27ethnicities and and how many white

1:02:32employees apple has as opposed to other

1:02:35races and what I said was you know hey

1:02:38this is really unfortunate or in and it

1:02:41looks like as I look at these numbers

1:02:43again fifty-six percent of apples white

1:02:45as a self-report today and a lot of

1:02:47people said in various degrees of

1:02:49obnoxiousness what's the appropriate

1:02:51amount of diversity because the country

1:02:52is I don't have those numbers in front

1:02:54of me but something like sixty or

1:02:56seventy percent white so if they're less

1:02:59white than America then what's the

1:03:02problem and I don't have any simple

1:03:05answer for that it's it's something I

1:03:07hadn't considered and it is a fair point

1:03:08but i would i would point out that not

1:03:12all of apple's in America and so that

1:03:15probably should be factored in as well

1:03:16but i think the the biggest thing to me

1:03:21was that Apple seems to have made not a

1:03:25lot of change from just a couple of

1:03:27years ago when they started reporting

1:03:29this which I admire they didn't make

1:03:31that much change yet it seems like the

1:03:35thing that was completely unacceptable

1:03:37couple a couple years ago is now look at

1:03:39us and how wonderful we are and that

1:03:41just seems kind of weird to me i don't

1:03:45know if you guys had any thoughts on

1:03:46this but that that's basically the thing

1:03:48that I still stand by which is it hasn't

1:03:52gotten that much better why are we

1:03:53really celebrating it like it's

1:03:54completely turned around so before we

1:03:57let John tell us the truth and why this

1:03:58is you know before the John basically

1:04:01like rock this and and do everything

1:04:03correctly i will just kind of speak for

1:04:05for myself and for i think a lot of

1:04:07people who listen to this kind of stuff

1:04:09and you see these kind of things I know

1:04:12that be having been just in the in the

1:04:15area of people discussing this for a

1:04:17while now I know now that it is like it

1:04:22is

1:04:23never that simple and so it like when

1:04:25you when you throw a stat like well you

1:04:27know America's x percent white and so

1:04:29that's fine like it is usually not that

1:04:32simple and so I know to keep my mouth

1:04:35shut with things like that because I

1:04:37know like I hear someone that I'd be

1:04:38like yeah that might make logical sense

1:04:40when you read it like 140 characters but

1:04:43I bet there's more to the issue than

1:04:44that and so it's important when you are

1:04:48responding to people not UKC but youyou

1:04:50the public and I'm talking myself to

1:04:52like don't jump to a conclusion that

1:04:55because chances are the issue is way

1:04:56more complicated than that

1:04:58so John tell us why I was gonna start

1:05:02getting anything but i'm i'm basically

1:05:03on what you just said that's a that's

1:05:06something I always notice on Twitter

1:05:08when we all notice that to some degree

1:05:09or another whether we're doing it

1:05:11ourselves we see other people doing it

1:05:12where people will start from the premise

1:05:18and then start throwing out facts and

1:05:19support of that and it's fine right

1:05:22because why wouldn't you you know you

1:05:24this is what you think and so you will

1:05:26find facts that support what you think

1:05:27of when you will list those facts but so

1:05:30is important to kind of ask yourself why

1:05:34like why am i seeking out the fact that

1:05:37i can find that align with this thing

1:05:39that I already think like am I actually

1:05:41thinking about the situation where I

1:05:43merely just trying to find something

1:05:44that was formerly so that maybe I mean

1:05:47it's you know it's like university but

1:05:49perhaps someone who was telling Casey

1:05:51hey you you said this percentage is bad

1:05:53but it's about the ratio of the you know

1:05:57the people united states its you have to

1:05:59go look up the ratio of the people

1:06:01United States you could do is this a bad

1:06:02example because they're probably

1:06:03completely intellectually honest but a

1:06:04lot of times it's easy to see an issue

1:06:05come up and people immediately go

1:06:09looking for something making throughout

1:06:11that supports their thing never really

1:06:13asking why am I so desperate to find

1:06:15things to that support this rather than

1:06:17you know looking at the looking at the

1:06:19issue from first principles are

1:06:21understanding that the wider context

1:06:22because they'll always be things

1:06:23especially you know statistics that can

1:06:25help support one side of the other but

1:06:28if you're only ever looking for and

1:06:29saying repeatedly to anyone who mentions

1:06:30anything about the topic

1:06:32the few facts you have in your back

1:06:33pocket that support the one thing that

1:06:35you care about you probably are missing

1:06:37the larger issue and what you really

1:06:38missing more is why is it so important

1:06:40to you that your current view of this

1:06:42exactly correct

1:06:44so anyway as for the specific stats i

1:06:47don't know i thought the stats were just

1:06:49us only so i think their percentages for

1:06:51you know the white employees that are

1:06:53actually accurate to the United States

1:06:56the United States has that percentage of

1:06:57the population and that's their you know

1:06:58so i thought that was an apt comparison

1:07:00but like Marco said it's never quite

1:07:02that simple because you know especially

1:07:04with statistics you can

1:07:06they're taking like all their employees

1:07:07right and all us employees Apple employs

1:07:11a lot of people and are not all those

1:07:13jobs are the same if you start looking

1:07:15at different sections of employees the

1:07:17percentages I think they do this don't

1:07:18break it down the page and the the

1:07:20percentages will change you know pretty

1:07:23dramatically by five to ten percent

1:07:25maybe even more if you look at like the

1:07:28top level of the org chart

1:07:30director-level c-level executives they

1:07:33are not 56% white right so and then all

1:07:38the way down to look at engineering vs

1:07:40QA vs retail versus you know all the

1:07:42different any other different

1:07:44categorizations if you want them all

1:07:46together it's nice to get a nice a great

1:07:47but you don't know what the pieces are

1:07:48right and so that that's not apples not

1:07:51trying to mislead anybody there they're

1:07:53not going to break down there they're

1:07:55hiring in their employees by individual

1:07:57job level and location and and what what

1:08:00is the population in the state of

1:08:01Maryland and one of the retail boys in

1:08:03the state of Maryland and what about you

1:08:04like it's just they're not going to do

1:08:06that right but that is Apple's job

1:08:07internally to work on this problem

1:08:10because if their overall numbers

1:08:12happened to work out to the same

1:08:14percentage of the United States but it's

1:08:16only because like all the leaders of the

1:08:18company or white men and all the rest of

1:08:20the company or not that's not that you

1:08:23not achieving your goal of diversity

1:08:25write your goal is not to make a bunch

1:08:26of numbers match a bunch of the numbers

1:08:28you can always pick numbers to make it

1:08:29look better or worse and some degree

1:08:31apple may be doing that in the

1:08:32high-level numbers that are up there but

1:08:34that's not the goal of the thing you

1:08:35you're not Apple is not an amorphous

1:08:37blob of people are combined into one

1:08:39google person they're a bunch of

1:08:40individual people and you want to know

1:08:43if you know if I get a job at out

1:08:45what are my chances of you know being in

1:08:48charge of all software an apple

1:08:50what are my chances of raising a rising

1:08:52to the level of the org chart where I'm

1:08:53taking meetings with the with tim cook

1:08:55and stuff what you know what are my odds

1:08:57of becoming a an engineering manager or

1:09:00leading whatever and if you look around

1:09:01you an apple and it's like gee everybody

1:09:02who's engineering manager is a white

1:09:04dude and everyone in the board meetings

1:09:07as a white dude and everybody was

1:09:09onstage WBC's a white dude as as you

1:09:11know to a first approximation that was

1:09:13the case

1:09:14not too many years ago and you know like

1:09:16that even now that file the company like

1:09:18that maybe you feel like oh sure I can

1:09:20get a job at an apple store they're good

1:09:22at it but like can I you know it

1:09:25do I feel like this company is a is

1:09:29giving equal opportunity to everybody

1:09:30and that's what Apple is working towards

1:09:32and I don't think they have yet achieved

1:09:33it was about to be putting out these

1:09:35numbers so that case he said it were if

1:09:40these numbers were given these big

1:09:41aggregates and they haven't changed that

1:09:43much from from year to year two shows

1:09:45they're making slow progress right and

1:09:47so it's not phenomenal progress but it

1:09:49certainly doesn't mean because you can

1:09:50say fifty six percent less than sixty

1:09:52percent therefore there done that

1:09:53absolutely doesn't mean that because it

1:09:55doesn't take much thinking to realize

1:09:57well that's just a bigger your number

1:09:58and it's not that's not the goal the

1:10:00goal is not to simply people are asking

1:10:01that i can honestly asking like what is

1:10:03the goal what numbers you supposed to

1:10:05see here to let you know that Apple has

1:10:07achieved its goal of diversity and the

1:10:09answer is those numbers are never going

1:10:11to tell you that right otherwise you

1:10:13know they wouldn't they wouldn't be

1:10:14continuing along this path they were

1:10:15just say when we're done we've

1:10:16accomplished our goal is not great right

1:10:18and that's before you even get into the

1:10:21office of like percentage of US

1:10:22population versus percentage of US

1:10:24population of working-age like not

1:10:26babies not children not retired people

1:10:29you know so it's numbers are complicated

1:10:31but I anyway I'm i am confident that

1:10:34Apple understands the problem before

1:10:35them and is working towards it just just

1:10:38as I am not particularly tied to the

1:10:41specific numbers that throughout on

1:10:42these pages are basically looking for is

1:10:43our little grass they put up going in an

1:10:46upward direction

1:10:47do they have enough of them every year

1:10:48so they show their making progress and

1:10:50how fast that progress one of the slopes

1:10:52look like yeah that's about it

1:10:57alright well we are out of follow-up

1:11:00thanks a lot 43 sponsors this week

1:11:02betterment tracker and automatic and we

1:11:04will see you next week

1:11:08now the show is over they didn't even

1:11:12mean to be in as it was accidental death

1:11:16was accidental

1:11:18John didn't research Marco and Casey

1:11:22would let me

1:11:24because it was accidental was accidental

1:11:28and you can find the show know today p

1:11:33dot and it for twitter follow them

1:11:41yes byl ISS so that's Casey list and a

1:11:45co-pay rm20 Marco Arment and our AC

1:11:53syracuse it

1:12:05don't want to talk about the Intel fan

1:12:10thing that's topic can we make that

1:12:13quick

1:12:15mmm yeah exactly

1:12:18maybe I don't do it let's take down what

1:12:21we have here is where we have an

1:12:22after-show then we can do Instagram

1:12:23stories and stories good ideas that's

1:12:25fluffy

1:12:26yeah it's it's this thing that appear on

1:12:28top of Instagram recently that's

1:12:30confusing all the old people like me

1:12:31you're not that old

1:12:33your wife's face is always up there she

1:12:37figured out very quickly because she is

1:12:40young at heart and also biologically

1:12:42then I am I am 2014 13 pounds i think

1:12:47like six about sex yet so I i am too old

1:12:52to understand snapchat so instagram has

1:12:55basically cloned a a major part of

1:12:58snapchat into the Instagram interface

1:13:00which has just confused

1:13:02instagram for me and now I have this row

1:13:05of heads on top of instagram i tap them

1:13:07and weird things happen and I don't

1:13:09really know what I'm supposed to be

1:13:10doing with this can you explain to me

1:13:11so John a year i don't think i've seen

1:13:14any post from you yet so you're not

1:13:16producing using this although i presume

1:13:19you're at least consuming it is that

1:13:20fair

1:13:21yeah I don't see myself wearing one of

1:13:23these videos but they're up there on the

1:13:24top of the faces

1:13:25ok so the idea here from what I can

1:13:29understand someone who's never used

1:13:30snapchat and didn't really have any

1:13:32interest in snapchat is with instagram

1:13:35stories you can take a photo or a video

1:13:40that is semi ephemeral so I believe it's

1:13:44after a day they will self-destruct and

1:13:47your phone explodes it's very

1:13:49inconvenient now they will self-destruct

1:13:50and then they won't be available anymore

1:13:53not to you not to anyone else and so the

1:13:55idea is if your Instagram profile in the

1:14:00photos that you post on Instagram proper

1:14:02is the like super stage super deliberate

1:14:06super-serious version of you

1:14:08instagram stories which is this

1:14:10ephemeral thing is more of the candied

1:14:14fun-loving like a here's here's me

1:14:16here's the real me

1:14:18take it or leave it sort of thing and I

1:14:20actually wanted to call out the latest

1:14:23episode of connected which is episode 13

1:14:25the hosted a really incredible job of of

1:14:30talking about kind of what the

1:14:31motivations are behind Instagram stories

1:14:33and how it's different than than regular

1:14:36Instagram but regardless I've really

1:14:40been enjoying it a lot more than I

1:14:42thought I would as someone who had not

1:14:44even the slightest interest in snapchat

1:14:46and one of the common complaints about

1:14:48snapchat that I've understood is that

1:14:50anyone who is over the age of about

1:14:52eighteen finds the interface completely

1:14:54inscrutable and everything's

1:14:56gesture-based and as much as we all hate

1:14:59tutorials as you install an app for the

1:15:01first time there this is one of those

1:15:04instances where a tutorial or

1:15:05walkthrough or onboarding would have

1:15:08perhaps been useful in snapchat because

1:15:11everything's gesture-based nothing

1:15:12obvious everything's weird Instagram

1:15:15there's a little bit of that Instagram

1:15:16stories but it's not nearly as bad and i

1:15:19was able to figure it out pretty quickly

1:15:21and I'd classify myself as an old man

1:15:24just as much as you two are and I've

1:15:27been enjoying both consuming and

1:15:28creating one of the things I've wondered

1:15:31as I consider whether or not to post a

1:15:33new entry in my Instagram story which

1:15:35sounds super cheesy but anyway one of

1:15:38these have been considering is why would

1:15:40anyone else give a crap about this so

1:15:42like as an example i was going to take a

1:15:44picture my setup at work and you know

1:15:47after having obscured that the sensitive

1:15:49parts of what was on my computer screen

1:15:51and then it occurred to me i don't think

1:15:53anyone really gives crap and so while i

1:15:55was at the beach you know it i'm on

1:15:57vacation we that's exciting and kinda

1:15:59cool or whatever I i posted a handful of

1:16:02times then but now that I'm back in my

1:16:05normal grind well it's not my normal

1:16:07high roller

1:16:08you know look at me I'm so awesome life

1:16:10it's my normal grind so I don't know if

1:16:12I'll be producing that much for posting

1:16:15that often of from now until the next

1:16:18time I do something interesting but i

1:16:21really liked the idea and I echo what

1:16:23Mike it said on unconnected with a nice

1:16:27talk about it and he and I both talked

1:16:28about this on analog as well that

1:16:30Instagram is kind of my happy social

1:16:32network i get a lot out of out of

1:16:35Twitter and I enjoy twitter but Twitter

1:16:37is nevertheless kind of a dumpster fire

1:16:41where I cannot remember a time that

1:16:44instagram has made me anything but happy

1:16:47and even with stories it makes me happy

1:16:50in a different way because it's kind of

1:16:51the fun ephemeral you know hey it's not

1:16:54gonna be perfect and that's okay because

1:16:55life isn't perfect and and it's brought

1:16:58and this is a super sleazy business

1:17:01eater but it's brought my engagement

1:17:02with instagram up even more and I check

1:17:04Instagram a handful times a day because

1:17:06I enjoyed so much but now I found myself

1:17:08checking it more often wondering which

1:17:10one of my friends because I tend to

1:17:11follow mostly friends on Instagram which

1:17:14one of my friends posted something new

1:17:15and something worth looking at and so I

1:17:17have really been enjoying it two thumbs

1:17:20up from me

1:17:21I think it's super super interesting and

1:17:23it seems to have traction so far don't

1:17:26think it's a flash-in-the-pan this isn't

1:17:29the next peach i don't think but remind

1:17:32me that in six months is physically yeah

1:17:34i mean like this i have i've been using

1:17:37Instagram or more recently been using

1:17:39the Explorer view a little search thing

1:17:40at the bottom which to pinterest me too

1:17:42and I didn't even realize that's what

1:17:44that was there I thought it was

1:17:45literally a search and I never wanted to

1:17:47search for anything so I never visited

1:17:49that tab and it turns out it's more like

1:17:51a browser Explorer view and you kinda

1:17:54let you know you like you pick topics so

1:17:55mine's like full of watches and puppies

1:17:57now which is awesome and I didn't know

1:17:59that either actually it is really nice

1:18:02to me i'm using Instagram Milano and I

1:18:04have the same things as you like

1:18:05Instagram is is just feel good place it

1:18:08is both to browse and to post it is a

1:18:11just nice place to be

1:18:12people aren't universally terrible on my

1:18:15Twitter well i'm in general like I'm

1:18:18saying even including all services that

1:18:20we have people in general are not

1:18:22terrible

1:18:23the design of a social thing online and

1:18:27social constructs social service like

1:18:28the whole design of it the structure

1:18:31what's allowed what's possible how

1:18:34things work who sees what that all

1:18:37effects how you how horrible people are

1:18:41are enabled to be and how and what kind

1:18:44of impact

1:18:44they're horribleness can have so the

1:18:46same like the same group of people can

1:18:49use two different services and act very

1:18:51differently on them and get very

1:18:53different things out of them because

1:18:54because of differences in their design

1:18:55and so even if the exact same even if

1:18:58Twitter's entire user base is also

1:19:00facebook is also instagrams exactly is

1:19:02your base you have better results on

1:19:04instagram because just the form and the

1:19:07structure and like the way everything

1:19:08works in a set of expectations of of it

1:19:11and the format and everything just

1:19:12encourage better behavior and limit the

1:19:15places people can be horrible to you

1:19:17so that's good what worries me about

1:19:20this and you know and I want to

1:19:22interview snapchat was you know the end

1:19:24again yet that conversation connected

1:19:25was awesome and willing to it one of the

1:19:28reasons i never had to get into it was

1:19:30because no one that I know really uses

1:19:33it like all the people i would be

1:19:35interacting with our on twitter and

1:19:36instagram so I don't need to use first

1:19:39of all don't use facebook which is great

1:19:41because i want to use facebook and i

1:19:43also and yes i know they own it and it

1:19:44doesn't matter and i also never had to

1:19:47use snapchat because just you know my

1:19:49people aren't there really so to bring

1:19:52that to Instagram to basically bring

1:19:54like a major feature of snapchat to

1:19:56Instagram is kind of okay

1:20:00on one level because it's like well I

1:20:02was gonna use snapchat anyway because my

1:20:05friends weren't there but on the other

1:20:07hand it really shows what you know

1:20:09Facebook has always been a very

1:20:10tasteless company

1:20:12facebook is pragmatic and ruthless and

1:20:14tasteless just like Microsoft was in the

1:20:17eighties and nineties really facebook

1:20:19now is this this conglomerate really

1:20:21they have tons of properties that are

1:20:24major on the internet they have they

1:20:26have Facebook itself which is probably

1:20:28the biggest web property in the world

1:20:31and probably will stay that way for

1:20:32quite some time

1:20:33anything that gets really big that they

1:20:35think might threaten that they just buy

1:20:37it they have enough money to keep doing

1:20:38that indefinitely and they are clearly

1:20:42totally okay with like well if you won't

1:20:45let us buy you were just gonna rip you

1:20:46off and we'll just kill you that way and

1:20:49that strategy doesn't always work it

1:20:52with you know Microsoft the same thing

1:20:54you know it

1:20:55Apple has done things that like on a

1:20:57smaller scale sure locking things but

1:20:58they don't usually do a huge scale job

1:21:00of it stuff like this google does it

1:21:02sometimes and that strategy tends to

1:21:05work like in general when companies when

1:21:08like the big powerful companies that

1:21:09already have all the users and all the

1:21:12usage and all the attention going to

1:21:14them and all this time spent on their

1:21:15services if they clone some smaller

1:21:18service there clone often wins and if

1:21:22they can't clone it they can buy it and

1:21:24this just kinda it's a bit of a warning

1:21:26sign of the power that Facebook has and

1:21:30i'm not saying snapchat will be killed

1:21:31by this in all likelihood snapchat is

1:21:33too big now and they'll probably be

1:21:35totally fine because that's really big i

1:21:37think they're bigger than Twitter

1:21:38they're really big but it does kind of

1:21:42scare me a little bit that facebook is

1:21:44willing to so so cavalierly just like

1:21:47completely rip off the stories feature

1:21:50called the same thing do the same things

1:21:52use the same gestures like completely

1:21:54rip it off it's so closely and so

1:21:57shamelessly and they just couldn't

1:21:59possibly care less because they are just

1:22:00that ruthless and we we live in an age

1:22:04now where these these these handful of

1:22:06big internet companies have a lot of

1:22:09control over us a lot of resources and

1:22:13like the centralization of power here is

1:22:16just getting more and more severe and

1:22:19that kind of worries me for the future

1:22:20of like internet things general things

1:22:23we do online that that use the internet

1:22:25that involve the internet or are the

1:22:26internet it is a little concerning thing

1:22:30that struck me about using this and

1:22:31having new snapchat a little bit is the

1:22:35one you often hear and you guys said

1:22:37before i go well you know I don't

1:22:39understand that put on the kids

1:22:40understand it or whatever I did yeah

1:22:42there is the demographic thing about

1:22:44what age groups use which application

1:22:45that's definitely true and more older

1:22:47people use Instagram or whatever but

1:22:50snapchat and also the extreme stories

1:22:53have bad interfaces they have bad user

1:22:57interfaces like there is no sensible

1:22:59structure to to the interface they don't

1:23:03use OS conventions that's for sure

1:23:05right so the normal affordances and sort

1:23:07of the things you're used to in

1:23:08generic uikit you know I don't know

1:23:11about another problem just talked about

1:23:12in iOS but like it doesn't look like iOS

1:23:13that's for sure

1:23:14right so it's kind of its own thing and

1:23:16within its own thing they don't really

1:23:18define any particular spatial metaphor

1:23:22or something to hang your hat on where

1:23:24you can say oh I see all of swipes to

1:23:26this i mean like basically if you you

1:23:28know if someone's like oh snap just only

1:23:30sensible what you'd basically be doing

1:23:32is listing this does that this does that

1:23:34this doesn't just like that there now

1:23:35and parts that's not a system that is

1:23:37just a list of actions and the resulting

1:23:39you know what you're looking for a

1:23:40system is like because I understand the

1:23:42system before you tell me the list of

1:23:44all things i can do i can predict oh

1:23:46then obviously the way to do X is going

1:23:48to be why that's when you know you have

1:23:50a system snapchat doesn't have a system

1:23:52or not a very good system it's a lot of

1:23:53arbitrary crap and they're not using

1:23:55native controls not using serve

1:23:57unprecedented right sort of setting

1:23:59their own standard and why does does

1:24:03that mean is that is that why young

1:24:05people get snapped and people don't know

1:24:07like there are other factors there in

1:24:08terms of what's advertised towards into

1:24:10the first users were and you know

1:24:11critical mass and social networks and

1:24:13look where my friends are all things

1:24:15that all factors that as well but I

1:24:16think there is something to the idea

1:24:18that young people will be full to do

1:24:22something there's a bunch of their

1:24:23friends are doing it they'll just figure

1:24:26it out

1:24:26they have a high tolerance for learning

1:24:28RG arbitrary stupid crap i think about

1:24:30all the arbitrary stupid crap that I

1:24:32learned which basically is the entire

1:24:33early years of computing where

1:24:35everything was terrible and nothing made

1:24:37sense and the mac was the only one had

1:24:38kind of an understandable system

1:24:40think of all the things that i memorize

1:24:41things i learned how to do in video

1:24:42games there was no system is arbitrary

1:24:44and kids have a lot of time and you're

1:24:47good memories and they'll just plug away

1:24:49until it becomes second nature to them

1:24:51so the kids snapchat makes perfect sense

1:24:53because they memorize all the the stupid

1:24:55gestures and commands even those

1:24:56gestures and commands are almost

1:24:58entirely arbitrary countered everything

1:25:00that every other UI paradigm the device

1:25:02they're holding has taught them and are

1:25:04just plain bad right

1:25:06they're just bad this is the bad you iso

1:25:08Instagram stories didn't use all the

1:25:10same things that snapped up but it's a

1:25:12similar thing we're Instagram was a very

1:25:13sensible straight up the middle iOS

1:25:15application with the reasonable UI using

1:25:17metaphors and controls and

1:25:19the interfaces that we understood so you

1:25:21can predict what would happen when you

1:25:22did thanks for the most part is always

1:25:23the weird edge cases and custom controls

1:25:25and stuff but then related stories it's

1:25:28in la-la land again where nothing makes

1:25:30sense and you can't you have no you have

1:25:32no nothing to hang your hat on except

1:25:34for like maybe it will behave like

1:25:36snapchat which again maybe that's

1:25:38establishing a standard maybe you're

1:25:39familiar with snapchat that might be a

1:25:40good way to go if you're trying to do

1:25:41like what marco said like oh we want to

1:25:43do a snapchat like then just copy

1:25:45snapshots gestures because all the kids

1:25:46who snapchat it makes perfect sense of

1:25:48them and they go do your thing and

1:25:49Swiper drag or tap and it doesn't do

1:25:52what they expect from using snapchat

1:25:53they think your thing is broken but

1:25:55again it's not a good you I just because

1:25:57they make it like snapchat for like or

1:25:59you know making similar snap try to get

1:26:01a little bar they've got on the top with

1:26:04a little segment and timers and what

1:26:06happens when you swipe in different

1:26:07directions and the little rotating

1:26:08effect versus the sliding vs up and down

1:26:11and tapping and it's just these are bad

1:26:14interfaces and so I maintained as the

1:26:16old cranky person but not only the

1:26:19actual reason like old people don't

1:26:21understand snapchat because they made a

1:26:23bad UI and because older people have

1:26:26less tolerance for figuring out a bad

1:26:28you want it doesn't look like the other

1:26:30ones doing it and I don't just think

1:26:32it's like it i think it's bad on many

1:26:34many levels and again this doesn't

1:26:35explain why old people don't use that

1:26:37there's so many other reasons but this

1:26:38is a factor I'm saying this is this is a

1:26:40factor in that giant students not the

1:26:41biggest factor is not you know even

1:26:43maybe a major factor but it's there and

1:26:45that annoys me because I feel like

1:26:47snapchat could have been successful with

1:26:50a clever innovative you i think a lot of

1:26:53games do this like you wanted to use no

1:26:55native controls totally custom doesn't

1:26:57use anything you like it is the same on

1:26:59Android and iOS but is delightful and

1:27:01understandable and fun to use and

1:27:04someone can pick up and learn you again

1:27:06at this is my hobby horse but it's

1:27:08really true if you make some kind of

1:27:10physical world based metaphor like I

1:27:13have stacks of cards and we can slide

1:27:14them side to side and up and down and

1:27:16bring them together and collapse them or

1:27:18i can crunch them up or like if you make

1:27:20some sort of thing like these aren't

1:27:21physical things but in some ways they

1:27:24behave and physical things are the is

1:27:25like a larger map that you can't see of

1:27:27this application once you learn the

1:27:29layout of that Matthew know how to move

1:27:31around it and you know how to make

1:27:32things happen and you know what a button

1:27:34looks like and you understand like any

1:27:36kind of metaphor like that even if he

1:27:38uses totally underrated non-native

1:27:40controls can become understandable when

1:27:42people start using it

1:27:43learn one or two things that all i see

1:27:45how this works and they can figure out

1:27:47how to do the other things without being

1:27:48told explicitly whereas snapchat like

1:27:51all mystery meat navigation a bunch of

1:27:53inscrutable icons and no real sense of

1:27:56where you are what you're doing where

1:27:57you going to just memorize how it works

1:27:59because you're young and you have a lot

1:28:00of time or use the application a lot and

1:28:02that becomes second nature you just

1:28:03think that's normal

1:28:04so if you ask any kid snapchat

1:28:06easy-to-use they like everyone knows how

1:28:07to use snapchats like i like i know how

1:28:09to write my name and I put one foot in

1:28:10front of the other walk and you snapped

1:28:12as the most intuitive interface on my

1:28:13entire phone

1:28:14no just use it a lot that's what that

1:28:16means that doesn't mean it's a good UI

1:28:17and the reason they like using it is

1:28:20because it makes people like us that

1:28:22angry not but that's not why where I'm

1:28:24about like I'm just saying it's a bad

1:28:25you I you angry about it because it's

1:28:27like you're making silly videos of

1:28:28yourself and send your friends and like

1:28:29all these kids these days shouldn't be

1:28:31taking video I don't care about that

1:28:32whatever take videos you live in the

1:28:34same thing when we take videos and our

1:28:35things and send our friends that's fine

1:28:37like whatever that I have no I have no

1:28:39anger towards people use snapchat I

1:28:41hangers the developers of snapchat for

1:28:43making a bad you I that users thumbs-up

1:28:48snapchat developers thumbs down