ATP

00:00:00 ◼   ► - So where's my new MacBook Pro, man?

00:00:02 ◼   ► - Looks like we're not gonna get it this year.

00:00:04 ◼   ► If we haven't gotten it yet,

00:00:05 ◼   ► they just did this iMac update.

00:00:07 ◼   ► - Yeah, I feel like I've been hearing from various sources

00:00:11 ◼   ► that it's coming, it's coming, it's coming, it's coming,

00:00:13 ◼   ► any second, any second, any second, any second,

00:00:15 ◼   ► but I don't know.

00:00:16 ◼   ► - Yeah, the 15-inch MacBook Pro is,

00:00:20 ◼   ► it's such the workhorse of the industry.

00:00:23 ◼   ► It gets so little attention, relatively speaking,

00:00:25 ◼   ► like in media coverage and people like us talking about,

00:00:29 ◼   ► well, not us, but most other podcasts talking about it,

00:00:34 ◼   ► talking about laptops and everything.

00:00:35 ◼   ► Everyone's talking about the new MacBook

00:00:37 ◼   ► or new MacBook Airs or whatever.

00:00:38 ◼   ► But when I go to conferences and events

00:00:42 ◼   ► or I go to somebody's office

00:00:44 ◼   ► and I see other developers working,

00:00:46 ◼   ► by far the most common machine in use

00:00:49 ◼   ► is the 15-inch MacBook Pro.

00:00:51 ◼   ► - Completely agree.

00:00:51 ◼   ► - Like it really is like the quiet workhorse

00:00:54 ◼   ► of the entire industry.

00:00:56 ◼   ► - Yeah, basically if you're doing something

00:00:57 ◼   ► that requires any more horsepower

00:01:00 ◼   ► than a MacBook Air can give you,

00:01:02 ◼   ► then immediately jump directly to 15-inch MacBook Pro.

00:01:06 ◼   ► Do not pass Go, do not pass 13-inch, do not collect $200.

00:01:10 ◼   ► That's a reference, John.

00:01:11 ◼   ► (electronic beeping)

00:01:13 ◼   ► - Last week, I brought up,

00:01:15 ◼   ► and we all talked about the thing

00:01:18 ◼   ► where the iPhone 6S and 6S Plus CPUs

00:01:21 ◼   ► were being manufactured by two different companies,

00:01:22 ◼   ► Samsung and TSMC, and some early benchmarks at the time

00:01:25 ◼   ► showing that the TSMC one got significantly better battery life than the Samsung one.

00:01:30 ◼   ► It was something on the order of like 15 to 20 percent better battery life under certain

00:01:34 ◼   ► CPU stressing benchmarks.

00:01:37 ◼   ► And this briefly became a thing and Apple kind of squashed it by issuing a PR statement

00:01:43 ◼   ► to a bunch of websites that basically said, it was the most Apple-y statement ever, it

00:01:51 ◼   ► It basically said nothing about whether it was true or not,

00:01:56 ◼   ► but it sounded like a denial.

00:01:58 ◼   ► And the reality of what they said is that,

00:02:02 ◼   ► yes, there is a difference,

00:02:04 ◼   ► but they say it's within 2% to 3%,

00:02:07 ◼   ► not 15% to 20%, under their testing,

00:02:10 ◼   ► and that the testing that revealed this larger difference

00:02:13 ◼   ► was considered invalid or unimportant

00:02:16 ◼   ► because it only showed basically maxing out the CPU.

00:02:21 ◼   ► And so if you are doing things that max out the CPU,

00:02:25 ◼   ► then there is a noticeable difference.

00:02:28 ◼   ► If you are not maxing out the CPU,

00:02:29 ◼   ► and most people are not maxing it out most of the time,

00:02:32 ◼   ► then the difference still exists but is smaller.

00:02:36 ◼   ► - I haven't seen enough really honest-to-goodness data

00:02:39 ◼   ► to make me feel strongly one way or the other.

00:02:42 ◼   ► Jon, what's your take on all this?

00:02:44 ◼   ► - I'm still kind of interested

00:02:45 ◼   ► and where the difference comes from.

00:02:47 ◼   ► I don't know enough about the Geekbench thing.

00:02:49 ◼   ► And of course, Apple statement is not very specific.

00:02:51 ◼   ► They think that they, they poopooed the,

00:02:54 ◼   ► that benchmark by saying that it spends

00:02:56 ◼   ► an unrealistic amount of time

00:02:58 ◼   ► at the highest CPU performance state.

00:03:00 ◼   ► Now, do they mean CPU as in the whole system on a chip?

00:03:02 ◼   ► Do they mean just the CPU part of the system on a chip?

00:03:06 ◼   ► I don't know if Geekbench is exercising the GPU

00:03:08 ◼   ► at the same time.

00:03:09 ◼   ► Like what could account for the large difference?

00:03:13 ◼   ► If it's only two to 3% of regular tests,

00:03:15 ◼   ► is that just because, you know, in regular usage,

00:03:18 ◼   ► you aren't maxing the CPU all the time?

00:03:20 ◼   ► Some people brought up games,

00:03:23 ◼   ► like what if I'm playing a game

00:03:24 ◼   ► that is close to maxing the CPU

00:03:25 ◼   ► the entire time I'm playing?

00:03:26 ◼   ► Then does it come back into the 20 or 30% range

00:03:29 ◼   ► or does that not matter because the game stresses the GPU

00:03:33 ◼   ► and it turns out the one with the good, you know,

00:03:35 ◼   ► there's just not enough details here.

00:03:37 ◼   ► What I was trying to think of was like,

00:03:38 ◼   ► maybe one is better, uses less energy coming from

00:03:42 ◼   ► going into a low power state or, you know, spinning parts of the chip up or down and the other one,

00:03:47 ◼   ► you know, is better at sustained high CPU usage. But if you were to throttle on, off, on, off,

00:03:55 ◼   ► it would get worse for, I don't know. It's mysterious. All we have is Apple's word to go on.

00:04:00 ◼   ► I don't think anyone has done any real world test yet. You would think if it was 20 to 30%

00:04:05 ◼   ► in real world, we would know about it. Not from someone running a benchmark, but from actual

00:04:09 ◼   ► people saying, boy, my iPhone 6s is terrible.

00:04:13 ◼   ► And someone else saying my iPhone 6s is great.

00:04:15 ◼   ► And I mean, people are saying that anyway,

00:04:16 ◼   ► but they probably all have the same CPU.

00:04:20 ◼   ► Anyway, it's mysterious.

00:04:21 ◼   ► Apple is not going to really give us any information on it.

00:04:25 ◼   ► As Apple says in the statement,

00:04:26 ◼   ► every chip we ship meets Apple's highest standards,

00:04:29 ◼   ► blah, blah, blah, blah.

00:04:29 ◼   ► Like I said, they're gonna make,

00:04:32 ◼   ► I think it is literally true that in Apple's testing,

00:04:36 ◼   ► the worst chip, like the bad one or whatever,

00:04:38 ◼   ► in the Geek Bench Test passes Apple's criteria.

00:04:41 ◼   ► It still means that perhaps the other ones

00:04:44 ◼   ► exceed Apple's criteria, and if you get one of those,

00:04:46 ◼   ► you're kinda lucky, and so it is still kind of like

00:04:48 ◼   ► a lottery, but without more information,

00:04:50 ◼   ► and I'm not sure how we're gonna get more information,

00:04:52 ◼   ► without more information, I don't even know if you're lucky

00:04:54 ◼   ► unless you spend all day running that benchmark.

00:04:57 ◼   ► How lucky are you if you get the quote-unquote good one?

00:05:00 ◼   ► - That's the thing, like their statement really,

00:05:03 ◼   ► it wasn't actually, like everyone was kinda treating it

00:05:05 ◼   ► as confirmation that this is wrong.

00:05:08 ◼   ► But in fact, it was confirmation that it's right,

00:05:11 ◼   ► that there is a difference,

00:05:13 ◼   ► but you probably won't notice it in average use.

00:05:16 ◼   ► - And it wasn't techy, they didn't go into the details.

00:05:18 ◼   ► So they're just, it's essentially,

00:05:20 ◼   ► it's first, the fact that they made a statement at all

00:05:23 ◼   ► is interesting, but second, what their statement said is,

00:05:27 ◼   ► "We feel we are covered.

00:05:28 ◼   ► We feel everything we've said is true.

00:05:30 ◼   ► Even though there's a difference,

00:05:32 ◼   ► we think the difference is small

00:05:33 ◼   ► in real-world usage, the end."

00:05:35 ◼   ► So you're right, it's basically just a confirmation of the thing, but they're not going to go into super techy detail and

00:05:39 ◼   ► Bottom line as long as the iPhone 6s gets okay battery life compared to what the 6 got which it probably does because everyone's 6s

00:05:47 ◼   ► Are like a year old right who gets a 6s most people aren't gonna get a success a month after they get a 6

00:05:52 ◼   ► It's everything seems to be fine. Yeah, so it's a non-story, but from a technical perspective. I'm still very interested in

00:06:00 ◼   ► How exactly this synthetic benchmark?

00:06:03 ◼   ► found this, you know, this weakness in this one CPU.

00:06:07 ◼   ► Our first sponsor this week is Casper. Casper is an online retailer of premium mattresses for a fraction of the price.

00:06:15 ◼   ► Casper mattresses are quality. They are one-of-a-kind

00:06:19 ◼   ► combination of memory foam and latex foam to give you just the right sink and just the right bounce for better nights and brighter days.

00:06:29 ◼   ► This is a hybrid, latex foam with memory foam,

00:06:32 ◼   ► that it really does have the best of both worlds.

00:06:35 ◼   ► You get the comfort of memory foam,

00:06:37 ◼   ► but the latex foam kinda tames some of the downsides of it,

00:06:40 ◼   ► like the hotness that a lot of people feel with it.

00:06:43 ◼   ► These are incredibly good values also.

00:06:45 ◼   ► Premium mattresses, you're lucky to get out

00:06:47 ◼   ► of a premium mattress store getting something nice

00:06:50 ◼   ► that's like queen or king size.

00:06:51 ◼   ► You're lucky to get out there for under 1500 bucks.

00:06:55 ◼   ► You gotta try really hard.

00:06:56 ◼   ► Usually you're closer to 2000 bucks.

00:06:59 ◼   ► Now Casper mattresses start at just $500 for twin,

00:07:02 ◼   ► $750 for full, $850 for queen, and $950 for king.

00:07:06 ◼   ► This is really about half the price

00:07:07 ◼   ► that you're gonna pay most of the time

00:07:08 ◼   ► for any other reasonable mattress.

00:07:10 ◼   ► And they understand that buying a mattress online

00:07:13 ◼   ► kinda sounds weird.

00:07:14 ◼   ► They have you covered, this is completely risk-free.

00:07:17 ◼   ► They offer free delivery and free returns

00:07:20 ◼   ► within a hundred day trial period.

00:07:22 ◼   ► So all you gotta do, you get it,

00:07:25 ◼   ► you try it out in your own home for a hundred days.

00:07:27 ◼   ► They ship it right to you,

00:07:28 ◼   ► like this little tiny compressed box, it's pretty cool.

00:07:31 ◼   ► You can carry the box in yourself

00:07:32 ◼   ► 'cause it's about the size of like a dorm room fridge.

00:07:34 ◼   ► Then you open it up and it just expands out,

00:07:36 ◼   ► sucks all the air out of the room,

00:07:37 ◼   ► and there you have a mattress.

00:07:39 ◼   ► And just sleep on it for up to 100 nights.

00:07:44 ◼   ► And if you don't like it after 100 nights,

00:07:46 ◼   ► you can call them up and they will arrange

00:07:47 ◼   ► for a return for free.

00:07:49 ◼   ► It is that simple.

00:07:50 ◼   ► Now this is again, an obsessively engineered mattress

00:07:53 ◼   ► at a shockingly fair price.

00:07:55 ◼   ► I mean, 850 for a queen, that's amazing.

00:07:58 ◼   ► Just the right sync, just the right bounce

00:07:59 ◼   ► with latex foam and memory foam

00:08:01 ◼   ► with better nights and brighter days.

00:08:03 ◼   ► So once again, risk-free trials, risk-free returns.

00:08:07 ◼   ► Try it for 100 days.

00:08:08 ◼   ► And these are made in America.

00:08:10 ◼   ► Go to casper.com/atp to try this out for yourself.

00:08:14 ◼   ► Using our code, you will even get 50 bucks off any mattress.

00:08:17 ◼   ► Casper.com/atp.

00:08:19 ◼   ► Thanks a lot.

00:08:21 ◼   ► - So big week for Marco.

00:08:23 ◼   ► You released Overcast 2.

00:08:25 ◼   ► - I did, yeah.

00:08:26 ◼   ► I even stopped saying the point O in my marketing materials

00:08:29 ◼   ► because I thought it sounded better without it.

00:08:32 ◼   ► - So delightfully appley of you.

00:08:34 ◼   ► - Like when I leave the cents off menus at fancy restaurants

00:08:36 ◼   ► or basically all restaurants now.

00:08:38 ◼   ► - They don't even give you a dollar sign, it's just like 12.

00:08:40 ◼   ► - Yeah, exactly.

00:08:41 ◼   ► Experimentation is proven that that makes people

00:08:45 ◼   ► more willing to buy overpriced meals.

00:08:47 ◼   ► - Well, it looks fancier if you give it like a nice

00:08:50 ◼   ► serif font and you just have the description of the item

00:08:53 ◼   ► with as much white space around it as possible,

00:08:55 ◼   ► as many words on there as possible that you can't understand, and just off to the right

00:08:58 ◼   ► somewhere it says like, twelve.

00:08:59 ◼   ► Yeah, it just says "peanut butter and jelly sandwich space space space thirty-four."

00:09:04 ◼   ► And you're like, "Alright, whatever."

00:09:06 ◼   ► Oh goodness, alright.

00:09:08 ◼   ► Well, we'll talk about overpriced things here in a minute, but why don't we start

00:09:13 ◼   ► with one of the marquee features of Overcast 2.

00:09:17 ◼   ► You finally got streaming working.

00:09:19 ◼   ► I did, yeah.

00:09:20 ◼   ► I mean, you know, there's not that much to say about it, I don't think.

00:09:24 ◼   ► I talked here before about all the different times

00:09:27 ◼   ► I tried to get it working in the past.

00:09:29 ◼   ► And I just tried so many different approaches

00:09:32 ◼   ► and I tried the correct approach multiple times

00:09:35 ◼   ► before actually getting my side of it right.

00:09:39 ◼   ► This is using Apple's low level audio file stream API

00:09:43 ◼   ► and it's incredibly low level and incredibly unforgiving

00:09:47 ◼   ► and also pretty sparsely documented.

00:09:51 ◼   ► So the core audio style of their documentation

00:09:56 ◼   ► is to be really conservative with the words.

00:10:00 ◼   ► And so if there's like some minute little detail,

00:10:04 ◼   ► you have to kind of read it like a lawyer

00:10:06 ◼   ► to really understand like, oh, what is this exactly?

00:10:11 ◼   ► This says audio sample, what kind of audio sample is that?

00:10:13 ◼   ► Or what kind of time stamp is this?

00:10:14 ◼   ► What time space is this in?

00:10:16 ◼   ► Things like that, it's really tricky to get right.

00:10:18 ◼   ► But I got it right eventually.

00:10:20 ◼   ► (laughs)

00:10:21 ◼   ► and with one exception,

00:10:23 ◼   ► which I did not uncover in any of my testing

00:10:25 ◼   ► and my beta did not uncover it either,

00:10:28 ◼   ► and that is it doesn't play AAC files

00:10:33 ◼   ► where the header is at the end of the file.

00:10:35 ◼   ► So this is, I'll go over it quickly why it does this

00:10:40 ◼   ► and why this is gonna be tricky to fix,

00:10:41 ◼   ► but basically the API is a streaming API,

00:10:46 ◼   ► so like I give it bytes and I tell it,

00:10:49 ◼   ► okay, I'm at this point in the file,

00:10:51 ◼   ► here are the bytes from that point,

00:10:52 ◼   ► and I just stream it in, and then it tells me,

00:10:55 ◼   ► okay, we got the header, we got the properties,

00:10:57 ◼   ► we got anything you need to know,

00:10:58 ◼   ► and now we're ready to give you samples.

00:11:00 ◼   ► So here's the audio samples.

00:11:02 ◼   ► And then as you come through, as you page through the bytes,

00:11:05 ◼   ► it gives you the samples for it.

00:11:07 ◼   ► The problem is that that requires the header data

00:11:10 ◼   ► about the file, which tells you really important things

00:11:12 ◼   ► like the sample rate of the audio,

00:11:14 ◼   ► how many channels is it, mono or stereo.

00:11:17 ◼   ► You can't really decode the audio properly

00:11:20 ◼   ► until you have those basics.

00:11:22 ◼   ► With MP3, it's fine.

00:11:23 ◼   ► With MP3, it puts everything important

00:11:26 ◼   ► to the beginning of the file

00:11:27 ◼   ► and also the beginning of every header,

00:11:29 ◼   ► of every little chunk of the file.

00:11:30 ◼   ► So with MP3, you can basically decode an MP3

00:11:34 ◼   ► starting from anywhere in the file.

00:11:37 ◼   ► You just kind of skim ahead for a certain byte pattern

00:11:39 ◼   ► and there's a frame and then you can just start.

00:11:40 ◼   ► It's amazing.

00:11:41 ◼   ► MP3 is so simple.

00:11:43 ◼   ► If you wanna have embedded metadata or chapters,

00:11:46 ◼   ► All of that is shoved up front in the ID3v2 block

00:11:49 ◼   ► on an MP3.

00:11:50 ◼   ► So you can just read the first couple kilobytes

00:11:53 ◼   ► or couple hundred kilobytes if there's artwork in there.

00:11:55 ◼   ► MP3 is amazing.

00:11:57 ◼   ► It's a great format and the patents all expire

00:11:59 ◼   ► in a couple of years.

00:12:00 ◼   ► Most of them already have.

00:12:01 ◼   ► So the last of the patents expire in a couple of years.

00:12:06 ◼   ► MP3 is great.

00:12:07 ◼   ► Unfortunately, about, I think last time I checked it,

00:12:11 ◼   ► it was something like three to five percent

00:12:13 ◼   ► of podcasts that Overcast has indexed.

00:12:16 ◼   ► So whatever I know about in the whole directory,

00:12:19 ◼   ► about three to 5% of episodes use AAC format,

00:12:23 ◼   ► or M4A or MOV, it's all the same format.

00:12:25 ◼   ► The QuickTime MOV/M4A format is less ideal

00:12:30 ◼   ► than MP3 on the reading side.

00:12:35 ◼   ► It is this incredibly architecture astronaut design format

00:12:39 ◼   ► where you can contain anything and you can embed anything

00:12:42 ◼   ► and you can have all these arbitrary tracks

00:12:43 ◼   ► It's a much more versatile format than MP3

00:12:47 ◼   ► 'cause it's like this kind of grand container format,

00:12:50 ◼   ► but it's really hard to read in a way,

00:12:54 ◼   ► without having the whole file.

00:12:55 ◼   ► You read it and it gives you this tree structure of atoms,

00:12:58 ◼   ► and within each atom is more atoms and more structures,

00:13:02 ◼   ► and the biggest problem is,

00:13:04 ◼   ► well first of all, the chapters are a disaster.

00:13:06 ◼   ► Chapters are interspersed throughout the entire file,

00:13:09 ◼   ► so you can't easily stream them.

00:13:11 ◼   ► And the other problem is that it is possible

00:13:14 ◼   ► to write all the audio data up front

00:13:18 ◼   ► and then to write the headers at the end.

00:13:21 ◼   ► So without reading the end of the file,

00:13:24 ◼   ► you don't know things like what format the audio is in

00:13:27 ◼   ► so you can't decode it, which means you can't stream that.

00:13:30 ◼   ► So if you've ever seen on like a file export dialog,

00:13:34 ◼   ► if you've ever seen those little check boxes

00:13:35 ◼   ► that say something along the lines of optimize for internet

00:13:38 ◼   ► or make it streamable or fast start, something like that.

00:13:43 ◼   ► What those do is they tell the encoder

00:13:46 ◼   ► to write that header up front.

00:13:49 ◼   ► Don't wait till all the data's written

00:13:50 ◼   ► and then write it at the end.

00:13:51 ◼   ► Write the header up front, or at least go back

00:13:54 ◼   ► and rewrite it up front after you're done

00:13:56 ◼   ► so that it's not sitting at the end of the file

00:13:58 ◼   ► so then people can stream it.

00:14:00 ◼   ► - There's not reliable support for byte range support

00:14:02 ◼   ► for byte range requests

00:14:04 ◼   ► so you can grab the end of the file?

00:14:06 ◼   ► - That's a separate issue.

00:14:07 ◼   ► I can grab the end of the file,

00:14:09 ◼   ► but the audio file stream API

00:14:11 ◼   ► is not compatible with this at all.

00:14:13 ◼   ► So I actually do byte ranges for seeking,

00:14:17 ◼   ► but the way I do it is I read the beginning of the file

00:14:20 ◼   ► every time to get the headers,

00:14:23 ◼   ► and then once I've gotten the headers

00:14:24 ◼   ► and the audio data starts, then I jump ahead

00:14:26 ◼   ► and send a second byte range request

00:14:28 ◼   ► for the part I actually need if it's further ahead.

00:14:30 ◼   ► So the problem is this audio file stream API

00:14:35 ◼   ► literally just does not support this at all.

00:14:37 ◼   ► Like this layout of having the header at the end,

00:14:40 ◼   ► even if I jump ahead, get it,

00:14:41 ◼   ► and then give it back to this API,

00:14:42 ◼   ► even if I have the whole file on disk,

00:14:44 ◼   ► that API does not support that.

00:14:46 ◼   ► So my options to fix this problem,

00:14:49 ◼   ► and I'll get back to byte ranges in a minute,

00:14:51 ◼   ► my options to fix this problem are either

00:14:54 ◼   ► to just, well, not support those files,

00:14:55 ◼   ► which is not a great idea.

00:14:57 ◼   ► I've already heard from a few people

00:14:58 ◼   ► who listen to some shows like that

00:15:00 ◼   ► that are very unhappy and that's understandable.

00:15:03 ◼   ► Or I can run those through a different API entirely.

00:15:07 ◼   ► The old API, I was using the ext audio file API.

00:15:12 ◼   ► That supports them, but that doesn't support streaming.

00:15:15 ◼   ► So I can have these two different code paths

00:15:17 ◼   ► of this pretty major part of the player

00:15:19 ◼   ► where I'm running it for certain files, not others.

00:15:22 ◼   ► And by the way, I don't even know which one

00:15:23 ◼   ► I need to call yet until I have a big chunk of the file.

00:15:27 ◼   ► Or I can do my third option,

00:15:29 ◼   ► which is what I'm probably going to end up doing,

00:15:31 ◼   ► which is actually fix the file on disk and then play it,

00:15:36 ◼   ► which is not a great option.

00:15:40 ◼   ► - Aren't you downloading it then?

00:15:42 ◼   ► - Yeah, I mean, I will probably end up

00:15:44 ◼   ► just not streaming these because it would be a lot harder

00:15:48 ◼   ► if I make it, 'cause then if I stream it,

00:15:49 ◼   ► then I have to actually go one level deeper in the API

00:15:53 ◼   ► and decode the format myself for both of these

00:15:57 ◼   ► and then feed raw data into a converter.

00:16:00 ◼   ► It's very, I would really rather not go into that level

00:16:02 ◼   ► if I don't need to.

00:16:03 ◼   ► - I'm surprised you're not going for the server-side solution

00:16:06 ◼   ► where these problematic files get downloaded once

00:16:09 ◼   ► to your server and then you put in the metadata for them

00:16:11 ◼   ► and then you encounter one.

00:16:13 ◼   ► In the wild, you just ask the server for the chapter info

00:16:15 ◼   ► which you already fetched, you know what I mean?

00:16:17 ◼   ► - Well, this isn't just chapters.

00:16:18 ◼   ► This is actually like playback.

00:16:20 ◼   ► I can't even play the files that are arranged this way.

00:16:22 ◼   ► - So you can't give it some, like, you know,

00:16:24 ◼   ► if you download, what I'm saying is download it once

00:16:25 ◼   ► on your server, get all the information you need

00:16:27 ◼   ► about the file, whether it's chapters or how to play it

00:16:29 ◼   ► or whatever, then when you encounter that file,

00:16:31 ◼   ► rather than reading the information about how to play it,

00:16:34 ◼   ► just ask the server for information about how to play it,

00:16:36 ◼   ► and it will tell you,

00:16:37 ◼   ► and you don't have to bother reading the end.

00:16:39 ◼   ► - Well, first of all, anything involving

00:16:41 ◼   ► like a server cache of the file has other problems.

00:16:43 ◼   ► So, you know, mostly it's, if I have one,

00:16:47 ◼   ► if I do one approach, which is to cache the file myself

00:16:50 ◼   ► and re-serve it from my servers to the app,

00:16:53 ◼   ► podcast publishers don't like that,

00:16:55 ◼   ► and I wouldn't like that.

00:16:56 ◼   ► - Yeah, no, you're not gonna serve the file,

00:16:57 ◼   ► just the metadata.

00:16:58 ◼   ► So then the question is, what happens if the metadata I'm serving is out of date because

00:17:02 ◼   ► the file has changed?

00:17:04 ◼   ► Then that's a problem.

00:17:05 ◼   ► I could e-tag it and everything, but then what if I don't have the data I need?

00:17:10 ◼   ► It's a problem.

00:17:11 ◼   ► I'm going to have to come up with some kind of crazy fix.

00:17:13 ◼   ► It's probably going to end up doing a temporary hack just to get these files playing again

00:17:17 ◼   ► in the next few weeks, and then a longer term I'll probably have to go to that lower level

00:17:21 ◼   ► API.

00:17:22 ◼   ► But I really don't want to do that.

00:17:24 ◼   ► I think the server side is going to end up coming back.

00:17:27 ◼   ► you obviously can do solutions that involve downloading it but if you actually want to stream

00:17:31 ◼   ► it someone's got to download it and get all the info about it right and then it's just a question

00:17:36 ◼   ► of you know like you can always fall once you have the download path in there you can always fall

00:17:40 ◼   ► back to that so you're like well if i couldn't get the information or this is a rare podcast that no

00:17:44 ◼   ► one has looked at or whatever or the information is out of date because the file has changed or

00:17:47 ◼   ► whatever then you fall back to downloading but i can't think of another way that you can successfully

00:17:52 ◼   ► stream something. I mean, I guess you're saying the lower level API, you can just do the jump

00:17:57 ◼   ► ahead to the end, get the header and jump back thing and just have a little delay?

00:18:00 ◼   ► Yeah, and that's, and of course that's if it supports byte ranges, which, and that's

00:18:04 ◼   ► a separate thing, so as I'm developing this, I've learned a lot of things about what servers

00:18:08 ◼   ► support and don't support. I also love the if range header. It's awesome. Whoever designed

00:18:13 ◼   ► that is awesome. The problem with byte range support is that podcast CDNs only support

00:18:20 ◼   ► it fairly loosely. So some of them support it every time, some of them never support

00:18:25 ◼   ► it, and some of them, like Libsyn, which powers most of Overcast's most popular podcast,

00:18:32 ◼   ► Libsyn supports it sometimes. So the way Libsyn works, as far as I can tell, is that they

00:18:39 ◼   ► are the front end to multiple backend CDNs. And when you request a Libsyn URL, you're

00:18:45 ◼   ► redirected to just a random backend CDN, as far as I can tell. And you don't get the

00:18:50 ◼   ► the same one every request.

00:18:51 ◼   ► So you can make a request for a file

00:18:54 ◼   ► that supports byte ranges, then make a second request

00:18:56 ◼   ► and it won't support them.

00:18:58 ◼   ► So you never know when calling a libsyn URL

00:19:01 ◼   ► whether the request you make will support byte ranges

00:19:04 ◼   ► even regardless of whether the previous one

00:19:07 ◼   ► to the exact same host did or not.

00:19:09 ◼   ► If I build the UI and everything assuming

00:19:11 ◼   ► that byte ranges will always be there,

00:19:13 ◼   ► I basically, I can't assume that basically.

00:19:15 ◼   ► So I have to also, like in a case like this,

00:19:18 ◼   ► I have to cover the case where I have to download the entire file first before I can even start

00:19:24 ◼   ► playing it, because that might be the only way to get it from the server.

00:19:27 ◼   ► So how did you not encounter this until after Overcast 2 was released? You just never downloaded

00:19:31 ◼   ► one of these podcasts for testing? Yeah, I mean, I honestly didn't download

00:19:36 ◼   ► a lot of M4A podcasts, or AAC, whatever you want to call it, the format, because all the

00:19:41 ◼   ► podcasts I listened to are MP3 format. And so I went and found some that I was using

00:19:47 ◼   ► for testing for things like AAC chapters and everything,

00:19:51 ◼   ► the enhanced AAC stuff, but none of those

00:19:54 ◼   ► happened to be encoded in this way

00:19:56 ◼   ► with the header at the end.

00:19:57 ◼   ► So I didn't test the entire catalog.

00:20:01 ◼   ► Maybe I should have written some kind of script to do that.

00:20:03 ◼   ► That would have been wise, but I didn't.

00:20:05 ◼   ► So now I have a bunch of test cases,

00:20:08 ◼   ► but basically it's gonna be a lot of work

00:20:11 ◼   ► to support what is a very small percentage of downloads.

00:20:16 ◼   ► But that's the job.

00:20:19 ◼   ► Well, since you've got all this server-side data,

00:20:21 ◼   ► you could actually use-- in the interim,

00:20:23 ◼   ► use the social engineering solution

00:20:24 ◼   ► where you find the most popular podcasts that

00:20:28 ◼   ► are like this, take their files, rewrite them losslessly

00:20:31 ◼   ► if possible, and send them an email and say,

00:20:33 ◼   ► hey, your files don't play on my player

00:20:35 ◼   ► because you didn't optimize for streaming.

00:20:37 ◼   ► Here's versions of your files that do.

00:20:39 ◼   ► Could you swap them in for the old ones?

00:20:42 ◼   ► You're not even asking them to re-encode.

00:20:44 ◼   ► You're not even telling them that going forward,

00:20:46 ◼   ► you're basically doing all the work for them.

00:20:47 ◼   ► I don't know how many it is, but maybe if you do that

00:20:49 ◼   ► with three podcasts, does that cover 90%

00:20:51 ◼   ► of your problematic files?

00:20:53 ◼   ► - Well, if you're gonna go a social route,

00:20:55 ◼   ► the easier social route is that all the people

00:20:57 ◼   ► who are encountering these podcasts are yelling at me

00:21:00 ◼   ► on Twitter and copying the people who publish them.

00:21:03 ◼   ► And some of the people who publish them have already said,

00:21:05 ◼   ► oh, I didn't realize that.

00:21:07 ◼   ► I've now converted the files with this different

00:21:08 ◼   ► checkbox option that doesn't do that.

00:21:10 ◼   ► - Yeah, they'll probably do it going forward

00:21:11 ◼   ► if people complain, but anyway, if you did it for them,

00:21:15 ◼   ► there would be even less work,

00:21:16 ◼   ► 'cause it would be like,

00:21:17 ◼   ► here's your stuff on the platter,

00:21:18 ◼   ► and I don't think, you can't do it for everybody,

00:21:19 ◼   ► but if, I don't know what the stats look like,

00:21:21 ◼   ► but if there's a real big peak of like,

00:21:24 ◼   ► these three podcasts account for 90% of the problems

00:21:26 ◼   ► 'cause they're popular,

00:21:28 ◼   ► that would also give you some breathing room.

00:21:29 ◼   ► - Yeah, I mean, it's really been a very small number

00:21:33 ◼   ► that I've heard about, so I'm not that concerned.

00:21:37 ◼   ► I really do think the right solution is just gonna be

00:21:40 ◼   ► to make the thing in the app

00:21:41 ◼   ► that just downloads the whole file,

00:21:43 ◼   ► and then rewrites it to be correct

00:21:46 ◼   ► and then feeds it into the parser.

00:21:47 ◼   ► 'Cause then I have way fewer code paths,

00:21:50 ◼   ► then I only have the one parsing code path,

00:21:51 ◼   ► I can stay with that high level API

00:21:53 ◼   ► and not do a whole bunch more work.

00:21:55 ◼   ► And then just those files won't stream.

00:21:59 ◼   ► They have to download all the way.

00:22:01 ◼   ► But that's kind of how they work already.

00:22:04 ◼   ► That's how Overcast worked constantly until last week.

00:22:08 ◼   ► And because it's such a very small number of these,

00:22:12 ◼   ► I think that's an acceptable trade-off,

00:22:15 ◼   ► but I don't know, I could change my mind, we'll see.

00:22:16 ◼   ► - So is it as far as 2.0 bugs as far as you know,

00:22:19 ◼   ► any other minor things?

00:22:21 ◼   ► - I broke the entire watch app, that was fun.

00:22:23 ◼   ► - And you didn't realize it?

00:22:24 ◼   ► Oh, good thing nobody uses that, huh?

00:22:26 ◼   ► - I don't even use it.

00:22:27 ◼   ► It's kind of embarrassing, but I don't use my watch app.

00:22:29 ◼   ► I hardly ever touch it.

00:22:31 ◼   ► The main reason that it broke is even more embarrassing.

00:22:35 ◼   ► Relatively late in the process,

00:22:38 ◼   ► I changed my artwork downloader

00:22:40 ◼   ► use a different URL session API because the old URL connection API was deprecated. I forgot

00:22:46 ◼   ► to check the box in Xcode for that one category file that said to include it in the linker

00:22:53 ◼   ► in the watchkit binary. So I had a missing symbol. When that function was called, it

00:22:59 ◼   ► would just throw an exception and crash. And the reason I didn't catch this is because

00:23:04 ◼   ► I wasn't using the simulator to test this. I was using my actual watch. And my actual

00:23:08 ◼   ► watch turns out has been very buggy recently. And one of the ways it's been buggy is that

00:23:15 ◼   ► I thought it was installing the new versions, like the newest build. Whenever I put a test

00:23:20 ◼   ► flight build or a dev build onto my phone of Overcast, I assumed that it was working

00:23:25 ◼   ► the way it always has worked before, which is that it was also copying over the watch

00:23:29 ◼   ► binary at the same time, or, you know, a few seconds later. Turns out it wasn't, and the

00:23:34 ◼   ► And the overcast build on my watch was like a month old and was before this change.

00:23:39 ◼   ► And so I was testing it on my watch and I thought it was fine.

00:23:42 ◼   ► Additionally, none of the beta testers caught it.

00:23:46 ◼   ► Possibly for similar reasons of test flight being buggy, possibly because none of them

00:23:49 ◼   ► used the watch app.

00:23:50 ◼   ► I don't know.

00:23:51 ◼   ► Doesn't matter.

00:23:52 ◼   ► Nobody caught it.

00:23:53 ◼   ► I didn't catch it.

00:23:54 ◼   ► No one else did.

00:23:55 ◼   ► App review didn't catch it either, which means app review didn't even try the watch app because

00:23:58 ◼   ► it literally wouldn't launch.

00:23:59 ◼   ► I have a fix in for that and for a playlist editing bug

00:24:03 ◼   ► where the very first time you would create a playlist,

00:24:06 ◼   ► it wouldn't save any of your settings,

00:24:07 ◼   ► which was really embarrassing.

00:24:09 ◼   ► I don't know why I didn't catch that, but it didn't.

00:24:10 ◼   ► Now it's fixed.

00:24:11 ◼   ► And a couple other little things.

00:24:13 ◼   ► But really that was it.

00:24:14 ◼   ► It was very minor besides the WatchKit app.

00:24:16 ◼   ► - It's kind of surprising the App Reviewer

00:24:17 ◼   ► didn't even launch the Watch app.

00:24:18 ◼   ► What kind of things can you sneak through

00:24:20 ◼   ► now that you know this?

00:24:21 ◼   ► - Yeah, I was also surprised by that.

00:24:23 ◼   ► Probably not worth pushing the boundaries on that,

00:24:25 ◼   ► but that's interesting.

00:24:27 ◼   ► - Well, the number one thing you sneak through

00:24:30 ◼   ► when you realize you have a loophole in App Review

00:24:32 ◼   ► is some sort of tethering back door.

00:24:34 ◼   ► Isn't that the rules?

00:24:35 ◼   ► - Yeah. - I thought that's

00:24:36 ◼   ► how this worked.

00:24:37 ◼   ► - That is standard protocol.

00:24:38 ◼   ► Either a tethering back door or an NES emulator.

00:24:41 ◼   ► - Right, right, one or the other.

00:24:42 ◼   ► To go back to streaming just for a minute,

00:24:45 ◼   ► I had a couple of minor questions

00:24:47 ◼   ► I wanted to ask you about that.

00:24:49 ◼   ► I feel like you've probably already answered this,

00:24:51 ◼   ► but what was the hardest part?

00:24:53 ◼   ► I mean, it sounds like the hardest part is yet to come,

00:24:54 ◼   ► which is figuring out this AAC stuff,

00:24:56 ◼   ► but is there anything else that was really hard

00:24:59 ◼   ► that's worth noting?

00:25:01 ◼   ► And this kind of works its way into the follow-on question,

00:25:05 ◼   ► which is what are you most proud of,

00:25:07 ◼   ► specifically within streaming?

00:25:09 ◼   ► - The hardest part was really just getting

00:25:12 ◼   ► all these pieces to work together,

00:25:14 ◼   ► 'cause there's a whole bunch of weird conditions,

00:25:17 ◼   ► weird concurrency, potential pitfalls,

00:25:20 ◼   ► a lot of different states

00:25:23 ◼   ► that all the different pieces can be in.

00:25:25 ◼   ► when you start downloading the file.

00:25:27 ◼   ► I mentioned I have this initial request

00:25:29 ◼   ► and then I can do a range request.

00:25:31 ◼   ► So which state are all those different requests in?

00:25:33 ◼   ► Once you have that data, as I'm passing it to the decoder,

00:25:38 ◼   ► what happens when I run against the end

00:25:40 ◼   ► of what I've downloaded and I have to send the decoder

00:25:42 ◼   ► a partial block or do I wait till I have a whole block?

00:25:45 ◼   ► And then what data is the decoder in?

00:25:46 ◼   ► Has it found the header yet?

00:25:48 ◼   ► Do I have what's necessary to decode yet?

00:25:51 ◼   ► Has it reached end of file?

00:25:52 ◼   ► What happens when all these other things reach end of file?

00:25:53 ◼   ► Do I actually play to the very last sample or do I cut it off somewhere because I messed

00:26:00 ◼   ► up a buffer somewhere?

00:26:02 ◼   ► All these little things, that's the hard part.

00:26:05 ◼   ► In addition to the aforementioned hard part which was just figuring out how the heck to

00:26:11 ◼   ► use this API, which is this low level C API, which like most of Core Audio is extremely

00:26:18 ◼   ► unforgiving and gives relatively unhelpful error codes.

00:26:23 ◼   ► So it was just really, really tough just getting all that stuff right.

00:26:28 ◼   ► What lessons did you learn while you were doing this?

00:26:31 ◼   ► Use AVPlayer.

00:26:33 ◼   ► Fair enough.

00:26:35 ◼   ► My biggest lesson is don't do this if you're starting from scratch.

00:26:38 ◼   ► However, I am very happy I did it because now I have, you know, the whole reason I had

00:26:42 ◼   ► to do this for Quick Review, anybody who wasn't familiar, the whole reason I had to do all

00:26:45 ◼   ► this craziness and not just use AVPlayer is that AVPlayer, as far as I'm aware, and

00:26:52 ◼   ► I've checked into this numerous times

00:26:54 ◼   ► and tried numerous different approaches,

00:26:55 ◼   ► but AVPlayer, as far as I know, cannot do smart speed.

00:27:00 ◼   ► So in order to do smart speed,

00:27:01 ◼   ► I had to write this whole audio engine

00:27:03 ◼   ► at a lower level than that.

00:27:05 ◼   ► And I didn't wanna just have a version that does smart speed

00:27:10 ◼   ► and a version that can stream

00:27:12 ◼   ► so that you can only have smart speed

00:27:13 ◼   ► while you're streaming.

00:27:14 ◼   ► I didn't wanna do that.

00:27:15 ◼   ► I wanted to have every feature available

00:27:17 ◼   ► whether you were streaming or not.

00:27:18 ◼   ► It was a lot of work, but I think it was worth it.

00:27:22 ◼   ► I assume I can fix this annoying little AAC limitation.

00:27:25 ◼   ► I think it was very much worth it.

00:27:28 ◼   ► Because now, even people who don't,

00:27:31 ◼   ► who never thought they would use streaming,

00:27:33 ◼   ► which I count myself among those,

00:27:35 ◼   ► it is really nice to have.

00:27:36 ◼   ► Even if you set it to automatically download everything

00:27:39 ◼   ► when it comes out, you will still, at some point,

00:27:42 ◼   ► run through a situation where, oh, something just came out,

00:27:45 ◼   ► I wanna listen to it right now.

00:27:47 ◼   ► and you can just tap it and it just starts playing immediately.

00:27:50 ◼   ► You don't have to wait for it to download.

00:27:52 ◼   ► And that's really nice.

00:27:53 ◼   ► And this was holding up a bunch of other possible features,

00:27:57 ◼   ► things like making inbound sharing links better.

00:27:59 ◼   ► Like now, I haven't done this yet,

00:28:01 ◼   ► but I can make it so that if you tap an Overcast share

00:28:04 ◼   ► link in something, I can pop up a little player

00:28:06 ◼   ► in the Overcast app and have you preview that or play

00:28:09 ◼   ► that right there.

00:28:11 ◼   ► Or just simpler things, like when notifications

00:28:13 ◼   ► come in for new episodes, I can have a play button on them

00:28:16 ◼   ► instead of just dismiss and wait for it to download.

00:28:18 ◼   ► You know, the simple stuff like that,

00:28:20 ◼   ► it's just really nice to have all these options.

00:28:22 ◼   ► So now I have the foundation that lets me actually do it.

00:28:26 ◼   ► - What lessons did you learn while doing all this,

00:28:28 ◼   ► other than testing the watch app and AV player?

00:28:30 ◼   ► Anything else?

00:28:32 ◼   ► - You learn to play one of each kind of podcast

00:28:35 ◼   ► that you know exists out in the wild, right?

00:28:36 ◼   ► So one MP3, one AAC with the headers at the end,

00:28:39 ◼   ► one AAC with the headers at the beginning,

00:28:41 ◼   ► one high bitrate, one low bitrate, you know, just--

00:28:44 ◼   ► I mean, basically what I learned there was,

00:28:46 ◼   ► I was already doing that, but what I learned there was,

00:28:49 ◼   ► there's another type that I forgot about

00:28:51 ◼   ► that I should have been included in the list. (laughs)

00:28:54 ◼   ► - It's funny to me that you have yet to mention

00:28:56 ◼   ► that you learned that unit testing may not be so terrible,

00:28:58 ◼   ► but I know that's a tree that it's not worth barking up.

00:29:01 ◼   ► Any other interesting stories about the development?

00:29:06 ◼   ► 'Cause it's been about a year, right?

00:29:07 ◼   ► What version of the streaming engine are we on, four or five?

00:29:11 ◼   ► - Three or four.

00:29:13 ◼   ► - Okay, so it was about a year.

00:29:15 ◼   ► Any other interesting stories worth sharing

00:29:17 ◼   ► before we talk about businessy things?

00:29:19 ◼   ► - Not really.

00:29:20 ◼   ► I mean, it was, you know, I did a few other things.

00:29:23 ◼   ► I converted a lot, a lot of things

00:29:25 ◼   ► that were previously firing notifications

00:29:28 ◼   ► to tell various other parts of the app to update.

00:29:31 ◼   ► And now I'm doing all those, I'm doing a lot of those

00:29:35 ◼   ► as KVO, using Facebook's KVO controller

00:29:38 ◼   ► instead of all this notification hell.

00:29:40 ◼   ► So that was actually a nice thing.

00:29:41 ◼   ► I'm also, I've serialized database access

00:29:45 ◼   ► onto the main thread, which is crazy sounding

00:29:48 ◼   ► to some people, and I know this is very technical

00:29:50 ◼   ► and very boring, so I'll go over it very quickly

00:29:52 ◼   ► for people who don't wanna hear it.

00:29:53 ◼   ► Basically, my previous version was using a version

00:29:56 ◼   ► of my FC model layer that had a background queue

00:29:59 ◼   ► for database access, and there was a number of challenges

00:30:02 ◼   ► with this and potential bugs when things were happening

00:30:05 ◼   ► in a background queue, and then the UI tries to update them,

00:30:08 ◼   ► and then what version of it does it get,

00:30:10 ◼   ► and does it update at the right time?

00:30:14 ◼   ► Does anything accidentally get called

00:30:15 ◼   ► in the background thread?

00:30:17 ◼   ► It's a touching UI kit, which you can't do.

00:30:19 ◼   ► So there's all sorts of these little possible bugs,

00:30:21 ◼   ► some of which became real bugs.

00:30:22 ◼   ► And the new version serializes all database operations

00:30:26 ◼   ► on the main thread.

00:30:28 ◼   ► You're typically told not to do this for a lot of things

00:30:30 ◼   ► because that is performance problematic.

00:30:34 ◼   ► Like if you have a big database operation

00:30:36 ◼   ► that's using the database for a couple of seconds,

00:30:39 ◼   ► the UI can't update.

00:30:41 ◼   ► In practice, if you have an operation that's that long,

00:30:44 ◼   ► the UI gets blocked anyway if you have it

00:30:47 ◼   ► on a background queue because at some point

00:30:49 ◼   ► the UI calls into the database queue

00:30:52 ◼   ► and the operation is ahead of it in the queue

00:30:54 ◼   ► that's blocking it all up.

00:30:55 ◼   ► So I found in practice any large operations

00:30:58 ◼   ► that were large enough to take a noticeable amount of time

00:31:00 ◼   ► on the database would block the UI anyway

00:31:03 ◼   ► even if it was on the background thread

00:31:05 ◼   ► 'cause something in the UI would call

00:31:06 ◼   ► into the database and have to wait.

00:31:09 ◼   ► - You don't have a, you know,

00:31:10 ◼   ► readers aren't blocked by writers kind of isolation,

00:31:13 ◼   ► or a separate queue for read and write, you know?

00:31:15 ◼   ► - You know, I don't have that.

00:31:18 ◼   ► I could add that at some point to FC model.

00:31:22 ◼   ► In practice, I really haven't needed it.

00:31:24 ◼   ► You know, in practice, having everything on the main thread

00:31:26 ◼   ► is both way simpler.

00:31:29 ◼   ► You know, it's, as I mentioned, all the possible bugs

00:31:31 ◼   ► that you get from having it on some other thread,

00:31:33 ◼   ► all those bugs are gone.

00:31:35 ◼   ► And also, it hasn't really been slower.

00:31:37 ◼   ► In fact, many things about it are faster.

00:31:39 ◼   ► So I have found no downside to this

00:31:43 ◼   ► in the kind of, in the ways that I actually use the database

00:31:46 ◼   ► which is, you know, every, I'm never doing like a table scan

00:31:50 ◼   ► or whatever SQLite calls a table scan.

00:31:52 ◼   ► I'm never doing that in the UI.

00:31:54 ◼   ► Everything that I'm querying is always indexed

00:31:57 ◼   ► and the data set really isn't that big relatively speaking.

00:31:59 ◼   ► So yeah, that's fine.

00:32:01 ◼   ► - All right, one final question

00:32:05 ◼   ► before we talk about something else that's awesome.

00:32:07 ◼   ► What's next on the roadmap?

00:32:09 ◼   ► Anything you're willing to share?

00:32:11 ◼   ► One benefit that we, well, air quote benefit

00:32:14 ◼   ► that we get being on the ATP emails

00:32:17 ◼   ► is that we also get about a 10th of your Overcast

00:32:21 ◼   ► support requests that somehow end up in the ATP inbox.

00:32:25 ◼   ► And I've seen a handful of people very perturbed

00:32:28 ◼   ► about the lack of authenticated feeds in Overcast.

00:32:32 ◼   ► Do you plan on doing that or is there anything else

00:32:34 ◼   ► you'd like to talk about with regard to your roadmap?

00:32:37 ◼   ► So right now, I crawl ATP once for all subscribers to it.

00:32:42 ◼   ► And then I collect that data

00:32:45 ◼   ► and then I distribute it to the people.

00:32:46 ◼   ► So okay, suppose you have a password feed.

00:32:49 ◼   ► Do I crawl it once and then re-serve that to other people?

00:32:54 ◼   ► That's obviously not great for security

00:32:56 ◼   ► and kind of defeats the purpose of password feeds

00:32:59 ◼   ► and could be like a piracy avenue.

00:33:02 ◼   ► So I don't wanna do that.

00:33:03 ◼   ► Do I crawl the feed separately for every user

00:33:07 ◼   ► that's subscribed to it with their username and password,

00:33:09 ◼   ► that is like the most semantically correct way to do it.

00:33:12 ◼   ► But that's also incredibly wasteful and resource heavy

00:33:16 ◼   ► on my crawling servers.

00:33:17 ◼   ► Suppose a really popular podcast

00:33:20 ◼   ► went with password protected feeds.

00:33:22 ◼   ► And then I had 60,000 people in my podcast app

00:33:26 ◼   ► trying to download this one thing

00:33:27 ◼   ► with 60,000 different passwords,

00:33:29 ◼   ► then that adds 60,000 crawling feeds I have to do

00:33:32 ◼   ► to every interval that I'm refreshing the feed,

00:33:36 ◼   ► that's both a burden on me and a burden on the other server.

00:33:39 ◼   ► So I don't know of anybody who does server-side crawling

00:33:43 ◼   ► and supports password feeds.

00:33:45 ◼   ► But I could be wrong.

00:33:46 ◼   ► I mean, you could do it, but I think at scale,

00:33:49 ◼   ► it starts to become problematic.

00:33:51 ◼   ► And so that's why I've been hesitant to do it.

00:33:53 ◼   ► It's also just among all the different feature requests

00:33:56 ◼   ► I get, it has been pretty low on the priority list

00:33:59 ◼   ► simply because not that many people request it.

00:34:01 ◼   ► And I think most podcasts these days

00:34:03 ◼   ► that are gonna do the password model

00:34:05 ◼   ► Generally speaking, I think that's fairly outdated

00:34:08 ◼   ► in that many of them are switching

00:34:10 ◼   ► to just having their own apps to play them back in.

00:34:13 ◼   ► And there's various reasons why I don't love that approach,

00:34:15 ◼   ► but that's the reality of the market.

00:34:17 ◼   ► So I don't have immediate plans to do this,

00:34:21 ◼   ► but I could do it in the future, I don't know.

00:34:24 ◼   ► - Fair enough, any other interesting things

00:34:26 ◼   ► planned for Overcast 3?

00:34:29 ◼   ► - I'm not even thinking about 3 yet.

00:34:31 ◼   ► No, I mean, honestly, I don't have any remaining

00:34:33 ◼   ► like massive ideas.

00:34:34 ◼   ► I have a bunch of small ideas I want to do,

00:34:36 ◼   ► little things I want to do,

00:34:37 ◼   ► a bunch of things that would be worthy of like 2.1, 2.2,

00:34:41 ◼   ► like that kind of update.

00:34:42 ◼   ► But I currently have nothing,

00:34:44 ◼   ► no concept of what 3.0 would be

00:34:47 ◼   ► or when it would come out, if ever.

00:34:49 ◼   ► I can take the 2.x line going for a long time.

00:34:53 ◼   ► - Fair enough.

00:34:54 ◼   ► All right, why don't you tell us

00:34:54 ◼   ► about something that's awesome.

00:34:55 ◼   ► - Our second sponsor tonight is Squarespace,

00:34:57 ◼   ► the all-in-one platform that makes it fast and easy

00:34:59 ◼   ► to create your own professional website, portfolio,

00:35:01 ◼   ► or online store.

00:35:03 ◼   ► For a free trial and 10% off,

00:35:04 ◼   ► at squarespace.com and enter offer code ATP at checkout.

00:35:08 ◼   ► Squarespace is simple and powerful,

00:35:10 ◼   ► with easy to use, what you see is what you get tools.

00:35:13 ◼   ► They're all drag and drop, they're customizable,

00:35:15 ◼   ► everything's great, your designs are beautiful.

00:35:17 ◼   ► They are professionally designed templates

00:35:19 ◼   ► that you can customize to your heart's content,

00:35:21 ◼   ► with no coding required.

00:35:23 ◼   ► So regardless of your skill level,

00:35:24 ◼   ► you can customize the look of your website.

00:35:26 ◼   ► And if you want, if you have a lot of skill,

00:35:28 ◼   ► you can jump in there and inject custom CSS,

00:35:31 ◼   ► custom JavaScript, whatever you wanna do

00:35:33 ◼   ► to customize the site template.

00:35:35 ◼   ► These are all of course responsive.

00:35:37 ◼   ► You can also embed commerce functionality.

00:35:39 ◼   ► Every website comes with free online store capabilities

00:35:43 ◼   ► for both physical or digital goods.

00:35:45 ◼   ► All this is supported 24/7 via live chat and email.

00:35:48 ◼   ► If you have any questions or need any help, 24/7 support.

00:35:52 ◼   ► All of this is state of the art technology

00:35:54 ◼   ► which powers your site to ensure stability and security.

00:35:57 ◼   ► This is trusted by millions of people

00:35:59 ◼   ► and some of the most respected brands in the world.

00:36:02 ◼   ► All this starts at just eight bucks a month.

00:36:04 ◼   ► If you sign up for a year, you get a free domain name.

00:36:06 ◼   ► And this is great.

00:36:06 ◼   ► I mean, there are so many cases

00:36:08 ◼   ► where this is the right answer.

00:36:09 ◼   ► For example, earlier today,

00:36:10 ◼   ► I was at one of my favorite coffee shops,

00:36:12 ◼   ► and the owner of the coffee shop said,

00:36:14 ◼   ► "Hey, I need to update my website.

00:36:16 ◼   ► "You know about computers.

00:36:17 ◼   ► "What should I do?

00:36:18 ◼   ► "How should I do it?"

00:36:19 ◼   ► And I said, "Just here, look."

00:36:20 ◼   ► I wrote down Squarespace form.

00:36:21 ◼   ► He hadn't heard of it

00:36:22 ◼   ► because he doesn't listen to podcasts, apparently, ever.

00:36:24 ◼   ► I wrote down Squarespace form.

00:36:26 ◼   ► I said, "Here, do this.

00:36:27 ◼   ► "Trust me.

00:36:28 ◼   ► "This is all you need.

00:36:29 ◼   ► "And hey, I'll let you know how that went."

00:36:31 ◼   ► and I go there about once a month,

00:36:34 ◼   ► so we'll follow up in future sponsors on how that went,

00:36:36 ◼   ► but I'm pretty sure it's gonna work out well.

00:36:39 ◼   ► This is a perfect example of exactly

00:36:40 ◼   ► what you need Squarespace for.

00:36:41 ◼   ► So anyway, check it out, start your free trial site today.

00:36:44 ◼   ► This is what I told him to do.

00:36:46 ◼   ► No credit card required at squarespace.com.

00:36:49 ◼   ► Use code ATP to get 10% off your first purchase

00:36:52 ◼   ► when you sign up.

00:36:53 ◼   ► Squarespace, build it beautiful.

00:36:55 ◼   ► - All right, the business model changed for Overcast 2.

00:36:59 ◼   ► What once was not free is now free.

00:37:02 ◼   ► - Yeah, basically.

00:37:04 ◼   ► So before the model was a free app

00:37:07 ◼   ► with an in-app purchase for five bucks one time

00:37:09 ◼   ► to unlock all the features.

00:37:11 ◼   ► So it was like there were limits,

00:37:12 ◼   ► certain features were not available,

00:37:13 ◼   ► and you could pay once to unlock them all.

00:37:17 ◼   ► And it worked fine.

00:37:18 ◼   ► It wasn't amazing, but it was fine.

00:37:20 ◼   ► The problem was that, of course, as these things go,

00:37:24 ◼   ► it's a one-time purchase, and so revenue,

00:37:28 ◼   ► Average revenue per month was going down,

00:37:30 ◼   ► as these things tend to do.

00:37:32 ◼   ► And I had this major new update,

00:37:34 ◼   ► and I had so many people asking me,

00:37:36 ◼   ► is it gonna be a paid upgrade?

00:37:38 ◼   ► There were people on both sides of that.

00:37:40 ◼   ► You know, like, about half the people who asked

00:37:43 ◼   ► were hoping the answer was yes,

00:37:45 ◼   ► and about half of them were hoping the answer was no,

00:37:46 ◼   ► 'cause they wanted to give me more money

00:37:48 ◼   ► to make sure the app survived.

00:37:51 ◼   ► Not 'cause they like it a lot.

00:37:51 ◼   ► So I had so many people who wanted the free update,

00:37:55 ◼   ► so many people who wanted to give me more money,

00:37:57 ◼   ► And I evaluated all these different options,

00:37:58 ◼   ► and the other problem was that,

00:38:00 ◼   ► you know, by having the app that has limits,

00:38:02 ◼   ► about 20% of the users actually paid to unlock it.

00:38:05 ◼   ► And as far as in-app purchase rates go,

00:38:07 ◼   ► that's incredibly good.

00:38:08 ◼   ► Like, that's a great conversion rate.

00:38:10 ◼   ► Most people with like a free thing, with a paid conversion,

00:38:13 ◼   ► they would love a conversion rate of 20%.

00:38:16 ◼   ► So I have no complaints about the conversion rate there.

00:38:19 ◼   ► However, that still meant that 80% of the users

00:38:22 ◼   ► were getting this limited, terrible version of the app.

00:38:26 ◼   ► and I wasn't using that version.

00:38:28 ◼   ► I wouldn't use that version, even for a day.

00:38:31 ◼   ► It was annoying to even have these different code paths

00:38:34 ◼   ► to test them, and I knew that 80% of my users

00:38:39 ◼   ► were using a terrible version of the app.

00:38:41 ◼   ► So I switched to a voluntary patronage model.

00:38:46 ◼   ► So basically, and this is not actually that new.

00:38:49 ◼   ► It's very, very similar to what I did with Instapaper,

00:38:53 ◼   ► about halfway through its ownership.

00:38:55 ◼   ► So you basically, the entire app is free.

00:38:59 ◼   ► Instapaper was paid, but that was a different story.

00:39:01 ◼   ► The entire app is now free, all features are unlocked,

00:39:05 ◼   ► and you pay if you want to to support ongoing development.

00:39:09 ◼   ► And you pay a dollar a month if you wanna do that.

00:39:12 ◼   ► And so far, that's working.

00:39:14 ◼   ► I haven't matched my previous income yet.

00:39:17 ◼   ► And I don't expect to for a while,

00:39:19 ◼   ► but I've gotten something like 40% there already,

00:39:23 ◼   ► I mean, after less than a week.

00:39:25 ◼   ► And the feature as it is right now is incredibly,

00:39:29 ◼   ► to a fault, unintrusive, like it's buried.

00:39:32 ◼   ► It doesn't, like if you just launch the app

00:39:34 ◼   ► and use it normally, you never even see it.

00:39:37 ◼   ► It's only in the settings screen.

00:39:38 ◼   ► If you go to the settings screen, then you will see it.

00:39:41 ◼   ► Otherwise you don't see it at all.

00:39:43 ◼   ► And over time, like that's mostly 'cause I just haven't

00:39:46 ◼   ► gotten around to making things that like promote it more.

00:39:49 ◼   ► Like you know, I was thinking maybe at the very bottom

00:39:51 ◼   ► of the podcast list screen, putting like a little thing

00:39:54 ◼   ► saying we're supported by Pay.

00:39:55 ◼   ► If you aren't a monthly patron,

00:39:57 ◼   ► putting a little thing there, saying here,

00:39:58 ◼   ► this is what we do.

00:39:59 ◼   ► Or when I add more features,

00:40:01 ◼   ► I can put a little welcome thing up saying,

00:40:04 ◼   ► here's what's new, and here's how you can join it

00:40:06 ◼   ► if you want to support this.

00:40:07 ◼   ► So there's stuff like that I can do,

00:40:08 ◼   ► but I haven't done it yet.

00:40:10 ◼   ► So having done a really invisible update,

00:40:14 ◼   ► and to already be about 40% towards my previous revenue

00:40:16 ◼   ► after less than a week, I consider that a success.

00:40:20 ◼   ► That is a faster uptake rate than I would have expected.

00:40:23 ◼   ► So I'm very happy with that so far.

00:40:26 ◼   ► - Good.

00:40:27 ◼   ► Why free?

00:40:28 ◼   ► I mean, don't you wanna make money, man?

00:40:31 ◼   ► I mean, I heard what you just said about

00:40:33 ◼   ► how you've got a pretty good uptake on this,

00:40:36 ◼   ► but you said you haven't quite reached

00:40:39 ◼   ► your old monthly revenue,

00:40:41 ◼   ► and that seemed to be working okay,

00:40:43 ◼   ► so why mess with the system?

00:40:45 ◼   ► - You know, part of it, as I said,

00:40:47 ◼   ► was the reasons of satisfying,

00:40:49 ◼   ► now everybody gets the good app, right?

00:40:52 ◼   ► so all my customers now are using the best features.

00:40:54 ◼   ► And you know, to me that's important,

00:40:56 ◼   ► like things like Smart Speed and Voice Boost,

00:40:59 ◼   ► this is why my entire custom audio engine exists.

00:41:03 ◼   ► This is why I had to write all that.

00:41:04 ◼   ► This is why I can't just use AV Player.

00:41:06 ◼   ► And it's a big reason that differentiates Overcast

00:41:09 ◼   ► from the built-in podcast app.

00:41:11 ◼   ► So when you're telling people like,

00:41:12 ◼   ► you know, you should use this app, well why?

00:41:15 ◼   ► The big, and you know, I have other competitors too

00:41:17 ◼   ► and that's fine, but the biggest competitor by a long shot

00:41:21 ◼   ► is the built-in Apple podcast app by a mile.

00:41:25 ◼   ► And that comes in, it comes with every phone,

00:41:28 ◼   ► it comes pre-installed.

00:41:30 ◼   ► As far as, I don't even think you can delete it,

00:41:31 ◼   ► I think it's always there.

00:41:33 ◼   ► So the question is like, how do I make anybody use my app

00:41:38 ◼   ► instead of the built-in Apple podcast app?

00:41:40 ◼   ► And a lot of the features that I locked behind that paywall

00:41:43 ◼   ► before, Apple gives those away for free.

00:41:46 ◼   ► So that really, it was not very competitive with Apple's app

00:41:50 ◼   ► And so before I had the scheme where the app was free

00:41:53 ◼   ► for some of the app and then you pay to get everything.

00:41:57 ◼   ► So it was always free for most people and pay for some.

00:42:02 ◼   ► And now it is still free for most people and pay for some.

00:42:07 ◼   ► But I've just changed what they're paying for and why.

00:42:12 ◼   ► And the result of that is a much simpler app

00:42:17 ◼   ► that's much better for everybody.

00:42:20 ◼   ► So that's why I think it's a win,

00:42:21 ◼   ► because it's still, it was always free for most people,

00:42:26 ◼   ► and I make money somewhere in there,

00:42:28 ◼   ► and now it still is that.

00:42:29 ◼   ► It is still free for most people,

00:42:31 ◼   ► and I make money somewhere in there.

00:42:32 ◼   ► Again, I've just changed the specifics of how that's done,

00:42:37 ◼   ► but it's still making money, it's still profitable,

00:42:40 ◼   ► and over the long term, it's probably gonna be

00:42:42 ◼   ► just as profitable, if not more so,

00:42:44 ◼   ► because, as I mentioned,

00:42:46 ◼   ► This is not my first paid app.

00:42:48 ◼   ► I've seen this train before.

00:42:50 ◼   ► That's not a saying.

00:42:51 ◼   ► Now it is.

00:42:52 ◼   ► (laughing)

00:42:53 ◼   ► I've seen this train before.

00:42:55 ◼   ► I know how this goes.

00:42:58 ◼   ► After my first year where I did this big thing

00:43:01 ◼   ► and I keep giving free updates,

00:43:03 ◼   ► average monthly revenue goes down

00:43:05 ◼   ► because I start to reach saturation

00:43:06 ◼   ► among my existing audience

00:43:08 ◼   ► of like the people who bought it last year

00:43:10 ◼   ► when I launched it, over time,

00:43:13 ◼   ► I'm not making any more money from those people

00:43:15 ◼   ► and I'm still giving all these new features

00:43:16 ◼   ► for free if I do a paid upgrade that has other problems, people generally hate those. Over

00:43:22 ◼   ► time if I would have stuck with the old model, that was a downward slope of the revenue.

00:43:27 ◼   ► It was slow, I was still doing okay, but the trend line was clearly slowly going down.

00:43:34 ◼   ► And that happens to every paid app that I've ever seen. So with this, first of all with

00:43:39 ◼   ► this I think the trend line will slowly go up. So that's a huge improvement right there.

00:43:45 ◼   ► And it also kind of levels it out a lot more.

00:43:47 ◼   ► You know, it's more predictable income,

00:43:49 ◼   ► and it gives people a way to give me more money

00:43:54 ◼   ► if they want to.

00:43:55 ◼   ► And so it really does solve a lot of problems.

00:43:58 ◼   ► And I don't think it's any,

00:44:01 ◼   ► I don't think it's that crazy.

00:44:02 ◼   ► You know, when you look at it as,

00:44:04 ◼   ► how I described it a minute ago,

00:44:05 ◼   ► it's just like, it was free for most people,

00:44:08 ◼   ► and some people pay before, and now it still is.

00:44:10 ◼   ► Just those things, like just what you pay for is different.

00:44:13 ◼   ► If you look at it that way, I don't think it's that crazy.

00:44:16 ◼   ► - Well yeah, but that's all fine for Marco, right?

00:44:18 ◼   ► - Oh God, this.

00:44:19 ◼   ► Yeah, you know, this was all Twitter

00:44:24 ◼   ► and blog drama this morning.

00:44:25 ◼   ► I don't wanna get really into it on the show,

00:44:26 ◼   ► but the short version is that anytime I do anything,

00:44:31 ◼   ► I hear from people saying, "Well, that's fine for you,"

00:44:34 ◼   ► copying the old "That's fine for Merlin" joke,

00:44:36 ◼   ► like, "Well, that's fine for Marco, but I can't do that

00:44:40 ◼   ► "because I don't have X," whether it's, you know,

00:44:42 ◼   ► of any Twitter followers as I do,

00:44:44 ◼   ► or the brand recognition that I do,

00:44:46 ◼   ► or whatever the case may be, the PR that I get

00:44:49 ◼   ► when I do things, always these arguments of,

00:44:54 ◼   ► well, everything I do is unique and invalid

00:44:59 ◼   ► to apply to anybody else.

00:45:02 ◼   ► And that's just not true.

00:45:04 ◼   ► And the fact is people are gonna convince themselves

00:45:07 ◼   ► of that no matter what I say, so it doesn't really matter,

00:45:08 ◼   ► it's not really worth arguing,

00:45:10 ◼   ► but the short version is the path I took.

00:45:13 ◼   ► The result that I'm at now of where my career is now,

00:45:18 ◼   ► what my audience, what people expect of me,

00:45:21 ◼   ► what people want to see from me,

00:45:23 ◼   ► yeah, that's hard to replicate in five minutes,

00:45:25 ◼   ► but I've been building it for like 10 years.

00:45:28 ◼   ► And the ways I have built it over 10 years

00:45:31 ◼   ► are generally accessible to other people.

00:45:34 ◼   ► If you also blog for 10 years and podcast for five years

00:45:39 ◼   ► and make a bunch of apps along the way,

00:45:42 ◼   ► then you have a better chance of getting

00:45:46 ◼   ► the kind of launch attention that I can get, et cetera.

00:45:50 ◼   ► But also, that launch attention is fleeting.

00:45:54 ◼   ► It's one time, that's it.

00:45:56 ◼   ► Launch attention alone does not carry things.

00:45:58 ◼   ► If it did, everything I did would succeed.

00:46:01 ◼   ► But in fact, most of the things I do don't succeed.

00:46:05 ◼   ► That's why I stopped doing them.

00:46:07 ◼   ► You know, like the magazine did not succeed.

00:46:10 ◼   ► The launch was, I think a couple data for the launch

00:46:14 ◼   ► was the most subscribers it ever had.

00:46:17 ◼   ► And then it just went down from there.

00:46:18 ◼   ► The magazine did not succeed.

00:46:19 ◼   ► Bugshot did not succeed.

00:46:21 ◼   ► Nursing Clock, of course, was kind of a joke

00:46:22 ◼   ► but that didn't succeed.

00:46:24 ◼   ► Even, I was having trouble keeping Instapaper afloat

00:46:26 ◼   ► when I sold it.

00:46:27 ◼   ► That's one of the reasons I sold it.

00:46:28 ◼   ► Because it was just not doing that well anymore.

00:46:30 ◼   ► You know, the fact is the things I do

00:46:33 ◼   ► don't succeed by default.

00:46:35 ◼   ► I am afforded the luxury of a stronger launch

00:46:38 ◼   ► than many people can get, but the value of that launch,

00:46:42 ◼   ► as I said, is temporary.

00:46:44 ◼   ► Look at our friends who made Vesper, right?

00:46:49 ◼   ► John Gruber, Brent Simmons, Dave Whiskas.

00:46:51 ◼   ► These people had massive audiences,

00:46:54 ◼   ► especially John Gruber, these massive audiences,

00:46:56 ◼   ► and yet Vesper hasn't taken over the world.

00:46:59 ◼   ► Obviously, the size of your audience is really,

00:47:02 ◼   ► it helps, but it's not all you need,

00:47:05 ◼   ► and it's not a guarantee.

00:47:06 ◼   ► So the people who really need to hear this won't hear it,

00:47:09 ◼   ► so it doesn't really matter what I say,

00:47:11 ◼   ► but I don't know, if there's anything to take away

00:47:13 ◼   ► from this, it is that almost nothing that I've done

00:47:17 ◼   ► is unique, and there is a road to take here

00:47:21 ◼   ► to increase your chances of success,

00:47:23 ◼   ► but it might take you 10 years and a lot of work,

00:47:26 ◼   ► 'cause that's what it took me.

00:47:28 ◼   ► - Yeah, one of the things you've mentioned in the past

00:47:30 ◼   ► was wanting to get your app, I think you mentioned it

00:47:34 ◼   ► your blog post, your pragmatic pricing blog post that we'll put in the show notes.

00:47:38 ◼   ► Wanting to get your app into the hands of as many people as possible, like you're

00:47:42 ◼   ► going for market share over profit, like rather than selling a $99

00:47:46 ◼   ► artisanal handcrafted podcast app to 17 people, you want

00:47:50 ◼   ► your app to be in the hands of the most people possible as a hedge against

00:47:54 ◼   ► what did you say, big money? Was that what you called it? Coming in?

00:47:58 ◼   ► Like, you didn't want podcasts to become like proprietary

00:48:02 ◼   ► your Facebook eyes, they're owned and controlled by a single company, so you

00:48:05 ◼   ► wanted to get your application out there to the broadest audience possible. And

00:48:11 ◼   ► that is in line with your original pricing model, which was free with an app

00:48:15 ◼   ► purchase because you reduced the barrier to someone tapping the little button in

00:48:19 ◼   ► the store and getting it on their device. And now again, same thing, you know, if you

00:48:24 ◼   ► want it on your device you can tap a button, you don't have to give any money,

00:48:27 ◼   ► only now it's a better app when you download it, so there's more chance that

00:48:30 ◼   ► people will tap the button. So that all works but one of the complaints from people about your new

00:48:38 ◼   ► pricing model has been revolving around what effect your new pricing model may have on the

00:48:45 ◼   ► other podcast clients that are for sale now. I guess not including apples because like they

00:48:51 ◼   ► don't care what you do and maybe not including the big guys that can afford to give away their

00:48:57 ◼   ► I don't know if there's any other big guys like that,

00:48:59 ◼   ► but basically other smallish independent podcast clients.

00:49:04 ◼   ► Do you think your new pricing model will have any effect

00:49:08 ◼   ► on the fortunes of those clients?

00:49:12 ◼   ► - I don't think it has any more effect than it did before,

00:49:15 ◼   ► which is I don't think it has any more effect

00:49:18 ◼   ► than what the Apple client has.

00:49:20 ◼   ► All of us are at a severe disadvantage

00:49:21 ◼   ► because the Apple client is built in and pre-installed

00:49:25 ◼   ► and free and has most features

00:49:28 ◼   ► that most people need built in.

00:49:30 ◼   ► - And it's reel to reel so the audio quality is better.

00:49:33 ◼   ► Has that warm sound you just can't get from vinyl, Casey.

00:49:36 ◼   ► Hi guys.

00:49:36 ◼   ► (laughing)

00:49:38 ◼   ► - And also, when you do a search in the app store

00:49:43 ◼   ► for podcasts or podcasts, it shows this giant banner up top

00:49:47 ◼   ► for the built-in podcast app.

00:49:50 ◼   ► To promote that, to basically divert your attention

00:49:53 ◼   ► back to that to dissuade you

00:49:54 ◼   ► from getting another podcast app

00:49:56 ◼   ► if you actually search the App Store.

00:49:58 ◼   ► - And you can't even delete that app off your phone,

00:50:00 ◼   ► so it's not as if you need to download it from the store.

00:50:03 ◼   ► Like, you have it.

00:50:04 ◼   ► - Yeah, so there's a reason why, I mean,

00:50:06 ◼   ► by most estimates, I think the Apple Podcast App

00:50:09 ◼   ► has something like 60 to 90% market share,

00:50:12 ◼   ► depending on who you ask.

00:50:13 ◼   ► I mean, it's a massive, massive player,

00:50:15 ◼   ► and that market share number, as far as I know,

00:50:17 ◼   ► is not going down.

00:50:19 ◼   ► So that's a huge disadvantage for anybody

00:50:23 ◼   ► in the market.

00:50:24 ◼   ► So mine being mostly free in the sense that

00:50:29 ◼   ► I still am asking people for money,

00:50:30 ◼   ► but just now you can get all the features for free.

00:50:32 ◼   ► It's a valid question that like,

00:50:34 ◼   ► am I killing these smaller apps?

00:50:36 ◼   ► I don't think I am.

00:50:37 ◼   ► I don't think I'm, I think I am,

00:50:41 ◼   ► first of all I'm competing,

00:50:42 ◼   ► and this is something that they can do if they want to.

00:50:45 ◼   ► By most estimates that I've been able to piece together

00:50:50 ◼   ► from like rank data,

00:50:51 ◼   ► I think they're all making more money than me doing what they're doing. So they've been

00:50:55 ◼   ► out grossing me, I think, for the whole, for the year, or at least coming very close, or

00:51:00 ◼   ► being very close. So I think they're doing fine. You know, when you have this giant built-in

00:51:06 ◼   ► app that is, you know, built into the phone, free, very full-featured, look at like the

00:51:12 ◼   ► iOS Notes app is another good example of this. Or any of the built, you know, the iOS Weather

00:51:16 ◼   ► app, any of the built-in iOS apps, calculator, I mean all this, reminders, all this stuff.

00:51:21 ◼   ► There are markets for all of those apps in the App Store,

00:51:24 ◼   ► for third-party versions of those,

00:51:26 ◼   ► and they're often very healthy markets

00:51:28 ◼   ► with many different competitors.

00:51:30 ◼   ► And even if Apple's app takes the 80% or whatever,

00:51:34 ◼   ► that still leaves a lot of people.

00:51:37 ◼   ► And I had this problem with Instapaper with ReadingList,

00:51:39 ◼   ► and as far as I could tell, ReadingList never really

00:51:42 ◼   ► had much of an effect on Instapaper either way.

00:51:44 ◼   ► It didn't seem like a negative or a positive.

00:51:46 ◼   ► Because the thing is, when you're making an app,

00:51:50 ◼   ► The, especially something as complex as a podcast app,

00:51:54 ◼   ► the sum of all your little decisions along the way,

00:51:57 ◼   ► which is tons of little tiny design decisions,

00:51:59 ◼   ► the sum of all those decisions is what makes the app

00:52:03 ◼   ► fit people or not fit people,

00:52:04 ◼   ► whether it kind of matches with the way you think

00:52:07 ◼   ► about things or whether it conflicts

00:52:09 ◼   ► with the way you think about things.

00:52:10 ◼   ► Whether it's a design you like or you don't like.

00:52:12 ◼   ► Before you even get to features or price,

00:52:14 ◼   ► those are all considerations.

00:52:16 ◼   ► And because people have different preferences

00:52:18 ◼   ► what they want, how they want it to look,

00:52:20 ◼   ► how they want it to work,

00:52:21 ◼   ► what features they need and don't need,

00:52:23 ◼   ► that creates tons of market potential for other apps

00:52:27 ◼   ► in every category, even categories that Apple

00:52:29 ◼   ► already has a built-in thing up front for.

00:52:31 ◼   ► The fact that I come into this category

00:52:34 ◼   ► and I give my app away for free,

00:52:36 ◼   ► I don't think, I think it's the same calculus,

00:52:38 ◼   ► I think it's the same situation

00:52:40 ◼   ► that the apps have always been in before.

00:52:42 ◼   ► Like, apps that provide something that people want

00:52:45 ◼   ► that do things a little differently than the built-in one

00:52:48 ◼   ► now than mine, those will find audiences the same way mine did. They will find audiences

00:52:54 ◼   ► regardless of how much my app costs.

00:52:56 ◼   ► All right, so next question. If it turns out that changing your pricing in this way did

00:53:04 ◼   ► reduce the sales of the competing applications to a degree that maybe one out of two of them

00:53:09 ◼   ► drop off or whatever, how does that fit in with your goals of trying to make sure big

00:53:16 ◼   ► money doesn't come into podcasting. Is it not affected at all? Is it negative? Is it

00:53:20 ◼   ► positive? And how would you feel about it?

00:53:22 ◼   ► Well, I would feel pretty bad if I actually like killed someone else's app, but I have

00:53:28 ◼   ► seen very little evidence to suggest that that kind of thing has ever happened in iOS

00:53:32 ◼   ► or any software market for that matter. That's not really how things tend to go. Usually

00:53:39 ◼   ► apps die or become unsuccessful or non-economical

00:53:44 ◼   ► to continue because they themselves just kind of

00:53:47 ◼   ► didn't do that well maintaining their own app

00:53:50 ◼   ► or keeping their own users around.

00:53:51 ◼   ► I had the same, when I had Instapaper,

00:53:55 ◼   ► I learned this pretty well too that I was always

00:53:57 ◼   ► worried about my competition, I was always worried.

00:53:59 ◼   ► I talked about this at XOXO, I was always worried

00:54:02 ◼   ► about what if somebody comes in tomorrow

00:54:05 ◼   ► and takes all my users away?

00:54:07 ◼   ► And that never happened.

00:54:08 ◼   ► Lots of new competitors came around,

00:54:10 ◼   ► some of them very big, some of them completely free.

00:54:13 ◼   ► In fact, most of them completely free.

00:54:15 ◼   ► Some of them are from big companies, some of them Apple.

00:54:18 ◼   ► And it never seemed to make any difference whatsoever

00:54:21 ◼   ► because people had chosen me for lots of reasons

00:54:26 ◼   ► and my app was kind of mine to screw up,

00:54:29 ◼   ► or it was mine to neglect or whatever the case may be.

00:54:33 ◼   ► So if a podcast app goes away, if it shuts down,

00:54:38 ◼   ► which Instacast did this past, I don't know,

00:54:41 ◼   ► six months ago maybe, Instacast did,

00:54:43 ◼   ► but I talked to the creator of that in the past,

00:54:45 ◼   ► and I think he was having trouble for a while keeping it up,

00:54:49 ◼   ► so I don't think I had anything to do with that really.

00:54:51 ◼   ► If an app goes away, yes, I would feel bad

00:54:54 ◼   ► if it was my fault, but I would have a really hard time

00:54:58 ◼   ► believing that it was really my fault.

00:55:01 ◼   ► Furthermore, if being free up front for this past year,

00:55:07 ◼   ► If that was really a big deal, my market share would be bigger.

00:55:12 ◼   ► But it isn't.

00:55:13 ◼   ► Like, Pocket Casts is the greatest counter example of this.

00:55:16 ◼   ► Pocket Casts has way more users than I do.

00:55:19 ◼   ► Way more.

00:55:21 ◼   ► They make way more money than I do.

00:55:23 ◼   ► By a large amount.

00:55:25 ◼   ► They are on both platforms.

00:55:26 ◼   ► They have a staff to maintain it.

00:55:28 ◼   ► Like, Pocket Casts is, by all objective measures, kicking my butt.

00:55:33 ◼   ► And they're paid up front.

00:55:34 ◼   ► And they've been paid upfront the entire time

00:55:36 ◼   ► that I've been free.

00:55:37 ◼   ► And the reason why people choose Pocket Casts

00:55:42 ◼   ► is not because that, oh, this is four or five dollars

00:55:46 ◼   ► or whatever they charge, I don't even know what they charge,

00:55:47 ◼   ► whatever it is, it's not because of that,

00:55:50 ◼   ► it's because they just like it better

00:55:52 ◼   ► or it does things that mine doesn't do

00:55:54 ◼   ► or it serves platforms that I don't serve.

00:55:56 ◼   ► It's for other reasons.

00:55:58 ◼   ► So I think people are putting way more emphasis

00:56:03 ◼   ► on this pricing model, then I think it's warranted.

00:56:08 ◼   ► - So for the big picture thing though,

00:56:10 ◼   ► ignoring whether you are the cause of it or not,

00:56:15 ◼   ► is it better for keeping podcasts

00:56:19 ◼   ► from being Facebook eyes or whatever?

00:56:22 ◼   ► Do you want to see lots of third party clients

00:56:27 ◼   ► out there for podcasts or do you not really care

00:56:29 ◼   ► as long as the overall market share

00:56:32 ◼   ► between the big money and the little guys is the same,

00:56:37 ◼   ► even if the little guys share is divvied up

00:56:40 ◼   ► between five people, seven people, 12 people, two people.

00:56:44 ◼   ► You don't really care.

00:56:45 ◼   ► - My goal here is diversity in the ecosystem.

00:56:49 ◼   ► So from that point of view,

00:56:50 ◼   ► a large number of smaller clients is better.

00:56:54 ◼   ► However, we've had a large number

00:56:57 ◼   ► of small clients for years,

00:56:59 ◼   ► and we haven't made meaningful inroads

00:57:01 ◼   ► into getting significant market share overall

00:57:05 ◼   ► for the independent category.

00:57:06 ◼   ► The winners have always been Apple's podcast app,

00:57:09 ◼   ► and then Down, and Stitcher,

00:57:11 ◼   ► and then things like iHeart and TuneIn

00:57:14 ◼   ► that are kind of not quite podcast players.

00:57:17 ◼   ► Now you have things like Spotify,

00:57:19 ◼   ► they're getting into podcasts,

00:57:20 ◼   ► and that's only going to continue.

00:57:22 ◼   ► Diversity is important,

00:57:24 ◼   ► but you also need some big players

00:57:27 ◼   ► that can be big enough to attract people away

00:57:30 ◼   ► from those other ones.

00:57:31 ◼   ► And that's what I'm trying to be.

00:57:33 ◼   ► Stitcher, I think by most measures,

00:57:35 ◼   ► had something like 5% of the entire podcast player market.

00:57:38 ◼   ► That's a lot.

00:57:39 ◼   ► I'm trying to reach that kind of level.

00:57:40 ◼   ► I know I'm not gonna have like 50% or more.

00:57:43 ◼   ► That's crazy talk.

00:57:45 ◼   ► I would love to reach 5%.

00:57:46 ◼   ► And I'm nowhere near it, but I'd love to get there.

00:57:49 ◼   ► - Isn't Apple kind of a good guy in this scenario?

00:57:51 ◼   ► Because even though they're big

00:57:52 ◼   ► and have all this money and everything,

00:57:54 ◼   ► their sort of vague disinterest in podcasts

00:57:56 ◼   ► means that their player just reads RSS feeds, right?

00:57:59 ◼   ► They're not trying to open up peg walls.

00:58:01 ◼   ► They're not trying to grab copyright or insert

00:58:04 ◼   ► their own ads into people's podcasts.

00:58:05 ◼   ► They seem pretty sort of benign, kind of doddering.

00:58:10 ◼   ► It's the built-in app.

00:58:11 ◼   ► It works OK.

00:58:12 ◼   ► It works better now than it used to.

00:58:14 ◼   ► But it's a straight up podcast app, right?

00:58:16 ◼   ► There's not even any weird iTunes DRM shenanigans

00:58:20 ◼   ► in there, right?

00:58:21 ◼   ► Oh, correct.

00:58:22 ◼   ► But I don't like the podcast.

00:58:24 ◼   ► I don't like theirs.

00:58:26 ◼   ► All right, right.

00:58:26 ◼   ► So the app isn't good.

00:58:27 ◼   ► But I'm saying, in terms of the openness versus closeness,

00:58:29 ◼   ► Like you can totally see how like Stitcher in your scenario,

00:58:33 ◼   ► like something like Stitcher is more the enemy

00:58:35 ◼   ► in terms of they want, you know,

00:58:39 ◼   ► I don't know if they want control of it.

00:58:40 ◼   ► Anyway, their view of the podcast world

00:58:42 ◼   ► is like through the lens of Stitcher, right?

00:58:45 ◼   ► It is a different kind of deal than,

00:58:47 ◼   ► well, people just put up RSS feeds

00:58:48 ◼   ► and this is just a client app that crawls them

00:58:50 ◼   ► and it lets people listen to things.

00:58:52 ◼   ► Like that's what you're trying to preserve essentially,

00:58:54 ◼   ► what we have now, which is,

00:58:55 ◼   ► hey, so you wanna put up a podcast?

00:58:57 ◼   ► It's just an RSS feed with a bunch of attachments.

00:58:59 ◼   ► Anybody can do it.

00:59:00 ◼   ► If you wanna get listed in the big popular directories

00:59:03 ◼   ► like iTunes or I don't know what else is out there,

00:59:05 ◼   ► there's no barrier to that entry.

00:59:07 ◼   ► Apple is not like the gatekeeper.

00:59:09 ◼   ► They don't like charge you money

00:59:10 ◼   ► or require that Apple ads be put in front of your stuff

00:59:13 ◼   ► or whatever.

00:59:14 ◼   ► It's all pretty open and straightforward.

00:59:16 ◼   ► Kind of like blogging used to be in kind of,

00:59:18 ◼   ► well, we'll talk about medium in a little bit,

00:59:20 ◼   ► but anyway, like blogging was in the old days

00:59:23 ◼   ► where it's very open.

00:59:24 ◼   ► And it seemed to me that what you're trying to guard against

00:59:27 ◼   ► is a new world where it's like,

00:59:29 ◼   ► well, if you wanna have a podcast,

00:59:31 ◼   ► you have to go through Stitcher

00:59:32 ◼   ► and Stitcher gets X percentage of your profit

00:59:34 ◼   ► and kind of like the App Store.

00:59:35 ◼   ► And they're the gatekeeper for everything involved.

00:59:39 ◼   ► And they reserve the right to insert their own ads

00:59:42 ◼   ► into your things and to resell your content.

00:59:44 ◼   ► I don't know, like I'm making up,

00:59:45 ◼   ► I have no idea what Stitcher's deal is.

00:59:46 ◼   ► But the idea that, or like Facebook,

00:59:49 ◼   ► like where also you want your articles

00:59:51 ◼   ► to be shown on Facebook,

00:59:52 ◼   ► Facebook controls what gets into Facebook,

00:59:54 ◼   ► Facebook can copy your stuff and republish it,

00:59:57 ◼   ► and Facebook Instant Articles are a thing

01:00:00 ◼   ► that you have to write to,

01:00:01 ◼   ► it's not like they just pull your stuff, anyway.

01:00:02 ◼   ► Or like Apple News, like that's, am I correct

01:00:05 ◼   ► in trying to get a handle on what it is

01:00:06 ◼   ► that you're, the doomsday scenario that doesn't yet exist,

01:00:09 ◼   ► but that you're trying to avoid?

01:00:10 ◼   ► - Yeah, so, you know, to go back to Stitcher for a second,

01:00:14 ◼   ► the thing I don't like about Stitcher is that

01:00:17 ◼   ► they have their own proprietary directory,

01:00:20 ◼   ► and so it's not a general purpose podcast player,

01:00:23 ◼   ► can only play their podcasts. And if you agree to be one of their podcasts, they call your

01:00:30 ◼   ► feed, they get updates, they download your episodes and re-host them themselves so you

01:00:35 ◼   ► don't see the download numbers. They insert their own ads between them, which is weird

01:00:42 ◼   ► and conflicting possibly with the ads that you might have. They transcode your audio

01:00:47 ◼   ► quality to be terrible. And last time I checked, they actually required you to promote Stitcher

01:00:51 ◼   ► on Beyond Your Shows, which would be why you always hear podcasters saying, "Find us

01:00:55 ◼   ► on iTunes and Stitcher," because they have to.

01:00:58 ◼   ► And we decided with this show, early on, when we got a couple emails saying, "Why aren't

01:01:03 ◼   ► you on Stitcher? I can't listen to you," we decided early on that based on the apparent

01:01:08 ◼   ► volume being fairly low, that we didn't think it was worth being on Stitcher, because

01:01:15 ◼   ► we weren't very happy with those terms. And so we decided, "No, it's not worth it.

01:01:20 ◼   ► we don't want to do that. And the only reason we had the option to say no is because Ditchers

01:01:27 ◼   ► Market Share was only like 5% or whatever. If they got any bigger than that, it would

01:01:33 ◼   ► be really hard to say no. So imagine what if they had 15%, 20%. That becomes real numbers.

01:01:41 ◼   ► And so imagine like big publishers like Gimlet or Slate or Radiotopia, like big publishers.

01:01:49 ◼   ► If some player is in the market like that and they start dictating terms like that,

01:01:54 ◼   ► they basically have to agree to them.

01:01:57 ◼   ► They don't have the luxury to say no to something that's controlling possibly 15-20% of the

01:02:02 ◼   ► market.

01:02:03 ◼   ► That's a huge, huge problem.

01:02:05 ◼   ► A player doesn't have to get to a majority stake, 50%, they don't have to get that big

01:02:11 ◼   ► to be able to dictate terms.

01:02:15 ◼   ► I don't want to reach that point.

01:02:16 ◼   ► We've come dangerously close a few times.

01:02:18 ◼   ► I really don't want to reach that point in this medium.

01:02:21 ◼   ► It's even worse, you know, Facebook is way worse,

01:02:23 ◼   ► obviously, because they have,

01:02:25 ◼   ► I think most publishers would tell you that

01:02:27 ◼   ► more than 50% of traffic comes from Facebook.

01:02:31 ◼   ► Like, it's crazy how much traffic Facebook drives.

01:02:34 ◼   ► So, you know, we at least in this medium,

01:02:36 ◼   ► we have the freedom that we don't have

01:02:38 ◼   ► those middlemen who can dictate so many terms to us yet.

01:02:41 ◼   ► However, Apple is one.

01:02:44 ◼   ► And, granted, a podcast app, you know,

01:02:45 ◼   ► Most podcast apps you can subscribe to any URL,

01:02:48 ◼   ► no matter where, you know, any URL that's an RSS feed,

01:02:50 ◼   ► you can subscribe to it and the app will play it.

01:02:53 ◼   ► But Apple still, Apple runs the iTunes podcast directory,

01:02:57 ◼   ► and that directory is the center of all knowledge

01:03:00 ◼   ► of podcasts for a vast majority of podcast players,

01:03:05 ◼   ► Apple's and otherwise, and Apple has rules.

01:03:07 ◼   ► Like I think they disallow like adult stuff

01:03:09 ◼   ► and stuff like that, but anyway, right now,

01:03:12 ◼   ► Apple is pretty hands-off with their directory.

01:03:14 ◼   ► Right now, they have this giant market share

01:03:17 ◼   ► in both the directory side and in the player app side.

01:03:21 ◼   ► But they've mostly been hands off, as you said.

01:03:24 ◼   ► They've mostly kind of ignored it.

01:03:26 ◼   ► But what happens if they don't?

01:03:28 ◼   ► You know, like what happens if they start

01:03:30 ◼   ► using that power they have and making changes?

01:03:33 ◼   ► They probably won't because podcasting,

01:03:35 ◼   ► you know, the reason why they haven't really

01:03:37 ◼   ► touched it much so far, as far as I can tell,

01:03:40 ◼   ► is because it just was never that important to them

01:03:43 ◼   ► relative to everything else they do.

01:03:44 ◼   ► There's this giant company with these giant products,

01:03:46 ◼   ► these giant initiatives.

01:03:47 ◼   ► Podcasting was always so small

01:03:49 ◼   ► that it wasn't really worth them

01:03:50 ◼   ► messing around with, really.

01:03:52 ◼   ► But podcasting is growing.

01:03:53 ◼   ► And Apple is getting, I don't know,

01:03:56 ◼   ► possibly a little bit desperate

01:03:57 ◼   ► in relevance on the music side.

01:03:59 ◼   ► So that might change.

01:04:02 ◼   ► Not only do I wanna make sure that nobody else comes in

01:04:05 ◼   ► and gets enough market shares to be able to dictate terms

01:04:07 ◼   ► to every podcast publisher,

01:04:09 ◼   ► but I also would like to eat away at Apple Share

01:04:11 ◼   ► bit because I'm not comfortable with anybody having that much power over a market even

01:04:17 ◼   ► when it is Apple and they've been pretty good about it so far.

01:04:20 ◼   ► The other side of this is suppose you want to do online video. Online video is just YouTube

01:04:26 ◼   ► these days. Like you might as well just say YouTube because that's what online video means

01:04:30 ◼   ► to most people, YouTube. It is so dominated by one company and also that company is constantly

01:04:37 ◼   ► messing with the terms and constantly changing the way it works. They are really not a great

01:04:42 ◼   ► owner of that entire medium because they have shown over and over again that they're willing

01:04:47 ◼   ► to change things around for their own benefit and to be opaque and to make changes that

01:04:51 ◼   ► might not be in your own best interest as a publisher and things like that. But if you

01:04:54 ◼   ► try to publish video really anywhere except YouTube, it's very hard to get any viewership.

01:05:01 ◼   ► So I don't want podcasts to ever reach that point.

01:05:05 ◼   ► Right now, they're vulnerable to that

01:05:08 ◼   ► with Apple's market share.

01:05:09 ◼   ► But only, you know, Apple is not the kind of company

01:05:12 ◼   ► that would do that generally.

01:05:13 ◼   ► But you know, things change, people change,

01:05:16 ◼   ► companies change, anything could happen.

01:05:17 ◼   ► So ideally, I would like to diversify the market so much

01:05:21 ◼   ► that not only does nobody get the power to dictate terms,

01:05:25 ◼   ► but that Apple doesn't have that power either.

01:05:28 ◼   ► - You don't have to post your video to Facebook,

01:05:30 ◼   ► or to YouTube to get your viewership,

01:05:31 ◼   ► just posted to Facebook. Yeah, there's that issue as well. Maybe it's a duopoly.

01:05:38 ◼   ► Choose your poison. Exactly. The two giants fight each other over who has

01:05:43 ◼   ► monopolies. Yeah, I think we're all thus far protected by a podcast being such a

01:05:49 ◼   ► drop in the bucket, but yeah, I don't know, like, I mean, every time I see these

01:05:53 ◼   ► stories about, you know, new podcast initiatives, the whole serial thing, and

01:05:58 ◼   ► Even the you know what the that front of yours whether the gimlet media or something like that

01:06:03 ◼   ► Anyway, anytime those stories go in you know when podcasts get rediscovered by the mass media briefly for yeah

01:06:11 ◼   ► I'm the two or three year cycle that it's on

01:06:13 ◼   ► people get excited about it being a thing, but then it's kind of like then it quiets down and

01:06:20 ◼   ► I'm not sure if it ever crosses the threshold into a

01:06:26 ◼   ► Matt a real mass media. Yeah, I'm not sure if even crosses the threshold into like reading like as in books

01:06:31 ◼   ► Or you know paper books or ebooks, which I still think it's just a massively larger business than podcasts will ever be

01:06:37 ◼   ► So it could be that this ecosystem is never interesting enough for Apple to wake up and try to rest control of it

01:06:46 ◼   ► But if they did, yeah

01:06:48 ◼   ► I don't know what the hedge is against it

01:06:49 ◼   ► The head is the head you and your little MySQL database with a bunch of podcasts in it

01:06:54 ◼   ► Is that it? Is that all we've got? Is it just, you know, I don't or stitcher that's not you know, oh god

01:06:59 ◼   ► No, I mean like ideally the hedge is lots of other people who has who has a directory because even you have your own directory

01:07:06 ◼   ► Right. That's what yeah, I have my own

01:07:08 ◼   ► But I I will still search iTunes as a fallback to you know to get stuff that I if I can't find anything, right?

01:07:13 ◼   ► Well, so who else has their own directory at all? Microsoft has one

01:07:17 ◼   ► I don't know how big it is, but they do have one

01:07:20 ◼   ► There's a kind of like one or two others around I know Google

01:07:24 ◼   ► I've heard many very, very strong rumblings that Google is working on a major podcast

01:07:30 ◼   ► initiative. I don't know anything about it, but I know they're working on a major

01:07:34 ◼   ► podcast initiative.

01:07:35 ◼   ► I bet it involves ads.

01:07:37 ◼   ► Almost certainly.

01:07:38 ◼   ► I don't know anything about it, but hmm.

01:07:40 ◼   ► Right. And, you know, and again, like this, how much power do you want Google to have

01:07:45 ◼   ► over this medium? The more diverse we can get it to be, the less they can dictate terms.

01:07:51 ◼   ► So this is not going to stay still.

01:07:55 ◼   ► Podcasts are becoming big.

01:07:56 ◼   ► They're getting lots of attention.

01:07:58 ◼   ► We have to be very defensive and skeptical about how this is going to go in the future.

01:08:04 ◼   ► We have to like, I think it's really worth fighting for this.

01:08:09 ◼   ► Because we lost video long ago, right at the start, we lost text mostly now these days.

01:08:16 ◼   ► I don't want to lose podcasting to these big, private,

01:08:20 ◼   ► centralized, proprietary things.

01:08:22 ◼   ► - You know, it's probably protecting podcasts right now

01:08:24 ◼   ► as the incompetence of car makers,

01:08:25 ◼   ► because I feel like that's still kind of the linchpin.

01:08:28 ◼   ► Like when we were kids, radio dominated

01:08:30 ◼   ► because cars had radios in them.

01:08:31 ◼   ► Like I think in the home, the advent of television

01:08:35 ◼   ► replaced radio for a lot of things,

01:08:36 ◼   ► although people still had radios at work

01:08:38 ◼   ► or whatever to listen to.

01:08:39 ◼   ► But for, you know, terrestrial radio

01:08:43 ◼   ► has been all screwed up or whatever,

01:08:45 ◼   ► and the replacement of, you know,

01:08:46 ◼   ► podcasts are the replacement, they're independent,

01:08:48 ◼   ► you can get what you want, they're free,

01:08:51 ◼   ► they're all internet powered,

01:08:52 ◼   ► and once our cars can all play podcasts

01:08:55 ◼   ► using their ubiquitous internet connections

01:08:57 ◼   ► and their apps and their Apple CarPlay,

01:08:59 ◼   ► that's been such a mess, like it's clearly not there yet.

01:09:01 ◼   ► If you could snap your fingers and say,

01:09:03 ◼   ► starting now, every car you buy anywhere in the world

01:09:06 ◼   ► can play any podcast.

01:09:07 ◼   ► And it's like, wow, podcasts have really made it now,

01:09:11 ◼   ► because then people who have never heard of podcasts,

01:09:13 ◼   ► They just grow up in a world where you go into a car and you somehow search for, you

01:09:17 ◼   ► know, This American Life, and it plays it whenever you want, and you don't know how

01:09:21 ◼   ► it happens.

01:09:22 ◼   ► In the same way that you grow up in a car and you press the little preset button and

01:09:26 ◼   ► the radio station comes on and you listen to music, right?

01:09:29 ◼   ► That is the final form, if you will, of podcasts, and that it is the true replacement for radio

01:09:34 ◼   ► only.

01:09:35 ◼   ► It's on demand and it's diversified and whatever.

01:09:37 ◼   ► But you can get that if Facebook owns podcasts or Google owns podcasts.

01:09:42 ◼   ► In fact, it may come faster if one company owns podcasts or whatever.

01:09:47 ◼   ► I'm trying to envision a world in which A, carmakers get their acts together to actually,

01:09:52 ◼   ► you know, sort of, I don't know, it's not like agreeing on a standard, but like, if

01:09:56 ◼   ► you get everyone in a room and say, "We all agree, right?

01:09:58 ◼   ► Podcasts are just an RSS feed and that's where they come from and no one is going to support

01:10:02 ◼   ► any particular company."

01:10:04 ◼   ► But if one of those companies got big, if it was Google or Apple or Microsoft or even

01:10:08 ◼   ► and a Stitcher gets big or something and somehow--

01:10:11 ◼   ► kind of like XM Radio and Sirius got their claws

01:10:14 ◼   ► into the auto industry for a while there,

01:10:16 ◼   ► where it used to be you could get radio in your thing

01:10:19 ◼   ► and then you get satellite radio.

01:10:20 ◼   ► And it was one of two companies, and then they merged.

01:10:22 ◼   ► They merged, right, those two?

01:10:24 ◼   ► One bought the other one out.

01:10:26 ◼   ► Luckily, satellite radio is terrible.

01:10:27 ◼   ► So don't have to worry too much about that.

01:10:29 ◼   ► Technology-wise, there are limitations there

01:10:31 ◼   ► that aren't going to go big.

01:10:34 ◼   ► But the final iteration of cars is,

01:10:37 ◼   ► hey, if cars had internet connection,

01:10:39 ◼   ► then cars could listen to music and they could listen to,

01:10:42 ◼   ► music is already proprietaryized, whatever the word is.

01:10:46 ◼   ► It's Spotify, it's Apple Music, it's RDO,

01:10:49 ◼   ► it's all these other things.

01:10:51 ◼   ► So there's no hope of that being like,

01:10:52 ◼   ► oh, if we just add support for this protocol,

01:10:55 ◼   ► anyone can publish, that's already proprietary.

01:10:57 ◼   ► Podcasts have a chance, have a chance, a slim chance

01:11:00 ◼   ► of remaining in this sort of neutral open state

01:11:04 ◼   ► long enough for cars to get their app back together,

01:11:07 ◼   ► such that all cars have some crappy podcast player in them.

01:11:11 ◼   ► And once the ball starts rolling on that, if it gets going,

01:11:14 ◼   ► it could end up being like the web,

01:11:17 ◼   ► I guess is the best example of like the web got out the door

01:11:19 ◼   ► before anyone could really get control of it.

01:11:20 ◼   ► Microsoft tried and basically failed,

01:11:22 ◼   ► but you can make something with a web browser now

01:11:25 ◼   ► and it can browse the web.

01:11:27 ◼   ► The web browser engines are open source.

01:11:29 ◼   ► You can make one of them.

01:11:30 ◼   ► It can load a webpage that's not owned and controlled

01:11:32 ◼   ► by a single company.

01:11:33 ◼   ► Podcasts have a chance at that, but right now,

01:11:36 ◼   ► I think the expectation is that if a person buys a new car,

01:11:38 ◼   ► that car cannot play a podcast,

01:11:39 ◼   ► except perhaps through a Bluetooth integration

01:11:42 ◼   ► with their iPhone using an application.

01:11:43 ◼   ► I think that is still too complicated for most people.

01:11:46 ◼   ► Most people just wanna go into a car

01:11:48 ◼   ► and have a preset button that's their favorite podcast

01:11:51 ◼   ► and press play and it start playing the next episode

01:11:53 ◼   ► or something like that.

01:11:55 ◼   ► - The example of satellite radio, I think,

01:11:57 ◼   ► was the best counter example to this,

01:11:59 ◼   ► which is that satellite radio, you're right,

01:12:02 ◼   ► It came in, it got great integration into cars

01:12:06 ◼   ► where now almost every car that you buy,

01:12:10 ◼   ► in the US at least, has an option for satellite radio.

01:12:13 ◼   ► Many of them, it's even bundled into other packages

01:12:16 ◼   ► that you might get anyway.

01:12:18 ◼   ► So it's very common.

01:12:20 ◼   ► - But you had to pay for it, which is a real killer.

01:12:22 ◼   ► Like obviously podcast integration wouldn't be like,

01:12:24 ◼   ► oh you have to pay X dollars a month.

01:12:27 ◼   ► - Right, but either way, the hardware was there

01:12:29 ◼   ► and it still hasn't caught on.

01:12:31 ◼   ► And I think what will protect podcasts in the car from that kind of like big integration

01:12:38 ◼   ► deal kind of world is the same thing that has made satellite radio even less relevant

01:12:43 ◼   ► today than it was before, which is, I always say, don't bet against the smartphone.

01:12:50 ◼   ► The smartphone is what is killing satellite radio finally.

01:12:53 ◼   ► I mean satellite radio has been like kind of half dead for a long time, I mean forever,

01:12:57 ◼   ► but it's, the smartphone will kill it for good.

01:13:01 ◼   ► The fact is, internet connected cars,

01:13:04 ◼   ► I think this is probably like a half step.

01:13:08 ◼   ► - Well, through your phone, obviously.

01:13:10 ◼   ► I'm not saying you're gonna pay,

01:13:11 ◼   ► but that would be a payment thing too.

01:13:13 ◼   ► You're gonna pay for your cell access

01:13:14 ◼   ► and then you're gonna have your phone with you

01:13:15 ◼   ► when you're in your car, but I feel like it's not there.

01:13:18 ◼   ► As someone who owns probably the lowest end possible car

01:13:20 ◼   ► you can get that does connect to your phone

01:13:22 ◼   ► and play audio, it works,

01:13:24 ◼   ► but it is not the type of thing that I would say,

01:13:28 ◼   ► just get this car and it will just,

01:13:29 ◼   ► you won't have to do anything

01:13:31 ◼   ► and it'll just figure it out and it's really slow.

01:13:33 ◼   ► Sometimes the Bluetooth doesn't connect.

01:13:35 ◼   ► This ties into Gruber's new theory

01:13:37 ◼   ► that Bluetooth is the worst thing ever

01:13:39 ◼   ► and preventing the future of the internet

01:13:41 ◼   ► of things or whatever.

01:13:42 ◼   ► But honestly, sometimes it just doesn't connect

01:13:44 ◼   ► to Bluetooth audio.

01:13:45 ◼   ► Sometimes it takes a long time.

01:13:46 ◼   ► You can't tell if it's working.

01:13:47 ◼   ► Sometimes you gotta toggle Bluetooth on and off.

01:13:49 ◼   ► When it does work, there's enough of a delay

01:13:51 ◼   ► that you're not quite sure whether it's working,

01:13:52 ◼   ► so you have to decide whether you sit there

01:13:53 ◼   ► and wait for the audio to switch over,

01:13:55 ◼   ► whether you start driving and hope that it will.

01:13:57 ◼   ► Like, it's not as seamless.

01:13:59 ◼   ► It's still, I feel like it's still a nerd experience,

01:14:01 ◼   ► but you're right that that's the way it has to go.

01:14:03 ◼   ► No one's going to pay a separate monthly fee for their,

01:14:06 ◼   ► although boy, don't tell Verizon,

01:14:08 ◼   ► but they would love that they could charge it.

01:14:09 ◼   ► Anyway, no one really wants to pay a separate monthly fee

01:14:12 ◼   ► for their car to have internet access.

01:14:14 ◼   ► And if you're gonna have your phone with you anyway,

01:14:17 ◼   ► we really just need, you know, good car,

01:14:20 ◼   ► I mean, getting back, I guess like CarPlay and all,

01:14:22 ◼   ► you know, good smartphone car integration.

01:14:25 ◼   ► Once that becomes, I think we're all,

01:14:28 ◼   ► Do you think we're there with audio for iPod car integration?

01:14:32 ◼   ► We kind of had that nice period where it was like,

01:14:34 ◼   ► car said 30-pin connectors in them or a USB type thing,

01:14:37 ◼   ► and you would have iPod integration.

01:14:38 ◼   ► I felt like that worked pretty reliably,

01:14:40 ◼   ► but that was obviously a pre-wireless technology.

01:14:44 ◼   ► I mean, I think we are really pretty much there now

01:14:48 ◼   ► for Bluetooth audio, which is even better.

01:14:51 ◼   ► Bluetooth is so much better.

01:14:52 ◼   ► And BMWs.

01:14:55 ◼   ► Even other cars, like if I get rental cars, I try it.

01:14:57 ◼   ► I know other people who try it.

01:14:58 ◼   ► Usually Bluetooth audio is not perfect,

01:15:03 ◼   ► but most of the time pretty good.

01:15:05 ◼   ► - Yeah, the audio is fine, it's just the connection.

01:15:07 ◼   ► 'Cause I have Bluetooth audio,

01:15:08 ◼   ► that's what I'm talking about with the integration.

01:15:11 ◼   ► You need more sophisticated integration for podcasts

01:15:13 ◼   ► if you wanna actually put up an onscreen display

01:15:15 ◼   ► that gives you more than just metadata

01:15:17 ◼   ► as if it's a music track, like you'd like to be able to,

01:15:20 ◼   ► I don't know.

01:15:20 ◼   ► - But that's all you need.

01:15:21 ◼   ► I mean, when you're playing it in a car,

01:15:23 ◼   ► that's all I need.

01:15:25 ◼   ► One of the reasons why I haven't explored options

01:15:29 ◼   ► like making a BMW app for it,

01:15:32 ◼   ► one of the reasons why is because

01:15:34 ◼   ► the Bluetooth integration is just good enough.

01:15:36 ◼   ► And it's really, really convenient

01:15:40 ◼   ► that you just get in the car

01:15:42 ◼   ► and the phone can stay in your pocket,

01:15:44 ◼   ► you just get in the car and a few seconds later

01:15:47 ◼   ► it starts playing your podcast

01:15:48 ◼   ► right where you left off on your phone.

01:15:50 ◼   ► - Much more than a few seconds in crappy cars

01:15:52 ◼   ► and sometimes never because it inexplicably doesn't connect.

01:15:54 ◼   ► Like I'm saying is that I think that's not that's not there yet for regular people

01:15:58 ◼   ► Like it's not it's not the type of thing where you can just assure somebody do you have do you have a smartphone period?

01:16:04 ◼   ► When you buy a new car any car

01:16:06 ◼   ► You will be able to listen to podcasts in the car and there's nothing you'll need to do and no manual you need to read

01:16:12 ◼   ► And no futzing you'll need to do and no caveats about remember don't start driving until the audio plays because it may not be connected to

01:16:17 ◼   ► Bluetooth and you better take care of that before you start moving

01:16:19 ◼   ► Otherwise, you'll be trying to use your touchscreen while you're moving. You're gonna run over a kid

01:16:24 ◼   ► You know John maybe it's just time for you to buy a nicer car

01:16:26 ◼   ► I'm saying like most people are buying cars like this and most people buying cars don't have any Bluetooth integration at this point

01:16:33 ◼   ► So it's I think it's starting to get really true. I don't know a lot of very very cheap cars

01:16:38 ◼   ► like my brother-in-law just got a

01:16:40 ◼   ► Brand new Civic and admittedly there are cheaper cars in the Civic

01:16:44 ◼   ► but I think the Civic is kind of a decent barometer for what a

01:16:48 ◼   ► reasonably priced cars these days and his has what I would call comfort access it has, you know, the proximity key I

01:16:54 ◼   ► Believe it as Bluetooth. It has a humongous touchscreen on it, which I don't think is navigation. Those are those are options

01:17:01 ◼   ► Those are all options

01:17:02 ◼   ► Sure, but I mean, I don't know a lot of people that buy a truly stripped car

01:17:08 ◼   ► I mean you didn't buy a truly stripped car, right? You didn't get the best

01:17:11 ◼   ► Package my first car didn't have a passenger side mirror. I'm talking about today

01:17:17 ◼   ► roll up windows.

01:17:19 ◼   ► - I'm surprised it was legal to not have a window,

01:17:21 ◼   ► like a mirror, like that's crazy.

01:17:23 ◼   ► - He's very old, Marc.

01:17:25 ◼   ► - The car is really, the car was really small

01:17:27 ◼   ► and honestly like, once you get used to it not being there,

01:17:31 ◼   ► I don't know, it wasn't, what I didn't like about it

01:17:34 ◼   ► obviously was the asymmetry, like it's upsetting

01:17:36 ◼   ► that this doesn't have the right, anyway, it was fine.

01:17:40 ◼   ► It also had the cool like a little joystick

01:17:43 ◼   ► to control the mirrors, like instead of power mirrors

01:17:45 ◼   ► the little-- everything was manual.

01:17:47 ◼   ► Nice.

01:17:48 ◼   ► Anyway, this is all getting off track of the podcasting, though.

01:17:52 ◼   ► But anyway, I feel like that is the--

01:17:54 ◼   ► with this podcast stuff, we're like, oh, these stories

01:17:56 ◼   ► in the paper and podcasts are big,

01:17:58 ◼   ► and there's lots of money involved,

01:17:59 ◼   ► and there's VCs, and there's cereal, and blah, blah, blah.

01:18:02 ◼   ► I feel like we're not over the hump yet with podcasts.

01:18:06 ◼   ► And to give an example, we got over the hump with digital

01:18:08 ◼   ► music.

01:18:09 ◼   ► iPods are everywhere.

01:18:11 ◼   ► iPods swept away.

01:18:13 ◼   ► Digital music swept away, digital music on plastic disks was swept away by digital music

01:18:18 ◼   ► on little tiny hard drives and eventually flash chips and stuff.

01:18:22 ◼   ► So that revolution happened.

01:18:24 ◼   ► The podcast supplanting talk radio for most people, I feel like has not happened.

01:18:31 ◼   ► Talk radio, as terrible as it is, I think is still the dominant form of people listening

01:18:37 ◼   ► to other people talking.

01:18:38 ◼   ► Yeah, but I do think podcasting is replacing it.

01:18:42 ◼   ► You know, it's not happening rapidly, but it is happening.

01:18:45 ◼   ► I mean, if you look at most like podcast,

01:18:48 ◼   ► you know, growth graphs or whatever

01:18:50 ◼   ► over the last few years,

01:18:51 ◼   ► it doesn't appear to be accelerating rapidly.

01:18:55 ◼   ► It's just going up slowly and steadily

01:18:57 ◼   ► the way it always has.

01:18:58 ◼   ► And I think that's going to continue.

01:19:01 ◼   ► You know, it is slowly, steadily getting more popular.

01:19:05 ◼   ► It is not really ever going down.

01:19:08 ◼   ► And so over time, it will replace talk radio

01:19:10 ◼   ► for most people.

01:19:11 ◼   ► It'll take a while.

01:19:12 ◼   ► I mean, a lot of people still read newspapers, right?

01:19:15 ◼   ► I mean, but that's not a growth industry, you know?

01:19:19 ◼   ► So I think we have, I think we are really in the early days

01:19:24 ◼   ► of a transition that is definitely happening.

01:19:26 ◼   ► - All right, well, we should move on

01:19:28 ◼   ► to your intentional destruction

01:19:30 ◼   ► of the open blogging platform on the internet.

01:19:33 ◼   ► But first, don't you have one more sponsor?

01:19:35 ◼   ► - Our final sponsor this week is lindsay.com,

01:19:37 ◼   ► an easy and affordable way to help you learn

01:19:39 ◼   ► with high quality, easy to follow video tutorials.

01:19:43 ◼   ► You can instantly stream thousands of courses

01:19:45 ◼   ► created by experts on software, web development,

01:19:48 ◼   ► graphic design, and more.

01:19:49 ◼   ► Go to lynda.com, that's L-Y-N-D-A dot com slash ATP

01:19:54 ◼   ► to see for yourself.

01:19:55 ◼   ► They offer courses for all experience levels,

01:19:57 ◼   ► whether you're beginner or advanced.

01:19:59 ◼   ► And every lynda.com course is produced

01:20:01 ◼   ► at the highest quality.

01:20:03 ◼   ► This is not like the inconsistent homemade videos

01:20:05 ◼   ► on YouTube or whatever.

01:20:06 ◼   ► Courses are broken up into bite sized pieces

01:20:08 ◼   ► so you can learn at your own pace

01:20:09 ◼   ► and learn from start to finish,

01:20:11 ◼   ► or you can just jump in and find a quick answer

01:20:12 ◼   ► with tools such as searchable transcripts.

01:20:15 ◼   ► You can search for what they're saying in the video,

01:20:17 ◼   ► you can see right there, it highlights as it moves along.

01:20:20 ◼   ► You can jump to any point you want

01:20:21 ◼   ► just by clicking the text, it is really quite amazing.

01:20:24 ◼   ► They also have playlists, certificates of course completion

01:20:26 ◼   ► you can publish to your LinkedIn profile.

01:20:29 ◼   ► You can even learn while you're on the go

01:20:31 ◼   ► with Lynda.com apps for iPhone, iPad, and Android.

01:20:34 ◼   ► The way this works is you don't pay per video.

01:20:37 ◼   ► You don't pay per course, you don't pay per subject.

01:20:40 ◼   ► You can watch whatever you want in their entire catalog,

01:20:44 ◼   ► no pressure, no commitment, whatever you want,

01:20:47 ◼   ► for one low flat monthly price of just $25 a month.

01:20:51 ◼   ► 25 bucks a month for unlimited access

01:20:54 ◼   ► to over 100,000 video tutorials

01:20:56 ◼   ► and they're adding more all the time.

01:20:58 ◼   ► They have all sorts of courses that you might love.

01:21:01 ◼   ► From app and web development

01:21:02 ◼   ► and many different programming languages

01:21:04 ◼   ► to productivity apps, creative pro apps

01:21:07 ◼   ► like Adobe's Creative Suite, Logic, and Final Cut,

01:21:10 ◼   ► and even professional skills

01:21:11 ◼   ► like management and negotiation.

01:21:13 ◼   ► So lynda.com is so useful that 30% of colleges

01:21:16 ◼   ► and universities, including most of the Ivy League schools,

01:21:19 ◼   ► offer subscriptions to their students and faculty members.

01:21:21 ◼   ► So this is really quite impressive.

01:21:23 ◼   ► Check it out, I've seen it, they are great.

01:21:25 ◼   ► lynda.com is offering a 10-day free trial

01:21:28 ◼   ► with all course access.

01:21:29 ◼   ► So watch whatever you want for 10 days

01:21:32 ◼   ► in their entire catalog.

01:21:34 ◼   ► All that for free, free trial.

01:21:36 ◼   ► If you visit lynda.com/atp, once again,

01:21:39 ◼   ► that's a 10-day free trial with access to all courses

01:21:42 ◼   ► at lynda.com/atp.

01:21:45 ◼   ► Thanks a lot.

01:21:47 ◼   ► - So you don't like the open blogging platform

01:21:50 ◼   ► you decided to post your, which one do you,

01:21:53 ◼   ► one of the ones about the new version of Overcast on Medium.

01:21:56 ◼   ► - Yes.

01:21:57 ◼   ► - A proprietary platform,

01:21:58 ◼   ► only controlled by a single company.

01:22:00 ◼   ► - I cross-posted.

01:22:01 ◼   ► - Does that make it better?

01:22:03 ◼   ► - Little bit.

01:22:03 ◼   ► - Eh.

01:22:04 ◼   ► Why did you decide to do that?

01:22:07 ◼   ► - Yeah, so my latest post on my site,

01:22:10 ◼   ► I posted on my site,

01:22:11 ◼   ► and also I posted a copy of it on Medium.

01:22:14 ◼   ► Medium is really big now.

01:22:16 ◼   ► I wanted to understand it.

01:22:18 ◼   ► And there's only so much understanding you can do

01:22:21 ◼   ► of a blogging platform, really.

01:22:24 ◼   ► It's a glorified social blogging platform.

01:22:26 ◼   ► There's only so much of understanding you can really have

01:22:29 ◼   ► without actually blogging on it at least once.

01:22:32 ◼   ► So this company is so big, I think it's unwise

01:22:36 ◼   ► to be unfamiliar with it, especially as,

01:22:40 ◼   ► you know, they aren't the biggest company on the web,

01:22:42 ◼   ► but they are incredibly influential and incredibly popular

01:22:46 ◼   ► among people like in our circles, among like the tech people,

01:22:50 ◼   ► the early adopters, the kind of tech media space.

01:22:54 ◼   ► It is disproportionately popular among that crowd.

01:22:57 ◼   ► So I wanted to understand it.

01:22:59 ◼   ► - Why is that, do you think?

01:23:00 ◼   ► - Why is it popular with that crowd?

01:23:02 ◼   ► - So having used it for one post,

01:23:04 ◼   ► which is not a lot of experience,

01:23:06 ◼   ► but it's a lot more than you have, right?

01:23:08 ◼   ► So having used it for one post,

01:23:10 ◼   ► I can say that the editor is fine.

01:23:14 ◼   ► Everyone says it's amazing, it's fine.

01:23:16 ◼   ► It doesn't support markdown, which is unfortunate,

01:23:18 ◼   ► but it's fine.

01:23:19 ◼   ► It was nice having social feedback right there,

01:23:24 ◼   ► like people doing the highlights,

01:23:26 ◼   ► people doing the little recommend or like.

01:23:29 ◼   ► I'm still not clear whether it's recommend or like,

01:23:31 ◼   ► if those are two different things, I don't even know.

01:23:33 ◼   ► But getting all the feedback right there

01:23:36 ◼   ► to see all the different people you know

01:23:39 ◼   ► who would heart recommend like it or whatever,

01:23:43 ◼   ► and then you see the little highlights and everything.

01:23:46 ◼   ► Yes, a few of the comments were kind of interesting.

01:23:49 ◼   ► So it was nice to have that level of incident feedback.

01:23:53 ◼   ► It was like the way Twitter provided incident feedback

01:23:57 ◼   ► when you tweet things.

01:23:58 ◼   ► It was like that, but different and more of it,

01:24:00 ◼   ► and directly on the blog post.

01:24:02 ◼   ► - Or like Tumblr at the little bottom,

01:24:04 ◼   ► where it's like who, re whatever.

01:24:07 ◼   ► Tumblr has a similar thing, right?

01:24:08 ◼   ► It doesn't have the inline ones in the right margin,

01:24:10 ◼   ► but you could post something on Tumblr

01:24:12 ◼   ► and immediately see a bunch of little people's

01:24:13 ◼   ► avatar icons appear on the bottom.

01:24:15 ◼   ► - Exactly, exactly.

01:24:16 ◼   ► So very similar to Tumblr, but focused on actual blog.

01:24:20 ◼   ► Like on Tumblr, you can use Tumblr

01:24:23 ◼   ► for a 12 paragraph blog post,

01:24:25 ◼   ► but very few people will read it.

01:24:28 ◼   ► 'cause that's not really the mode you're in.

01:24:29 ◼   ► When you're browsing Tumblr, you're in like a skimming mode

01:24:32 ◼   ► 'cause all the rest of the content is skimmable stuff.

01:24:35 ◼   ► So nobody wants to stop and read some giant long post.

01:24:38 ◼   ► Whereas Medium, that's the whole point of the service,

01:24:41 ◼   ► is to read people's text.

01:24:43 ◼   ► That is the whole point.

01:24:45 ◼   ► And so it's normal to go there and see something

01:24:48 ◼   ► that is 12 paragraphs long.

01:24:50 ◼   ► So I decided, yeah, I wanted to try it.

01:24:53 ◼   ► I wanted to see what it was like.

01:24:55 ◼   ► And the feedback and the community aspects of it

01:24:59 ◼   ► do seem pretty nice.

01:25:00 ◼   ► In the past, I've been critical of Medium

01:25:03 ◼   ► because I've said, which I still agree with,

01:25:06 ◼   ► that you are not writing for yourself,

01:25:09 ◼   ► you're writing for Medium.

01:25:11 ◼   ► You know, in the same way, like,

01:25:12 ◼   ► when you give Twitter a whole bunch of your content there,

01:25:15 ◼   ► you know, you're really doing Twitter a big favor.

01:25:17 ◼   ► Yourself, you know, you're kind of,

01:25:19 ◼   ► it's of mixed value, right?

01:25:20 ◼   ► So the reason why you might wanna publish on Medium

01:25:24 ◼   ► Now, I said this before, is that if your goal,

01:25:29 ◼   ► if your primary goal is not to become a writer,

01:25:34 ◼   ► necessarily, or to develop your own audience,

01:25:38 ◼   ► but to spread a message, to spread some idea,

01:25:41 ◼   ► to spread some, like, a post.

01:25:45 ◼   ► If your main job isn't writing,

01:25:48 ◼   ► and you're writing it for some other reason,

01:25:50 ◼   ► like to make an argument or to promote something,

01:25:52 ◼   ► or whatever the case may be, it is really good for that.

01:25:56 ◼   ► And so I kinda wanted to try it from that point of view.

01:25:59 ◼   ► I'm writing this post that is, it is an idea,

01:26:03 ◼   ► it is kind of to promote overcast,

01:26:05 ◼   ► or at least that's an ancillary benefit of it.

01:26:08 ◼   ► My site has been pretty slow recently.

01:26:12 ◼   ► That's totally my fault, of course.

01:26:14 ◼   ► So the idea of taking any attention away from my site,

01:26:19 ◼   ► there wasn't that much attention on my site to begin with,

01:26:22 ◼   ► So it wasn't that big of an expense to try it.

01:26:25 ◼   ► It also wasn't exclusive.

01:26:26 ◼   ► They don't require it to be exclusive.

01:26:28 ◼   ► So I got to have all those benefits

01:26:30 ◼   ► of all that attention there

01:26:32 ◼   ► while not taking away from my site.

01:26:34 ◼   ► - Why does Medium get you more attention

01:26:37 ◼   ► than you posting it on your own site?

01:26:38 ◼   ► Those are just two URLs on the web.

01:26:40 ◼   ► Why is it that when posting it to your site

01:26:42 ◼   ► and posting it to Medium,

01:26:42 ◼   ► why do more people see it on Medium?

01:26:45 ◼   ► - That is a very good and very relevant question.

01:26:48 ◼   ► And I don't know the overall answer to that.

01:26:51 ◼   ► but certainly you can kind of tell on the web today,

01:26:55 ◼   ► it is pretty hard to get good traffic to a blog post.

01:26:59 ◼   ► It is much easier to get good traffic

01:27:02 ◼   ► in social environments.

01:27:04 ◼   ► And the fact is, you gotta go where the readers are.

01:27:08 ◼   ► You know, if you want something to spread like that,

01:27:10 ◼   ► you have to go where the people actually are,

01:27:12 ◼   ► where the consumers are.

01:27:13 ◼   ► And for the kind of things that I was writing,

01:27:16 ◼   ► Medium has a whole lot of those.

01:27:19 ◼   ► - Yeah, but I mean like physically, mechanically speaking,

01:27:23 ◼   ► how are the people there?

01:27:24 ◼   ► How does anyone find your post on Medium

01:27:26 ◼   ► other than seeing you link to it from your blog?

01:27:28 ◼   ► Like I don't understand,

01:27:29 ◼   ► obviously I don't use the service, so I don't understand.

01:27:31 ◼   ► - Well, there's like social recommendations

01:27:33 ◼   ► and there's like top voted in this time period

01:27:36 ◼   ► or among your friends or whatever.

01:27:37 ◼   ► - So it's kind of like tech meme or whatever you think.

01:27:40 ◼   ► People are browsing the front page of Medium.

01:27:41 ◼   ► I know there is a follower thing

01:27:43 ◼   ► 'cause I always get emails saying people followed me

01:27:44 ◼   ► on Medium and I feel bad

01:27:45 ◼   ► because I don't think I've ever written anything there.

01:27:48 ◼   ► But anyway, so you have an account

01:27:49 ◼   ► you have followers and presumably if I ever posted something to Medium my 17 followers

01:27:52 ◼   ► on Medium would see that I posted it, but are people I guess like using it kind of like

01:27:57 ◼   ► Reddit where you go to the front page of Medium and look at the top things top voted by people

01:28:03 ◼   ► you follow or something?

01:28:04 ◼   ► I suppose. I mean I haven't been using it enough to know. I mean none of us are doing

01:28:09 ◼   ► that obviously right? I mean whenever I see a lot of Medium posts you're right but when

01:28:12 ◼   ► I see them I see them in tweets basically. That's where I see links to Medium. And in

01:28:18 ◼   ► context from my perspective that could just as easily have been a link to Marco.org and

01:28:21 ◼   ► I would have seen it just as much but maybe other people are using Medium differently

01:28:25 ◼   ► and they're going to it like they go to Reddit pages and just or like Techmeme or anything

01:28:29 ◼   ► like that and just going to the what the hell is the front page of Medium? Let me go look.

01:28:32 ◼   ► It's like some like editorial collections of stuff I think. So in recent years every

01:28:37 ◼   ► time I write a post on my site these days I also tweet about it and the main reason

01:28:43 ◼   ► I do that is because the fact is way more people are reading Twitter than subscribe

01:28:50 ◼   ► to my RSS feed and checking my RSS feed regularly. And also Twitter provides feedback mechanisms

01:28:57 ◼   ► and ways for people to spread it with retweets and links and re-blogs or whatever. So there's

01:29:03 ◼   ► all these values that Twitter brings me in my publishing. Medium is another one of these

01:29:10 ◼   ► venues where there is a lot of activity happening there.

01:29:14 ◼   ► There's a lot of people reading it.

01:29:16 ◼   ► There's a lot of people recommending and sharing stuff

01:29:18 ◼   ► there to other people there.

01:29:20 ◼   ► And so the idea of cross-posting major posts there

01:29:24 ◼   ► doesn't sound that crazy to me anymore