74: One Of Us Shipped Something


00:00:00   Let me start by saying that I have started work on the Showbot title ordering, but it is not done.

00:00:07   And I apologize for that.

00:00:08   There was a poll request. Someone sent you the code.

00:00:10   I know. I know. But I don't know if I've only glanced at it, but I'm

00:00:14   a little opinionated about how I want it done. So I'm going to give it the college try.

00:00:19   One of us shipped something.

00:00:21   You're so mean to me. I'm going to give it the college try and see if I can do it the way I want

00:00:27   to do it and if I can't do it in basically the one time I sit down to do it, then I'll

00:00:33   take the pull request and walk away.

00:00:35   How about this alternate strategy?

00:00:37   Accept the pull request and let us use that while you work on your solution, then revert

00:00:41   that integration and put in your change transplant.

00:00:44   To be honest, that's probably the better approach, but that's all right.

00:00:49   All right, so nothing really happened this week, so let's start with follow-up.

00:00:53   Apparently some stuff we've said in the past has been wrong.

00:00:56   Really?

00:00:57   I would like to talk now for a half hour to explain why it was wrong.

00:01:00   All right.

00:01:01   John?

00:01:02   This first item is, I didn't put this first one in there.

00:01:04   Who put the first item in?

00:01:05   Oh, that might have been me actually.

00:01:07   We got a lot of feedback that when we were talking about what would cause someone to

00:01:13   switch to Android from an iPhone and we got a piece of feedback that we talked about an

00:01:19   episode or two ago that said, "Hey, a lot of these new versions of iOS can slow down

00:01:28   your phone and make the response time to everything a lot less quick."

00:01:34   And that's maybe why people switch.

00:01:36   And so somebody wrote in, I don't have who it was, but somebody wrote in, "Many people

00:01:39   have noted the perceived slowdown on older phones could likely be purchasing an old phone,

00:01:45   For example, buying a 4S today, which is still on store.apple.com.

00:01:50   Actually I think I wrote that.

00:01:52   The point being that you can still get a 4S today, and iOS 8 is presumably coming out

00:01:58   in September, and probably isn't going to run too well on a 4S, because I believe it

00:02:03   is supported.

00:02:04   Is that right?

00:02:05   The 4S is supported on 8, that's correct.

00:02:07   4S and up.

00:02:08   So all the A5 devices are supported on 8.

00:02:10   Which is interesting because there's a lot of them.

00:02:12   Like if they cut off the A5 device,

00:02:14   if they cut off the original iPad mini,

00:02:17   which is still for sale.

00:02:18   - Exactly, so there's a lot of older devices being sold

00:02:22   that are still going to be supported.

00:02:24   And so that's something to consider

00:02:27   and something that would perhaps drive you away from Apple

00:02:30   because here you get this perhaps not amazing experience.

00:02:34   You know, what if this was your first iPhone

00:02:36   or first smartphone and you buy a 4S today

00:02:39   and it runs all right with iOS 7

00:02:41   And then in a couple of months, you see that iOS 8 is available, you go and update it,

00:02:46   and all of a sudden it runs like crap.

00:02:48   I mean, that's no bueno.

00:02:50   So that's just something to consider.

00:02:52   We talked about that in past shows.

00:02:53   We were talking about why the heck the iPad 2 was around so long.

00:02:56   They finally stopped selling that, right?

00:02:58   I don't know.

00:03:00   You're probably right.

00:03:01   Anyway, yeah, we've had that same complaint.

00:03:03   And another one of my hobby horses is not putting enough RAM in things, especially when

00:03:08   you can't upgrade it.

00:03:10   It's short-sighted from a sort of customer satisfaction and brand satisfaction level

00:03:15   to skimp on RAM, even though it helps your margins, because some people will say, "Well,

00:03:20   I'm just going to get it with the minimum," and if the minimum is not a reasonable amount,

00:03:24   like if you're just setting someone up for -- it's much better now in the days of SSDs,

00:03:27   but when it was spinning disks, it was just torture to get a Mac with the minimum amount

00:03:30   of RAM and then use it for several years.

00:03:32   And by the end of its lifetime, the OS and all the apps need way more RAM, and the thing

00:03:35   was always swapping, especially it was like a 5400 RPM laptop drive and it was bad.

00:03:43   When we talked about this, the whole having an upgrade that slows down your thing, I don't

00:03:47   think any of us thought that wasn't a phenomenon that happens.

00:03:51   I think we've all actually owned devices that we've either decided not to upgrade to iOS

00:03:56   7 or regretted upgrading to iOS 7, and all the way back.

00:03:59   I remember on one of my original iPod touches, I think I didn't want to upgrade to 4 or something,

00:04:05   and then eventually I gave in and regretted it.

00:04:07   This is a thing that happens

00:04:08   if you keep devices long enough.

00:04:12   I actually, I did add two tiny notes

00:04:13   to this bit of follow-up here, I guess.

00:04:15   The first one is that if you buy a device,

00:04:18   if Apple sells you something

00:04:20   that they probably shouldn't still be selling,

00:04:21   like you get an iPad 2

00:04:23   and when it was so clearly long in the tooth

00:04:25   or you end up with a current non-retin iPad Mini

00:04:28   or poor Marco's mom buying the iPhone 4

00:04:32   when she shouldn't have been.

00:04:34   That's Apple's fault for selling those things.

00:04:37   And the customer gets it and they get a bad experience.

00:04:40   Sometimes it's just a bad experience

00:04:41   right out of the box, right?

00:04:43   But the idea that the solution to that

00:04:47   is to try another vendor, in some sense,

00:04:50   makes sense if you don't know anything about Apple

00:04:53   and its products or whatever.

00:04:53   You bought an Apple thing, it sucks,

00:04:54   you try a different one, right?

00:04:56   But for tech savvy people,

00:05:01   I would imagine that they could have a similar reaction,

00:05:04   but for a different reason.

00:05:05   Tech savvy people are gonna do it for the airline reason,

00:05:07   where like it's vindictive, where you're like,

00:05:09   well, I bought this thing, I was satisfied with it,

00:05:12   I used it for a while, I upgraded it,

00:05:14   Apple shouldn't have even allowed it to be upgraded,

00:05:16   so that's Apple's fault, I hate Apple now,

00:05:18   my solution is for my next phone,

00:05:20   for my next tablet or whatever,

00:05:21   I'm going to buy from a different vendor.

00:05:24   And that can very quickly turn into,

00:05:26   you know, I just hate Apple not buying their stuff anymore,

00:05:28   when in reality the optimal solution for that person

00:05:31   who maybe used and enjoyed Apple products for a while

00:05:33   is the next time they buy,

00:05:35   you know, buy a higher end Apple product.

00:05:38   Spend more money, I'm sure Apple would love this,

00:05:40   spend more money on Apple products

00:05:41   and actually it'll work well.

00:05:42   Again, it's Apple's fault for selling these old devices

00:05:44   too long, it's Apple's fault for allowing the upgrades

00:05:46   to go on them when they don't perform well.

00:05:48   Arguably it's Apple's fault that the software

00:05:50   doesn't perform well, but like,

00:05:53   if for example I had a relative who had that experience

00:05:56   And I knew that they were satisfied with Apple devices in general.

00:05:59   I wouldn't recommend, oh, well, for your next one, you should get an Android phone,

00:06:02   if I think that they would not get a better experience out of that.

00:06:06   Yes, they get the satisfaction of saying, well, now I'm not going to buy Apple stuff anymore.

00:06:09   But it's like, really, you'd be super satisfied with an iPhone 5S for your next phone.

00:06:13   Right? So if you say you had a 4S, you upgraded to get slow.

00:06:16   If you really loved your iPhone for all of its life, except for this last part, which again, totally Apple's fault for putting them in this situation and everything.

00:06:25   it's like cutting off your nose to spite your face

00:06:27   by saying, well, the next phone I'm gonna get

00:06:28   is an Android phone, because that probably won't be

00:06:31   a satisfactory experience, especially if someone

00:06:32   just is used to iOS or whatever.

00:06:34   So that's one angle on this, the whole idea of just,

00:06:38   I'm angry, I'm gonna get revenge at Apple,

00:06:40   and I'm gonna do something that's technically not really

00:06:42   in my best interest.

00:06:43   And by the way, this is why, again,

00:06:45   everyone loves to spend money on Apple.

00:06:46   So this is why I recommend people,

00:06:48   don't get an Apple device if the only one you can afford

00:06:50   is the cheapest one they sell,

00:06:51   'cause the cheapest one they sell always sucks.

00:06:53   Like, that's true of, you know, you get an iMac,

00:06:56   but it was like the $1,000 iMac,

00:06:58   it's like 15% less expensive and 50% slower.

00:07:02   Like, and you know, if you can't get the expensive one,

00:07:06   you actually are probably better off with something else.

00:07:09   And then the other point I had on this topic is,

00:07:11   at a certain point, hardware and software

00:07:16   sort of catches up to the baseline of what the OS,

00:07:20   what people wanna do with the OS.

00:07:22   If you think about the early history of OS X,

00:07:24   where it was just super slow for just years at a time,

00:07:27   just years and years, it felt gross,

00:07:29   it felt like molasses, it was really slow,

00:07:31   it was compositing on the CPU,

00:07:33   even when they got the compositing on the GPU,

00:07:35   the drawing and the window resizing

00:07:37   and all the event handling, it was just slow, it was crappy.

00:07:41   At a certain point, the cheapest Mac you could buy

00:07:44   could pull down Windows fast, could scroll fast.

00:07:47   Like these days, no matter what Mac you buy,

00:07:49   buy the least expensive new Mac from Apple,

00:07:52   it will scroll a page, okay?

00:07:53   And that sounds stupid, but we went through years

00:07:56   where it couldn't scroll

00:07:57   and where everything was just super slow.

00:07:59   I think iOS devices are very close to getting to the point

00:08:04   where no matter what piece of crap you buy from Apple,

00:08:06   you'll be able to scroll through a webpage, okay?

00:08:08   It'll load a webpage, okay?

00:08:09   It won't, like, no OS upgrade will screw them

00:08:12   in the way that iOS 8 is probably able to screw

00:08:15   the lowest supported device today.

00:08:17   We're getting very close to that tipping point where the hardware is good enough for the

00:08:21   basics to work.

00:08:23   We're not there yet, though, and I think this is a painful period.

00:08:26   And in the beginning of iOS, like, forget it, you know.

00:08:29   The iOS 1 was like, once the 3G and especially the 3GS came out, you didn't even want to

00:08:33   look at those old devices anymore.

00:08:35   You could not go back.

00:08:36   But you do eventually hit some minimum threshold of, like, you know, on the Mac menus work,

00:08:41   window resizing works, scrolling works.

00:08:44   You can launch applications without waiting a year.

00:08:46   There's still tons of other things that you can do without, that will make your thing

00:08:50   feel slow or whatever, but we just need the basics to work and we're not quite there yet.

00:08:53   So I'm hoping this time period where Apple keeps doing the strategy of selling devices

00:08:57   for a long time, giving them OS upgrades for a long time, will eventually work out for

00:09:03   them when the cheapest device they're selling meets the minimum threshold required for just

00:09:08   the basics to not be awful.

00:09:11   - I think certain generations age better than others too.

00:09:14   Like in the case of the iOS devices,

00:09:16   where you'll have basically the old models

00:09:19   being kept around for two years or so.

00:09:22   Certain chips and models have a lot more headroom in them

00:09:28   than others, and so the original iPad was not kept around

00:09:33   because it did not have a lot of headroom,

00:09:34   had way too little RAM.

00:09:36   The A5 with the iPad 2 at the time had tons of headroom.

00:09:41   which is why they're still selling A5 devices today.

00:09:44   And the original iPad mini, for an iPad these days,

00:09:50   it's kinda sluggish, but it still works.

00:09:53   So that's why they can still sell it.

00:09:54   But running iOS 7 on the iPhone 4 was a bad move.

00:10:01   And I think the A6 and A7, especially the A7,

00:10:08   really has a lot of headroom in it.

00:10:09   So like the current 5C is the A6.

00:10:14   I don't know how long they're going to keep selling A6 devices.

00:10:17   There aren't any iPads with it anymore.

00:10:19   But I would guess that the 5C internals from the iPhone 5 with the A6,

00:10:26   I would guess those stick around for another couple of years.

00:10:28   And that really won't be that bad.

00:10:31   The RAM problem is going to be an issue for-- even for the A7 device.

00:10:34   It's not like the CPU is going to get bad.

00:10:35   But again, Apple's been so stingy with RAM,

00:10:37   They've gone multiple generations without bumping the RAM specs and these things.

00:10:42   And that is going to hurt it much sooner than like, you know, you're not going to top out

00:10:46   the A7 doing basic stuff.

00:10:48   It's fine.

00:10:49   The A7 is completely adequate for sliding views back and forth, scrolling them, responding

00:10:54   to tap events, doing all the basic things you're going to need that we were just talking

00:10:58   about.

00:10:59   But if applications start using more memory because the applications become more sophisticated,

00:11:04   that A7's not gonna save you if everything keeps getting swapped out and you can't fit

00:11:07   things in RAM, you play one game and it boots out all your Safari tabs and it's all sorts

00:11:11   of, you know, so that's why I say we're getting close to the threshold but we're not there

00:11:15   yet.

00:11:16   So I would think even the most expensive, fanciest iOS device you can get today has

00:11:19   still not met the minimum threshold of you can keep that device for like five years and

00:11:24   it will still be basically competent.

00:11:26   Whereas, I mean, I'm sitting next to a Mac right now, granted it's the top-end one, but

00:11:30   But again, if you bought a top-end Mac in 2008, it still runs OS X perfectly well.

00:11:36   I'm running a current version of Photoshop, I run lots of applications.

00:11:40   Granted I upgraded the RAM and all that stuff.

00:11:42   That's not true today.

00:11:43   If you buy a top-end iPhone today in a couple years, it will not be up to the task of running

00:11:49   all the applications that...

00:11:51   Even just the basics of just being able to...

00:11:52   I feel like, "Oh, every time I launch this, this other thing, it gets booted out of memory,

00:11:57   and I can't even run this game because it requires more RAM this thing has.

00:12:01   Well that's something that, well first of all real-time follow-up, I was wrong. They

00:12:05   are still selling the iPad 4, which has the A6 CPU. It's an A6X, but it's the same

00:12:10   CPU. So, my mistake. Also, I think you're right that high-end Macs can last longer,

00:12:19   but also don't forget, you're talking about a Mac Pro where almost everything's upgradeable.

00:12:23   And modern Macs, not only are most Macs sold not Mac Pros, but modern Macs don't even have

00:12:29   a lot of upgradeable parts anymore.

00:12:30   It used to be, you know, if you bought a MacBook or a MacBook Pro three years ago, five years

00:12:35   ago, you could upgrade the RAM and hard drive over time.

00:12:38   You could maybe swap out the old spinning disk for the SSD and get another couple of

00:12:41   good years out of it.

00:12:42   You can max out the RAM, you know.

00:12:43   Which is exactly what I did.

00:12:45   Right.

00:12:46   A lot of people do that.

00:12:47   It's a great idea to do that, especially now that SSDs are so cheap.

00:12:50   But the devices we buy today, because they are so not upgradeable, almost every Mac that's

00:12:59   sold now doesn't have replaceable RAM anymore.

00:13:01   That's what I was saying.

00:13:02   But if you buy a top-end Retina MacBook Pro today, everything's sealed up, can't upgrade

00:13:06   anything, I think that machine will be fine for the basics in five years.

00:13:10   Yeah, but I would say if you're specing out a MacBook Pro, I would say get the most RAM

00:13:16   you can afford up front.

00:13:19   the RAM will make the biggest difference over time with how well it ages.

00:13:23   Yeah, someone just asked me on Twitter recently if they should—I forget which machine they

00:13:26   were specing out—should I take the CPU upgrade to an i7 or should I take the extra RAM? And

00:13:31   it's a difficult choice, but I have to go with the RAM, because again, like, you know,

00:13:37   you can always do—like, if things take a little bit longer, fine, but RAM is like capabilities.

00:13:41   Speed versus capability. Do you want to be able to run these two programs at the same

00:13:44   time without swapping, or do you want these programs to run 15% faster? And being able,

00:13:51   like capability is more important than speed.

00:13:53   Right, and the difference between 8 and 16 gigs of RAM, say, or especially, is there

00:14:00   anything still with 4 gigs of RAM in the lineup, maybe the low-end MacBooks? But the difference

00:14:04   between the base amount of RAM and twice as much of that is a much bigger difference usually

00:14:10   than the CPU choices you have where, you know, for the same price of doubling the RAM, you

00:14:15   might get an extra, you know, 5% CPU performance. It isn't a big jump on the CPU for the money.

00:14:22   It used to be better, but these days, yeah, you don't get much for your extra CPU money.

00:14:28   Exactly. And, you know, going back to the iOS devices, I think this is why you're right

00:14:31   to focus on RAM because that is the limiting factor so often. Like, that was definitely

00:14:37   the case with the iPad one. And feedback from the chat, yeah, tons of stuff with 4GB of

00:14:41   RAM still coming as stock, some of the low end stuff.

00:14:44   I was going to make a snarky comment, but I'm sure Apple is still selling stuff with

00:14:47   4GB, but I wouldn't have actually believed that this list is true, so now I'm even more

00:14:51   depressed.

00:14:53   We are sponsored this week, once again, by our friend Matt Alexander and his company,

00:14:58   Need. Need is a refined retailer and lifestyle magazine for men. Each month, Need curates

00:15:05   and sells a limited quantity of exclusive products from the world's top men's brands.

00:15:10   These collections are presented in the form of a monthly editorial, built around a certain

00:15:14   theme, and are shot by local independent photographers.

00:15:18   Beyond Clothing, which they have a lot of really nice clothing, but Beyond Clothing,

00:15:22   Need also sells coffee, literature, furniture, and so forth.

00:15:25   So you can tell this is written by a British person because he uses the phrase "and so

00:15:28   forth" in the script.

00:15:29   Soon, they'll localize to certain cities around the world, the first of which will

00:15:32   be London.

00:15:34   just launched volume 8 assembly it features some of the best products for

00:15:38   evenings with friends hosting parties and so forth see look again very British

00:15:43   their favorites this month are the Stewart throw which Matt tells me throw

00:15:48   means blanket I guess but it's fancier if you call it a throw it could at all

00:15:53   isn't that also a pillow is couldn't they read a throw pillow that that

00:15:56   exists right I think so John you're domestic is that right you're asking me

00:16:01   about the terms for furniture items from a British ad.

00:16:04   - That's correct.

00:16:05   - Not my area of expertise.

00:16:06   - Okay, well, if you want it to become

00:16:08   your area of expertise, go to need.

00:16:10   So their favorites are the Stuart Throw,

00:16:12   which might mean a blanket, the Schwood sunglasses,

00:16:16   that has to be something cool,

00:16:17   and the classic Oxford watch.

00:16:20   See, I wish I could have a British accent

00:16:22   'cause I feel like saying the classic Oxford watch

00:16:25   would be so much cooler if I had a British accent.

00:16:27   - That's true, and it is a damn fine looking watch.

00:16:29   It really is.

00:16:30   way better than all the smartwatches

00:16:32   that our industry's coming up with these days.

00:16:35   All right, so there's also a special offer for ATP listeners

00:16:39   anyone who places an order with need and was sent from ATP,

00:16:42   send an email to hello@neededition.com

00:16:46   with the subject line overcast trousers.

00:16:49   (laughing)

00:16:51   Once again, that is send an email to hello@neededition.com

00:16:54   subject line overcast trousers.

00:16:57   - And to be clear, to be clear, that's need edition

00:16:59   like a newspaper edition, not edition like mathematics.

00:17:03   - Maths, if you're British.

00:17:05   - Correct. - It's maths.

00:17:06   So yeah, no, no, it's edition, EDI,

00:17:08   you know, that kind of edition, neededition.com.

00:17:12   Subject line overcast trousers.

00:17:13   Anyway, they will throw in a bunch of extras

00:17:15   with those orders, so you know, things like magazines,

00:17:18   field notes, socks, scarves.

00:17:20   See, what I've learned from need is that a scarf

00:17:23   is not only something you should wear in the wintertime

00:17:25   when it's freezing.

00:17:27   You can also wear them when it's less freezing as a fashionable item to improve your fashion

00:17:32   score.

00:17:34   So anyway, they will, and then also if you do that you will also be put on a list to

00:17:38   get 25% off your next order.

00:17:40   So go to NeedEdition with an E not an A, NeedEdition.com and once again email them after you place

00:17:47   your order at hello@neededition.com, subject line overcast trousers to get all sorts of

00:17:52   cool stuff and a discount on your next order.

00:17:55   Thanks a lot to Need, once again.

00:17:58   They can really show you how to be much more fashionable

00:18:01   and cool than I could ever possibly explain to you,

00:18:03   because I am not fashionable or cool.

00:18:05   However, I've had some things from Need,

00:18:06   and they make me look much more fashionable and cool

00:18:09   than I really am.

00:18:10   (laughing)

00:18:14   - Well done.

00:18:14   No, thanks to Matt Alexander and the Need crew.

00:18:17   That's extremely cool.

00:18:19   - I got red shoelaces from them.

00:18:21   That was amazing.

00:18:22   It turns out if you take a regular pair of shoes

00:18:25   and you put red shoelaces in them,

00:18:27   they become fashionable shoes.

00:18:28   - We have some more follow up from Melissa Savage,

00:18:32   who is the wife of one of the co-hosts

00:18:35   of the Mobile Couch podcast,

00:18:37   and she actually has her own podcast,

00:18:39   which I don't remember what it is,

00:18:40   so I'm a terrible person.

00:18:41   - Very helpful.

00:18:42   - Yeah, thank you.

00:18:43   She wrote in to say,

00:18:45   "I thought I'd just drop you a note about screen size

00:18:48   "because it illustrates why it's important

00:18:50   "to have more women in tech.

00:18:52   We are on average smaller than men, meaning our reach is not so large.

00:18:55   Despite what you say about getting over the bigger screen within a week,

00:18:58   the iPhone 5 annoys me daily because I can't quite stretch my right thumb to the left corner,

00:19:03   site of a thousand and one key functions for most apps."

00:19:06   This is where, of course, the back button is.

00:19:08   So she goes on to say, "Not only this, but women's clothes almost never have pockets,

00:19:13   or when they do, they are crappy tiny ones.

00:19:15   The iPhone 5 doesn't fit properly into any of my jeans' pockets.

00:19:18   Modern purses have phone slots, but these are not large, maybe four inches by a half an inch.

00:19:23   So, real time follow up, by the way, her podcast is Silver Screen Queens. Anyway, the point being

00:19:30   that we were lamenting the fact that our jeans pockets, or if you're Matt Alexander,

00:19:35   our trouser pockets are not big enough. Wait, are those the same things? I thought

00:19:39   trousers were like khakis or are trousers all pants if you're a British person?

00:19:44   I thought they were all pants. Don't say pants to a British person.

00:19:48   I thought they were all the things that we put over our underwear.

00:19:51   We'll have to get some clarification with Matt Alexander on the underwear hierarchy.

00:19:55   Any one of those CPG Grey videos to show the hierarchy of the lower body of men.

00:20:04   So a random stranger in the chat room with no credibility is saying that trousers are

00:20:08   all pants.

00:20:09   And by pants you mean not underwear.

00:20:11   Yeah, that's a good question.

00:20:13   This is very confusing.

00:20:14   Oh, God.

00:20:15   Why did the British take our language and mess it up so badly?

00:20:18   Seriously, they ruin everything.

00:20:20   So yeah, so to come sort of back to point, the idea is that here it is we're lamenting

00:20:25   the fact that our trouser/jeans pockets are not big enough, and we actually have pockets

00:20:31   that are designed to fit things.

00:20:33   Whereas women don't often have pockets at all, and occasionally when they do have pockets,

00:20:39   they're tiny.

00:20:40   And that is something that, quite frankly, I did not even consider.

00:20:42   Well, not just the—I mean, there's two points here.

00:20:45   One is the clothing thing.

00:20:47   And the second is small hands and reaching the corners of the screen.

00:20:53   Even when the iPhone, when the iOS devices got taller, it was a little bit harder to

00:20:57   reach over there.

00:20:58   But here's the thing about the women in small hands in reach issue.

00:21:02   When I see people out on the street with comically large phones, it seems to be most frequently

00:21:08   women with comically large phones.

00:21:11   And I don't know why that is.

00:21:12   You would think if they, on average, have smaller hands, they would favor smaller phones.

00:21:17   That is not what I see.

00:21:19   I mean, at the very least it's equal, but my impression is that you are more likely

00:21:25   to see just a really big phone in use by a woman than a man.

00:21:30   I don't know if that—does that match your experiences?

00:21:33   No.

00:21:34   Oh, really?

00:21:35   See, I have seen only a handful of comically large phones, but the first one that I saw

00:21:41   amongst my peer group was a dear friend of mine who is extremely petite, has a just ridiculously

00:21:50   large phone. To me it looks like an iPad mini.

00:21:52   - See, for me, I think first of all, the smaller hands and reach issue is not just a woman's

00:21:58   issue. My hands are kind of medium-sized and when Apple increased the screen size with

00:22:05   the iPhone 5, I had changed the grip I had on the phone. I used to hold it a certain

00:22:10   with all the old size and then I had to like kind of kind of choke up a little

00:22:15   bit on the phone like in in t-ball terms because that was last time I played

00:22:19   sports I had to choke up a little bit to get a better reach when the five came

00:22:24   out and I there's still like certain things you gotta like kind of kind of

00:22:27   hold it back a little bit in your hand like dude I had to adjust the way I even

00:22:30   touch it to hit certain corners and so I don't think it's necessarily a men and

00:22:35   women issue I think it's just like I mean there's already even within you

00:22:39   Within both genders, there's a pretty big range

00:22:41   of hand sizes.

00:22:42   And I think a lot of, like iOS has,

00:22:45   with 7 especially, when they added the edge swipe gesture

00:22:49   to go back and stuff like that,

00:22:51   like they're making the interface more screen gesture based

00:22:55   instead of based on tapping a certain button

00:22:57   in a certain spot on every screen,

00:23:00   overall helps this problem because it makes it

00:23:02   so that you don't have to touch

00:23:04   a very certain far away area so often.

00:23:07   You can just use these gestures.

00:23:08   And I think this is the kind of problem that can be worked around with those kinds of decisions

00:23:13   for the most part.

00:23:14   I mean, obviously within reason.

00:23:15   You can't really use an iPad one-handed.

00:23:16   It kind of sucks.

00:23:19   And the five-and-a-half inch phone, which apparently might now be fake or delayed or

00:23:23   – it's one of those things where the analysts – there's basically zero evidence for

00:23:27   the five-and-a-half inch phone anymore, so the analysts who all predicted it are now

00:23:30   saying, "Well, it's going to be delayed until next year," because they don't want

00:23:33   to just say they were wrong.

00:23:36   I forgot where I was even going with this.

00:23:40   You were onto something with the grip thing, because when I see people using large phones

00:23:44   just really big ones that you notice, you're like, "Whoa, is that a tablet?" No, that's

00:23:48   their phone. I see them using it in a different way. They're not one-handing it

00:23:52   with a thumb on the thing. Maybe they're two-handing it. Maybe the ones that have a stylus

00:23:56   obviously, that's totally the Galaxy Note things or whatever.

00:24:00   You use it in a different way. And you do see, remember back when

00:24:04   we had sliders with QWERTY keyboards and everything,

00:24:06   the whole deal was people would pull out the phone,

00:24:09   hold it sideways, flip down the keyboard,

00:24:10   and it was kind of like, you know,

00:24:12   in landscape orientation,

00:24:14   where you're holding it on the side

00:24:16   like it's a Game Boy Advance.

00:24:17   So I'm making game references, you guys don't get it.

00:24:19   Anyway, landscape.

00:24:21   And using two thumbs on it with the keyboard,

00:24:23   but now that it's touch-oriented, that grip has gone,

00:24:26   and for a while we were in the one-hand thing

00:24:28   with the thumb on the thing,

00:24:29   but now with the larger phones, I see more two hands,

00:24:31   and like two-hand portrait,

00:24:33   two-hand landscape, one hand holding it, one hand tapping it,

00:24:35   like old people use remotes, you know?

00:24:37   You hold the remote in one hand,

00:24:40   and then you take the second hand and come from above

00:24:42   with a pointer finger and you hit the buttons

00:24:43   on the remote.

00:24:44   None of these things are, I think, necessarily bad.

00:24:49   People like the devices and are choosing to,

00:24:52   again, that's why Apple's gonna make a phone

00:24:53   with a bigger screen.

00:24:54   People like the bigger phones.

00:24:55   They're adapting their usage behavior to adjust for it.

00:24:58   But as Melissa writes in,

00:25:00   If you like to use your phone with one hand,

00:25:03   like with the traditional sort of iPhone style grip,

00:25:07   you will be annoyed with,

00:25:08   along with the rest of the people who like that,

00:25:10   no matter what your hand size,

00:25:12   that the phones keep getting bigger.

00:25:14   'Cause I mean, like I said, when the phones got taller,

00:25:16   I found it harder to reach some sections on the screen.

00:25:18   You had to do a little shimmy where you,

00:25:19   remember we talked about that

00:25:20   when the iPhone 5 birth came out?

00:25:22   - Yep. - You had a little

00:25:23   grip shimmy, right?

00:25:24   And it's gonna be even worse.

00:25:25   We're going to have to adjust our habits

00:25:27   and if Apple stops selling phones of this size,

00:25:29   then we'll just suffer along with everybody else.

00:25:32   - Yep.

00:25:33   All right, a little bit of quick follow-up.

00:25:36   John, this is for you, I believe, about TVs.

00:25:39   - Oh yeah, we're talking about 4K and plasma,

00:25:42   and one of the reasons that Panasonic

00:25:44   exited the plasma business is it's gonna cost them

00:25:45   a lot of money to try to get a 4K plasma TV,

00:25:48   because those little cells or pits

00:25:51   that are in plasma televisions

00:25:52   that make up the picture elements,

00:25:54   if you were to quadruple the number of those,

00:25:57   or roughly quadruple the number of those

00:25:59   in the same area, they have to get much, much smaller, and that's harder to manufacture.

00:26:03   And AWACS—I think this was on Twitter—sent me a link to say that Panasonic actually did

00:26:08   make a 4K plasma television.

00:26:11   It's 152 inches diagonal, though.

00:26:13   That way, you don't have to make the cells any smaller.

00:26:16   You just make the television massive.

00:26:17   They said it weighs 1,300 pounds and draws 3,700 watts.

00:26:21   So it's not cheap either.

00:26:23   That must have a six-figure price tag.

00:26:26   It is, yeah.

00:26:27   I don't think I remember what the price was.

00:26:28   But anyway, we'll put the link in the show notes just to go to see.

00:26:30   Again, it's not the fact that 4K is an impossibility for plasma.

00:26:34   It's the technology to make the little picture elements, whatever they're called in plasma,

00:26:38   smaller.

00:26:39   They would have to do a lot of R&D to get that working.

00:26:41   But I'm assuming this gigantic television was simply whatever their latest and greatest

00:26:46   regular high-definition plasma process was and just made it way bigger with the little

00:26:50   cells being exactly the same size.

00:26:53   If only there was already a word for "little picture elements."

00:26:56   not pixels, because the pixel is the resulting image on the screen. I'm trying to talk

00:27:00   about—and there is no equivalent in CRTs because it's just a big vacuum and an electron

00:27:03   gun and scanning over the phosphorous. I think they're called cells.

00:27:08   We are also sponsored this week by our friends at Squarespace. Squarespace is the all-in-one

00:27:12   platform that makes it fast and easy—can you believe that? Both fast and easy—to

00:27:18   create your own professional website, portfolio, or online store. For a free trial and 10%

00:27:23   visit squarespace.com and enter offer code ATP at checkout. A better web starts with

00:27:30   your website. Squarespace is simple and easy. They have drag and drop content editing, very

00:27:37   easy page builder graphical tools, and also you can jump in there and you can get down

00:27:41   to the markup and you can inject your own script tags if you want to, you can write

00:27:45   raw markup if you want to, the whole range easy to expert. They have a beautiful design,

00:27:51   beautiful designs for you to use on your site.

00:27:53   All of their designs come with responsive mobile designs

00:27:56   that match your style.

00:27:57   So your site looks good on every device every time.

00:28:00   They offer 24/7 support through live chat and email,

00:28:03   and they have a massive support staff located

00:28:06   in both New York City and Dublin, Ireland.

00:28:08   Squarespace plans start at just $8 a month,

00:28:11   and that includes a free domain name

00:28:12   if you sign up for a whole year upfront.

00:28:13   And you might as well, because let's face it,

00:28:15   you're not gonna move your hoof that often.

00:28:16   So start at $8 a month if you sign up for a whole year,

00:28:19   Get your free domain.

00:28:21   Squarespace also features commerce on any plan.

00:28:24   All their plans come with this.

00:28:26   You can add, if you want to, a store,

00:28:28   an online store to your site.

00:28:30   And you can sell physical goods, digital goods.

00:28:32   They manage all sorts of stuff for you,

00:28:34   tracking, orders, and everything else.

00:28:36   Really fantastic.

00:28:37   Start a trial today of Squarespace

00:28:39   with no credit card required.

00:28:41   It's a real free trial.

00:28:42   No credit card required.

00:28:43   You know, this is where you forget to cancel it

00:28:45   and they automatically bill you, nothing like that.

00:28:47   Start building your website today.

00:28:49   Check it out.

00:28:50   When you decide to sign up for Squarespace,

00:28:52   make sure to use offer code ATP

00:28:54   to get 10% off your first purchase

00:28:57   and to show your support for our show.

00:28:59   Thank you very much to Squarespace for sponsoring ATP.

00:29:02   Squarespace, a better web starts with your website.

00:29:05   - So Marco, big day?

00:29:07   - Little bit.

00:29:09   I'm so tired.

00:29:10   - When'd you wake up?

00:29:13   - I didn't sleep well last night,

00:29:16   so it's kind of a blurry line.

00:29:18   Well, that was my next question was did you sleep any?

00:29:21   When did you actually remove yourself from bed?

00:29:25   - About eight, I decided I was probably done sleeping

00:29:28   and I had to come downstairs

00:29:30   'cause I had the launch scheduled for 11

00:29:33   and I had it in the state in iTunes Connect

00:29:35   where it just has an availability date set in the future.

00:29:38   Because the problem was, you know,

00:29:41   so I wanted to launch it today.

00:29:42   If you set it to be released in the App Store today,

00:29:46   then it'll start going on at midnight last night.

00:29:49   And I didn't want to stay up all night if there were server

00:29:51   issues.

00:29:52   Because that was the big question mark was,

00:29:54   will the server stay up under the load of however many people

00:29:58   tried out on day one?

00:29:59   I had no idea, because I didn't really know how heavy

00:30:03   is each user on the server.

00:30:04   I really had no measure of that.

00:30:07   You didn't do load testing with a bunch of VMs

00:30:09   and hit it from-- that's the great thing about EC2 and stuff.

00:30:12   You can just get a fleet of instances briefly

00:30:15   and bombard your service to simulate the load

00:30:17   of thousands of users.

00:30:18   And I mean, it costs some money,

00:30:20   but you can just ditch them all when you're gone.

00:30:21   - Yeah, I mean, I thought of that.

00:30:23   It would have been very hard like to mimic

00:30:26   the exact usage pattern of overcast users

00:30:29   without having a bunch of iPhones connecting at weird times,

00:30:31   having podcasts being released at certain times.

00:30:33   It would have been hard to simulate well.

00:30:36   And so instead I opted to use that wonderful cloudness

00:30:39   of modern hosting to instead just temporarily

00:30:43   to get way more resources than I actually needed.

00:30:46   So I temporarily, a couple days ago,

00:30:48   I went from two web servers on Linode to eight,

00:30:51   and a very easy way to create more if I needed to.

00:30:54   And it turns out I ended up needing way less than that.

00:30:56   Anyway, so I was scared of what the servers would do,

00:31:00   and I didn't wanna be kept up all night.

00:31:02   So instead, I set it to be released tomorrow,

00:31:05   and then today at eight a.m., I changed the date back,

00:31:08   which makes it immediately get released at that point.

00:31:10   But immediately released on the App Store

00:31:13   doesn't mean it's immediately available to everybody

00:31:15   whenever they check and showing up in search results.

00:31:17   Because it has to propagate to different CDMs

00:31:20   around the world, different servers being replicated.

00:31:22   However Apple does it, it usually takes a few hours.

00:31:25   So I wanted to launch at 11 a.m.,

00:31:27   so I made the switch of the date at eight a.m.,

00:31:29   figuring that by 11 it would be available everywhere

00:31:32   and showing up in search results.

00:31:34   And turns out that was too early.

00:31:37   It actually showed up by about 10 or 9.30 even.

00:31:39   It only took like an hour and a half.

00:31:41   But I waited anyway and I had all the people

00:31:44   who had advanced beta copies, the depressed people,

00:31:47   I had told them all, don't publish before 11.

00:31:49   So I waited and I didn't publish before 11 either.

00:31:52   And yeah, so spent the morning just looking over everything,

00:31:56   tweaking everything and then I launched it

00:31:58   and the response was strong immediately.

00:32:03   I was scared of lots of potential things.

00:32:08   I knew that I had made an app that I was very happy with.

00:32:11   And my beta testers were generally very happy

00:32:14   with it as well.

00:32:15   But that's only like 30 people.

00:32:17   And this app was downloaded by thousands of people today.

00:32:20   So the opinions of 30,

00:32:23   especially this relatively undiverse group,

00:32:27   are not very reflective of the web at large.

00:32:30   So I didn't know what to expect from either the server load

00:32:33   or from just the reception of the app.

00:32:35   And are there any really crazy bugs

00:32:37   that are just really edge case that none of these testers

00:32:40   or I hit that would be major showstoppers

00:32:44   and that I'd have to freak out

00:32:46   and issue a quick emergency update.

00:32:48   And it turned out none of that happened.

00:32:50   It was great, it was just smooth.

00:32:52   I was so worried, I haven't eaten well

00:32:57   for a day before this.

00:32:59   Yesterday I was feeling terrible all day.

00:33:01   It ended up being nervousness.

00:33:03   Now my back is a mess.

00:33:05   I've been holding all this tension in my back

00:33:08   and my lower back is just destroyed today, it's awful.

00:33:12   'Cause I was really taking this nervousness inwardly.

00:33:17   But it seems like everything went very well.

00:33:21   I mean, the downside is I spent most of today

00:33:25   reading thousands of tweets and hundreds of emails.

00:33:28   I still, I'm mostly keeping up with the tweets.

00:33:32   I still have, at the time of reading this,

00:33:35   524 emails that I haven't gotten to yet.

00:33:37   - What are people email, is this like support emails

00:33:39   to the support address or personal emails to you

00:33:41   or is there a difference?

00:33:43   - It's, well there isn't really a difference yet.

00:33:45   I have a support person I'm going to hire

00:33:47   but I wanted to receive all the emails on day one.

00:33:51   First of all, just from a pragmatic standpoint,

00:33:54   I wanted to know about bugs immediately

00:33:56   so I could try to fix them.

00:33:58   - But these are emails that like from the application,

00:34:00   you're in the application, there's a button somewhere

00:34:01   you can hit that emails you,

00:34:02   that's what you're talking about, right?

00:34:03   - Yeah, yeah, and it's mostly people

00:34:08   with little feature requests, some bug reports,

00:34:11   some tech support kind of stuff.

00:34:13   I would try to be careful and set expectations

00:34:18   for tech support very low.

00:34:20   I don't use the word support anywhere.

00:34:24   And the descriptions of all these email addresses

00:34:27   is feedback.

00:34:28   And I say everywhere, I can't respond to everything.

00:34:32   - Well you have an auto responder that says,

00:34:34   please explain to me why you wouldn't want

00:34:36   the number one podcast app available on the app.

00:34:38   (laughing)

00:34:39   - Wow.

00:34:40   So anyway, I have a lot of email to go through.

00:34:44   I don't think it's that bad of a thing.

00:34:45   I'll see what happens day to day.

00:34:47   I didn't know if I should hire someone

00:34:51   because as far as I know, maybe by next week

00:34:53   it'll be like five emails a day.

00:34:55   I have no idea.

00:34:56   I don't know what to expect yet for daily volume.

00:34:58   And so until I know that, I'm not going to,

00:35:02   I'm not gonna make any decisions about

00:35:03   whether I'm gonna hire someone or not,

00:35:05   but I probably will end up doing it if I can.

00:35:07   But yeah, most of the emails are stuff

00:35:11   that takes a few seconds to look at

00:35:13   and maybe a minute to respond to

00:35:15   if I have to respond, a minute at most,

00:35:18   but when you have hundreds and hundreds and hundreds

00:35:20   of them, that adds up quickly.

00:35:22   - So the launch went well, the servers didn't croak,

00:35:25   all is good in the world?

00:35:27   - Yeah, I mean the database server load

00:35:29   never went above like .4,

00:35:31   which is awesome, 'cause that's the hard one to scale.

00:35:35   The web servers were negligible.

00:35:37   - Have you cranked back on some of the expanded footprint?

00:35:41   Like, have you gotten rid of some of those VMs,

00:35:43   or are they still--

00:35:44   - No, they're all still there.

00:35:44   I mean, if I keep all of these up for an entire month,

00:35:48   the entire hosting bill of Overcast is like $540 a month,

00:35:53   something like that.

00:35:54   When I sold Instapaper,

00:35:55   the hosting bill was about $7,000 a month.

00:35:58   - For more or less hardware?

00:36:00   Well, it was more overall horsepower,

00:36:02   but not by a whole lot,

00:36:04   because hardware has gotten better since then,

00:36:06   and Linode upgraded all their stuff.

00:36:08   So Linode's current pricing is almost cheaper

00:36:13   than most dedicated pricing for similar things.

00:36:17   It's extremely competitive.

00:36:19   Before this, I was gonna host it at Limestone Networks,

00:36:24   which is a really cheap, dedicated host.

00:36:27   And I had a server there for a while

00:36:28   when I was developing, doing crawling and stuff.

00:36:30   And that server was like 260 bucks a month

00:36:33   for what I'm, you can now get at Linode for about 160.

00:36:38   I mean, it's crazy how good Linode is right now.

00:36:41   - This is another example of hardware catching up

00:36:44   to something that doesn't change as much.

00:36:46   So yes, hardware always gets better,

00:36:48   but for textual data in small amounts,

00:36:53   it used to be that no matter how small

00:36:54   your textual data was, if you just wanted to store

00:36:56   an address book for people that's like bounded in size, only text, not a lot of things. Even

00:37:01   that would be super expensive just because hardware hadn't caught up.

00:37:04   Hardware has gotten so much better, people's address books are similar in size. Maybe they're,

00:37:10   you know, twice as big, five times as big, still mostly text, maybe with little pictures.

00:37:14   But that grows much more slowly than computing power, price performance, memory, and stuff

00:37:18   like that. So I think the cost of like, you know, fast forward 15 years, the cost of doing

00:37:22   a similar service to Overcast, like podcast feeds, keeping track of what podcasts are,

00:37:26   all their metadata, what their feed addresses, all that stuff.

00:37:29   That's not going to grow at the same rate as how cheap memory and storage and everything

00:37:33   grows.

00:37:34   So you're getting to the point where it'll be like, "Oh, I run my podcast metadata service

00:37:39   on a puck in my home that I keep in the attic and never look at."

00:37:43   Right.

00:37:44   And that's...

00:37:45   Well, honestly, the dream is not that.

00:37:46   The dream is closer to what Linode offers, because this thing runs somewhere else that

00:37:49   you don't have to think about it.

00:37:50   I know, but I'm talking about hardware requirements.

00:37:52   it would be like, oh, it's like $5 a month,

00:37:54   unlimited CPU and memory.

00:37:56   - Oh yeah, exactly.

00:37:57   I was thinking actually as a side note,

00:38:00   which I guess all side notes are temporary,

00:38:02   if Apple updates the Apple TV with an A7 chip

00:38:06   to take advantage of metal

00:38:07   and making a game console version and everything,

00:38:10   I wonder about the web hosting potential.

00:38:13   It's like Mac mini colo times two,

00:38:15   the web hosting potential of Apple TVs,

00:38:18   'cause A7s are really good chips.

00:38:19   - But it's not, I mean, for web hosting,

00:38:21   It's like you got to have memory to have all this the the disk stuff cache in memory

00:38:26   So you don't actually you know see you don't have to pull it off the disk ever your disk caches are all warmed up

00:38:31   Right so you need a lot of memory for that and you probably need lots of small weak cores more than what is a seven?

00:38:37   I've still just two two fairly expensive course. I think still something like I mean

00:38:42   That's why the server chips are all like tons and tons of relatively weak small cores, so you could be processing lots of

00:38:48   things at the same time and each one of those things is no longer that demanding if you're just getting and

00:38:53   Sending again sending and receiving small bits of textual information

00:38:56   so apparently

00:38:58   A similar company to to Mac Mini colo or whatever was anyway similar company called a Mac Mini vault

00:39:06   They host they already have a page host on an Apple TV

00:39:10   We'll put that link in the show notes. Hey, no fans in the Apple TVs

00:39:15   You don't have to, you know, I guess you have to add your own fans. Maybe if you packed enough Apple TVs in,

00:39:20   I mean, they do produce heat. If you packed enough of them in, you need something to push that heat into the hot aisle in the data center,

00:39:25   so you'd have to like add fans to the front of it. Very strange.

00:39:28   I imagine like, you know, computing power to density ratio has to be pretty good with an A7-based Apple TV.

00:39:36   That has to be pretty good.

00:39:37   Yeah, but I mean, like the Mac Minis though, these are not purpose-built hardware for the data center.

00:39:42   And the people who do make purpose-built hardware for the data center have access to more or

00:39:46   less the same technologies, right?

00:39:48   So it's not as if, you know, this is the optimal solution.

00:39:50   It's just funny and interesting from like, imagine if you just got a bunch of these pucks

00:39:53   and you put them in this little drawer.

00:39:55   But realistically speaking, Google's still making their own blades out of God knows what,

00:39:59   and it's, you know, even better.

00:40:01   So first day went well.

00:40:04   Do you care to share how many users you have?

00:40:05   I saw you tweeting earlier today how many signups you had.

00:40:10   What are we at in at least a vague order of magnitude?

00:40:14   - 21,800.

00:40:16   - Now is that accounts registered server side

00:40:19   or are those number of downloads from the App Store?

00:40:21   - That is accounts registered server side.

00:40:23   Number of downloads I won't have until tomorrow morning

00:40:25   when I get the stats.

00:40:26   - I mean you figure it would be similar

00:40:28   'cause it's, you know, you can't really use the app

00:40:30   until you sign up.

00:40:32   So if people downloaded it and didn't launch it I guess.

00:40:34   - Yeah, yeah, but yeah.

00:40:35   So I mean I would imagine the number of downloads

00:40:38   is probably substantially higher

00:40:40   because there are gonna be some people who downloaded it

00:40:42   and just haven't launched it yet.

00:40:44   Some people who downloaded it got to the login screen

00:40:46   and said, "I don't wanna create an account,"

00:40:48   and deleted it, or said, "I'll come back to this later

00:40:49   "and do this."

00:40:51   So I would guess number of actual downloads

00:40:54   might even be as high as twice that.

00:40:56   - Do you have the FAQ linked from that first screen

00:41:00   that says, you know, create an account or whatever?

00:41:02   - That's why I wrote the FAQ,

00:41:04   and that's why it's called the Skeptics FAQ.

00:41:06   It's because I knew people would say,

00:41:09   why do I need an account?

00:41:10   And that's why that's the very first question.

00:41:12   Or no, I think it's the second question.

00:41:14   You should keep track of it.

00:41:16   If they launch the app for the first time

00:41:18   and don't hit either one of those buttons,

00:41:21   the second time they launch the app,

00:41:22   you should make the Skeptic's Fact more bold.

00:41:27   Because I don't even remember seeing it there, but you want--

00:41:30   there is an explanation.

00:41:31   You might as well give it now.

00:41:33   There is an explanation of why do

00:41:34   I have to make an account of any kind for a podcast app.

00:41:38   - Yeah, and this was, I've had a couple,

00:41:41   I didn't know, this was another kind of risky thing.

00:41:44   I didn't know if there would be a lot of people

00:41:45   who were offended by this and who would skip it,

00:41:48   who would cancel their efforts to try Overcast entirely

00:41:52   just because they didn't wanna do that.

00:41:54   The reason the accounts are there

00:41:56   is because this is an entirely server-backed app.

00:41:58   The server does all the crawling, all the updating,

00:42:00   all the notifications and everything.

00:42:02   The server does a lot of the work.

00:42:04   And the app doesn't even have an XML parser.

00:42:07   the server does everything and just outputs JSON

00:42:09   to the app which it can decode quickly and natively.

00:42:12   Anyway, and so there has to be some kind

00:42:14   of user identification method between the app and the server.

00:42:17   Now if all I ever had was an iPhone app,

00:42:20   or even if all it ever was was iOS apps

00:42:24   that were all in the same account,

00:42:25   I could do things to kind of hack around this.

00:42:27   Like I could do like, you know,

00:42:28   generate a random ID the first time you use it

00:42:30   and then use that as your username behind the scenes

00:42:32   and never even show the user what their username is.

00:42:35   and then that could work just fine.

00:42:38   I could also store a user ID string in iCloud

00:42:42   in the key value store so that way they could launch it

00:42:45   on their iPad after I make an iPad version

00:42:47   and then it syncs over.

00:42:49   There are things like that I could do.

00:42:50   The big problem with that is first of all,

00:42:52   that would almost completely rule out web functionality.

00:42:55   Second of all, if I then had a way for them

00:43:00   to add an email address and password to this account,

00:43:04   Believe me, I know what would happen,

00:43:08   because I had similar issues with people

00:43:09   making accounts at Instapaper.

00:43:11   What would happen is people would make duplicate accounts.

00:43:14   They would, rather than associate an email address

00:43:17   with their current account, they would create

00:43:19   a new account inadvertently with the email address.

00:43:23   And then you have two accounts, and they email me saying,

00:43:26   all my podcasts are gone, because I logged into the new one.

00:43:29   They didn't know.

00:43:31   And then you have to find ways to merge accounts.

00:43:34   And it basically becomes a big issue with

00:43:37   perceived customer stability and data loss,

00:43:41   'cause they will do something weird

00:43:43   that makes them think they lost everything,

00:43:45   and then blame me and be very unhappy.

00:43:47   And also it's just a massive burden on support email.

00:43:51   And so I chose, you know what, let me try instead,

00:43:54   just require the account right up front.

00:43:57   Explain it as best I can.

00:43:58   I know a lot of people are still gonna not do it,

00:43:59   but let me just require the account up front,

00:44:02   I make it always require an email address

00:44:04   so I can do password resets.

00:44:06   And let's just see if that works.

00:44:09   Maybe that'll be fine.

00:44:12   And in the future, if it ends up a lot of people

00:44:14   are being turned off by that

00:44:15   and I really want their business,

00:44:17   then maybe I can add the system where,

00:44:19   okay, it just starts up with an anonymous ID

00:44:21   and then you can add an email address later.

00:44:22   But I would rather try this first

00:44:24   because if this can work well enough,

00:44:27   it is so much easier to support

00:44:29   and it's so much easier as a user

00:44:31   that it's always the same.

00:44:32   All right, and you say all these things as though you've been burned once or twice

00:44:36   before.

00:44:37   I don't know what that's about.

00:44:38   Yeah, exactly.

00:44:39   All right, so let me start with a couple of obvious questions, and then I have some hopefully

00:44:44   less obvious questions, and then as we talk, I'm sure Jon will start tearing apart little

00:44:49   bits and pieces of what you're saying.

00:44:50   I'm so looking forward to this.

00:44:52   Yeah, I do want to save that for probably the end if we don't get to it in the middle.

00:44:57   Why freemium?

00:44:58   I'm not that confident in the market for a paid up front app anymore, especially because

00:45:03   I wanted to charge a good price for it.

00:45:06   The model is, in summary, the app is free, there are some limits, the in-app purchase

00:45:11   removes the limits.

00:45:12   That's it.

00:45:13   It's one purchase, one time, five bucks, and that removes all the limits.

00:45:17   So it's kind of like a trial version.

00:45:19   It's kind of my hacky way of doing a trial version, except that the entire app doesn't

00:45:25   just certain things just don't work.

00:45:28   Certain things don't work unless you pay.

00:45:30   And the two big things that don't work,

00:45:33   the smart speed and voice boost,

00:45:35   you can actually demo them without paying for five minutes.

00:45:39   Like there's like a five minute trial of those features.

00:45:42   I actually wasn't even sure if Apple would allow that,

00:45:44   but they did.

00:45:45   So it is basically a trial version.

00:45:50   And again, so going back a minute, so that's five bucks.

00:45:53   I don't think, if I launched today in the App Store, I'm sure my day one sales at five

00:45:59   bucks would be decent, but first of all, I know I got way more people as this model than

00:46:06   I would with that model.

00:46:07   I know that.

00:46:09   Second of all, I know over time that would be very hard to sustain because once the initial

00:46:16   PR is over, and once all your friends and all your blog readers have bought it, and

00:46:23   once everyone who's going to write about it

00:46:24   has written about it.

00:46:26   Then the sales of every app just tail off like crazy.

00:46:30   They just drop.

00:46:32   If you look at the graph, it looks like a roller coaster.

00:46:35   And they settle into a point,

00:46:38   and then that eventually lowers and lowers and lowers.

00:46:41   If your app is paid up front, that happens faster,

00:46:43   and it happens more severely.

00:46:44   I've seen this happen.

00:46:45   And Instapaper was always that model the entire time I owned it.

00:46:48   It's still that model today.

00:46:52   I know that model very well, the paid-up-front model.

00:46:55   I also know that in today's App Store, in a competitive category where I don't even

00:46:59   have the most features and people are very, very picky with what they want and what they

00:47:04   don't want, I knew that a $5 paid-up-front app was not a good long-term solution.

00:47:12   So that's why.

00:47:14   So I found a way to make free work.

00:47:16   I saw with Instapaper there were so many people who I would come in contact with in real life

00:47:22   even, even like family friends. I would be visiting them and I'd see on their phone,

00:47:28   they were still using Instapaper free, even like two years after it just continued. There

00:47:33   are so many people who, no matter how much they like you, they don't pay for apps.

00:47:40   It isn't just that they won't, it's that they, in their minds, don't. That's just

00:47:45   something I don't do. That's the kind of mindset it is.

00:47:49   - Who buys batteries, Marco?

00:47:50   - There's a lot of people, this is not a small group,

00:47:56   a lot of people who really just don't buy apps.

00:48:01   And I knew that the biggest podcast app in the world

00:48:06   by a very large margin is Apple's podcast app.

00:48:11   The second biggest by a pretty large margin is Stitcher.

00:48:14   Both of those are free.

00:48:17   I wanted to make an app that's good and free.

00:48:19   'Cause the fact is the Apple podcast app is not bad.

00:48:22   It's not great, but it's not bad.

00:48:25   That's the biggest competition in the market,

00:48:26   and it's free, and it has a lot of features.

00:48:29   And it has some features I can never have,

00:48:32   like the integration with iTunes.

00:48:33   There are alternatives here.

00:48:35   I could have, like somebody in the chat said,

00:48:36   I could have done ads, but I've never seen an ad

00:48:38   in an app that I thought made the app look good.

00:48:41   Ads in apps are very intrusive,

00:48:45   because they take up so much space.

00:48:47   and the inventory is usually,

00:48:48   like I don't mind doing podcast ads here

00:48:50   because our advertisers are good

00:48:52   and because for the rest of the show,

00:48:53   we're giving like six minutes of ads

00:48:56   out of a 90 minute show.

00:48:57   When an ad's in an app,

00:49:00   not only are the advertisers usually terrible,

00:49:02   but it's taking up a pretty big chunk of the screen

00:49:04   all the time.

00:49:06   That's a much bigger cost on the user

00:49:08   than the kind of ads I respect,

00:49:10   like podcast and well done blog ads like the deck ads.

00:49:13   It's a very different ratio

00:49:15   and the advertisers are all cheap and crappy,

00:49:17   and so I just, I don't like app ads at all.

00:49:20   There are also features in the app,

00:49:22   like the Twitter feature, where, you know,

00:49:24   so I offer these like Twitter-based recommendations,

00:49:26   where you can connect your Twitter account,

00:49:27   and then you can get recommendations

00:49:29   based on the people you follow,

00:49:31   based on what they listen to,

00:49:32   and, you know, if they've connected to their accounts.

00:49:35   And that kind of feature works best

00:49:38   the more people you have, it's a social feature.

00:49:40   And this is why all social apps are free, right?

00:49:42   'Cause the social network value, exponential,

00:49:44   blah, blah, blah, and so this is the same way

00:49:47   where that feature becomes a lot better

00:49:50   if I have more people using it.

00:49:51   So I decided rather than get five bucks from everybody,

00:49:55   I was going to try to get five bucks

00:49:59   from a few of the people who use the app

00:50:01   and just try to make the app as cheap as possible to host.

00:50:05   And I focused on that from day one

00:50:07   because I did not want to get to the point

00:50:10   where I was with Instapaper where it was very,

00:50:12   very expensive to host this.

00:50:14   With Overcast, I have made every possible focus

00:50:18   on keeping it as cheap as possible on the servers.

00:50:21   That's why I'm on Linode.

00:50:22   That's why my big expanded version of the hosting

00:50:26   was only gonna be 500 bucks a month,

00:50:27   and I'm probably gonna get away with more like two or 300

00:50:30   after this is all done.

00:50:31   The whole point of this was cheap hosting.

00:50:34   Make this sustainable that way.

00:50:37   And so the next question is,

00:50:38   which a few people have already asked,

00:50:39   is why not a subscription price?

00:50:42   And I thought about that a while too.

00:50:45   My model for the in-app purchase,

00:50:47   I've been all over the place with it in planning this.

00:50:50   Originally I was gonna do everything is unlocked

00:50:53   and you just pay what you want for the app.

00:50:56   And that, I thought about that,

00:50:59   I got some input from some trusted people

00:51:01   and that just, it wasn't very good.

00:51:03   - Is that even allowed on the App Store?

00:51:05   How would you get the money from them?

00:51:07   - I actually asked them that.

00:51:08   (laughs)

00:51:09   Because that's a good question.

00:51:11   I talked to a couple people at Apple,

00:51:13   whoever, you know, the people I could find at Apple.

00:51:16   They don't necessarily publish a directory or anything,

00:51:18   but the people I could find who might have been relevant

00:51:20   to it at Apple, I tried emailing in and got various places

00:51:24   and all of them kind of ended with, well, maybe.

00:51:27   So it was kind of a question mark

00:51:29   as to whether that would be allowed.

00:51:30   And generally with App Store stuff,

00:51:33   you don't wanna live on a question mark edge of a rule.

00:51:36   What you said earlier, Jon, is exactly right,

00:51:38   which is the cost of hosting each user goes down with time.

00:51:42   So the only way that my, if I'm responsible

00:51:46   about how I host this and how I manage the resources,

00:51:49   the only way that the costs really meaningfully

00:51:51   go up over time is if the user base

00:51:53   is growing substantially over time.

00:51:55   Because otherwise if the user base stays the same

00:51:57   and usage stays the same, then the cost of hosting

00:51:59   will slowly decrease as hardware gets better

00:52:01   and hosting gets cheaper.

00:52:03   - And I won't be like Instapaper where people like me

00:52:05   just have this massive backlog where you're saving

00:52:07   just thousands and thousands of articles, even when you archive them, right? I forget

00:52:10   what your policy was on archive things. It used to be like a window, but if you paid

00:52:13   for it, you've had various reels. But anyway, people's Instapaper collections could in theory

00:52:17   grow, whereas podcasts, there's some working set, and then after you've played an episode,

00:52:22   you're not retaining that info, are you? Well, I'm retaining the row in the database, which

00:52:27   is the row of five integers of the podcast ID, the user ID, whether you completed it

00:52:33   or not, and when. Well, I was saying you could trim that stuff

00:52:36   Like, you don't need to keep a record of,

00:52:38   I listened to this episode three years ago,

00:52:41   because it's no longer visible in the app,

00:52:42   hasn't been visible in the app for years.

00:52:45   - Well actually, if the podcast has fewer than 1,500 entries,

00:52:48   it is still visible in the app.

00:52:50   But that's, you know, the reality is,

00:52:53   like I mean, I know from Tumblr how these tables grow,

00:52:56   and what that means, and what it costs to host,

00:52:58   and the fact is, a table full of, you know,

00:53:01   hundreds of millions of integer associations like that

00:53:05   is really nothing to host.

00:53:07   It sounds like a big number, but in reality,

00:53:10   a table with five integer rows,

00:53:12   your service could be as big as Tumblr

00:53:16   and that table might only be like nine gigs.

00:53:18   We're not really talking a massive amount of data here.

00:53:21   And that's the kind of table where the index

00:53:24   is bigger than the table usually.

00:53:25   - And you were saving the textified versions of web pages

00:53:29   in the Instapaper database, right?

00:53:30   Because if a--

00:53:31   - I was saving them on S3, I wasn't saving the database.

00:53:34   - All right, well, but somewhere you were saving them

00:53:36   because you had the, what do you call it,

00:53:38   if someone had a different view of a website

00:53:42   than somebody else, you couldn't just save that webpage

00:53:44   as they saw it and allow everyone to see that.

00:53:46   - Right, it was a way to work with login barriers,

00:53:50   so that if you were logged in,

00:53:52   if a site required a login barrier,

00:53:54   if you saved it with a bookmarklet,

00:53:56   the bookmarklet would send a copy of what you were seeing

00:53:58   to the servers which would save it just for you.

00:54:01   And anyway, so, where was I?

00:54:03   Oh yeah, so subscriptions.

00:54:04   So yeah, I didn't do subscriptions

00:54:06   because my costs aren't high enough

00:54:09   where I really need them.

00:54:10   Like I can get along just fine without them.

00:54:14   And I might do them in the future.

00:54:15   Like with Instapaper, when I sold Instapaper,

00:54:17   I don't know what it is now,

00:54:18   I haven't been bugging them about stats or anything

00:54:21   'cause honestly that's not my job anymore.

00:54:23   But when I sold Instapaper,

00:54:25   it was making about half of its income from subscriptions.

00:54:30   And the Instapaper, it was an interesting model.

00:54:31   It was, you know, the app was a few bucks,

00:54:33   Most of its life it was five, then it became three.

00:54:36   About halfway through the time I had it,

00:54:38   I launched these monthly subscriptions

00:54:40   that would allow me to offer a very small subset of users

00:54:43   very expensive features.

00:54:44   It started out with search.

00:54:46   Well, it started out with nothing,

00:54:47   then I added search.

00:54:48   Most people who subscribed did not subscribe

00:54:52   because of search.

00:54:53   Most people who subscribed subscribed out of

00:54:55   a sense of goodwill to me for providing

00:54:57   years worth of app updates without charging them

00:55:00   and basically building up a surplus of goodwill

00:55:04   among my audience.

00:55:06   That's kind of my plan here,

00:55:07   is if I need more money later on

00:55:08   to address hosting costs or whatever,

00:55:11   I'll do a few years of free updates.

00:55:14   People will like me enough that if I put in a tip jar

00:55:17   or something or a monthly voluntary subscription

00:55:19   that does nothing, you can make good money with that.

00:55:23   I mean, really, think about it.

00:55:24   I was making half of my income from those

00:55:27   when I sold Instapaper,

00:55:27   and most of those people never searched.

00:55:30   you know, that was not the reason they were buying it. So that's the plan for money,

00:55:36   is what I'm saying. Did I miss anything there? I'm sorry, it's been a long day. I'm kind of

00:55:40   mentally fried. So that last bit, though, you were saying, like, the idea of having

00:55:44   the equivalent of a tip jar, even if it technically is for search or whatever,

00:55:48   or having a subscription thing, you mentioned that those are both viable. It's not like you're

00:55:52   saying, "Oh, that can't work," I mean, because you know from experience that it can work in a

00:55:56   in a reasonable fashion.

00:55:57   What is it about that approach that you didn't like

00:56:00   that made you change your mind this time

00:56:03   and go with the $5 to unlock

00:56:05   the rest of the functionality thing?

00:56:07   - It really is a combination of not having

00:56:09   a lot of faith in that, 'cause keep in mind,

00:56:11   I sold Instapaper over a year ago.

00:56:13   And when I sold it, sales were not great.

00:56:15   And so, I saw the writing on the wall

00:56:20   for the paid-up front model back then,

00:56:22   and I'm not very confident in it now.

00:56:26   Part of it is also that now is a different model.

00:56:29   Instapaper had basically one and a half competitors.

00:56:32   The podcast app market has lots of competitors

00:56:34   and many of them are free.

00:56:36   And so I recognize there was stiffer competition here.

00:56:42   I was going into it, I was going into a very mature market.

00:56:45   Like with Instapaper, even though there was competition

00:56:48   from six months after I launched,

00:56:51   the competition and I evolved together.

00:56:56   Like we both started from the same spot of not much.

00:56:59   And then we were able to grow over time

00:57:02   in complexity and advancement.

00:57:04   With the podcast market, I'm going into it,

00:57:07   while everyone else who's in this market

00:57:09   has been writing their apps for four or five years.

00:57:11   These are now old apps.

00:57:13   They've been around the block.

00:57:15   They have tons of features, tons of infrastructure in place.

00:57:19   I was starting late to this game,

00:57:21   So I was going to come in with fewer features by necessity.

00:57:24   And so that's another thing, I tried with that.

00:57:28   I realized that I would need to do something

00:57:32   a little more dramatic than just launch to five bucks

00:57:35   if I wanted to get a decent user base.

00:57:37   - Well, what I was thinking of was comparing it to,

00:57:39   also launch it free, but have a $20,

00:57:43   I like Marco 'cause he makes nice things thing

00:57:45   that unlocks some feature that no one's gonna use,

00:57:48   like search, but people would do

00:57:49   out of the goodness of their heart.

00:57:51   Like that's the model I'm comparing it to.

00:57:52   Not the $5 upfront, but the large in-application,

00:57:57   quote unquote, subscription, whatever you wanna call it,

00:57:59   that isn't really unlocking anything.

00:58:02   - Like the big tip jar.

00:58:03   - Yeah.

00:58:04   - The main reason why I didn't do something like that

00:58:07   is that I just didn't think most people

00:58:08   would buy something that expensive.

00:58:09   And like one of my ideas was to have like a three-tier

00:58:13   system of like pay what you want, you know,

00:58:15   $2 a year, $5 a year, $10 a year, something like that.

00:58:20   And I was a little afraid that by making it complicated,

00:58:24   'cause those things are all kind of complicated,

00:58:26   by making it complicated,

00:58:28   then I think you reduce dramatically

00:58:31   the number of people who will do it.

00:58:33   Because some of those words kinda scare people away.

00:58:37   I was even thinking maybe using

00:58:39   the public radio vocabulary,

00:58:41   calling these things pledges.

00:58:42   Or you could be a supporter of the app.

00:58:46   - Would you run a telethon once or twice a year then?

00:58:50   If I did that, maybe I'd have to consider it.

00:58:52   I thought of all those models in various showers

00:58:57   I've taken over the last two years.

00:58:59   And I, 'cause that's in the showers

00:59:02   when I think about pricing for my app most.

00:59:05   I don't know why.

00:59:06   That's my life.

00:59:09   So I went with this model because it's simple.

00:59:12   And because it's a low barrier to entry.

00:59:16   The price is relatively high.

00:59:17   If it's five bucks.

00:59:18   I was tempted to charge 10 actually,

00:59:21   but I knew nobody would buy that.

00:59:22   And who knows, I might discount the in-app purchase

00:59:27   in the future, I'll play it with the price,

00:59:29   it's not hard coded anywhere.

00:59:31   I decided in the end to just keep it simple

00:59:34   because it's simpler for everyone,

00:59:36   it's simpler for me, it's simpler for the users,

00:59:38   this gives me flexibility, I can do things like,

00:59:41   in certain places, people who can't use in-app purchase,

00:59:45   I can put up a web Stripe buying form,

00:59:48   and then sync that over with their account

00:59:51   so it unlocks the app.

00:59:52   By keeping it simple, I have many options.

00:59:56   So that's really it.

00:59:58   - All right.

00:59:58   - We are also sponsored this week by Fracture.

01:00:01   Fracture, they sponsored us a while ago back,

01:00:03   I think in May they sponsored us.

01:00:06   Fracture is great.

01:00:08   Fracture prints photos directly on glass in vivid color.

01:00:12   It's really interesting.

01:00:13   So I have a bunch of Fracture prints around my office

01:00:16   'cause they're good.

01:00:17   I mean, you know, it started out obviously,

01:00:18   they were a sponsor, that's how I learned about them,

01:00:20   so disclaimer, but even when I'm not being sponsored

01:00:25   by them, I have on multiple occasions bought Fracture prints

01:00:29   at full price because I like them.

01:00:31   Fracture, their printing method, it's great,

01:00:35   like you upload a picture or whatever

01:00:36   and then you get it printed, and it's really,

01:00:39   the picture is printed directly on a thin piece of glass

01:00:43   that's then mounted to a thin piece of foam board

01:00:45   so that you can hang it up easily.

01:00:47   So these prints are relatively lightweight.

01:00:50   Like you don't have to worry about them

01:00:51   like falling off the wall and pulling your wall down

01:00:53   and shattering or anything.

01:00:55   They're just immensely practical

01:00:57   'cause what you get is a really nice looking glass print

01:01:02   that's border to border, frameless,

01:01:04   and then it is kind of its own frame.

01:01:07   You don't need to then like get a picture frame.

01:01:09   And compared to getting a picture framed,

01:01:12   it's an amazingly good value.

01:01:14   So I have a bunch of these things.

01:01:15   I have a couple of big ones,

01:01:17   showing off various nice pictures I've taken.

01:01:19   And then I have these three

01:01:21   that are the smallest size they have,

01:01:22   which I believe is five by five inches.

01:01:24   And I use those to print out app icon pictures

01:01:28   of the apps I've worked on.

01:01:29   So I have this row hanging in the wall of my office,

01:01:32   this row of app icons that I've worked on.

01:01:35   It's kind of like a trophy collection

01:01:36   for the things I've done in my life.

01:01:38   Because when you make apps,

01:01:40   there aren't a lot of physical artifacts like that.

01:01:43   And it's great because the 5x5 print is just $12.

01:01:48   So it's really no barrier to entry here if you want to get a couple of app icons made.

01:01:52   You know, yeah, spend $12.

01:01:53   It's no big deal.

01:01:55   Anyway, Fracture puts everything you need to get your photo on the wall right in the

01:02:00   box.

01:02:01   They give you your own little picture anchor thing or if you get the desk version, it has

01:02:06   a little stand already built into the frame.

01:02:09   Prices start at just $12 for the prints.

01:02:11   They're very reasonable.

01:02:12   The big ones are really reasonable.

01:02:15   And every fracture is handmade and checked for quality by a human being in their small

01:02:19   team in Gainesville, Florida.

01:02:21   It is the thinnest, lightest, and most elegant way to display your favorite photos.

01:02:26   And the best thing is you can even get 15% off with a coupon code ATP.

01:02:32   So please use coupon code ATP to get 15% off.

01:02:36   Go to fractureme.com.

01:02:38   That's F-R-A-C-T-U-R-E me dot com.

01:02:41   Thanks a lot to Fracture for sponsoring our show once again.

01:02:45   I definitely recommend them, they are great.

01:02:46   I'm looking at, let's see, five of them right now

01:02:49   in my field of view.

01:02:50   - Yeah, they are really good and I recommend them as well.

01:02:53   Let me ask another obvious question,

01:02:56   then I'll get into one or two that are less obvious.

01:02:59   What took so damn long?

01:03:00   - I started writing this in October 2012

01:03:03   and that was when I still own Instapaper in the magazine.

01:03:07   That's when I had the idea for the audio processing stuff

01:03:09   and so I wanted to make a little prototype

01:03:10   to see if it was even possible.

01:03:12   So I did, and it was, and it was great.

01:03:14   But I didn't really have time

01:03:15   to make a full podcast app around it at that point.

01:03:17   I had other projects.

01:03:19   Then, you know, eventually other projects went away,

01:03:23   and I had time, so I worked on them.

01:03:25   And so then, you know, by last fall,

01:03:27   when I announced Overcast at XOXO last September,

01:03:31   so almost a year ago, that's when I announced it,

01:03:34   by that point I'd been working on it full-time

01:03:36   for, I don't know, four or five months,

01:03:38   something like that.

01:03:39   I really thought it was almost done.

01:03:42   Because as a typical programmer, I was like,

01:03:44   oh well, you know, it works for me for the most part.

01:03:48   It works on my phone, it's fine.

01:03:50   So yeah, I should have it out in what, two months?

01:03:53   That turned out not to be the case.

01:03:55   It just, podcast apps are so incredibly complicated.

01:04:01   And I didn't quite fully appreciate that at the time.

01:04:05   You know, I didn't, I knew about stuff like,

01:04:08   know, feeds, you know, having to manage weird feeds and everything else, like, you know,

01:04:12   edge cases there. But just the interface, there's so many screens in a podcast app,

01:04:18   it's crazy. Like I have rinsed up about this briefly a couple weeks ago in the after

01:04:22   show, so I'm, you know, it's just so, there's so much that goes into a podcast app. And

01:04:28   the, I now have 545 emails in my inbox, at least half of which are asking for features

01:04:37   that I haven't even done yet. It's a very demanding market as I'm learning today.

01:04:44   I thought I was launching with a lot for a 1.0. Turns out, for the most part, yeah, people

01:04:52   are fine with it, but there's a lot of people out there who are really demanding more, a

01:04:58   lot more, even from day one. And so I didn't want one point. I anticipated some of this,

01:05:05   So I didn't want to disappoint people a lot on day one.

01:05:10   So I didn't want to leave like massive gaping holes.

01:05:12   Now streaming is a big one and you know video is a medium sized hole.

01:05:17   I don't ever plan to support video because it's kind of a different medium and it demands

01:05:21   different things.

01:05:22   I wouldn't be able to use my audio engine and any of the effects and everything else.

01:05:26   I could use the compressor but I couldn't use smart speaks it would be weird.

01:05:30   There's a bunch of stuff that I couldn't,

01:05:33   that it would be harder to do with video.

01:05:35   And the whole interface, you have to accommodate video

01:05:37   and then you have the question of like,

01:05:39   all right, well, are you allowed to mix video into playlists?

01:05:42   And if so, what do you do?

01:05:43   Like, do you start playing video podcast episodes

01:05:46   right in the middle of audio podcast episodes?

01:05:48   What if you're in the car?

01:05:49   Like, there's all sorts of like weird things

01:05:50   with supporting video that I don't think are worth it

01:05:54   because I don't think video podcasts

01:05:56   are that big of a requirement.

01:05:58   I would venture a guess,

01:05:59   the majority of podcast listeners don't listen

01:06:01   to any video podcasts.

01:06:03   It's just such a different medium.

01:06:05   Video, I think, was, for me at least,

01:06:07   it was an easy decision not to support.

01:06:09   You don't, you just don't do it at the same time

01:06:12   as audio podcasts.

01:06:13   It's a different demand, and I think also most videos

01:06:16   move to YouTube these days.

01:06:18   - That's what I was gonna say, like,

01:06:19   are video podcasts even a thing anymore?

01:06:21   Hasn't YouTube channels more or less replaced them?

01:06:23   But I mean, I suppose they're still out there,

01:06:24   but like, that definitely sounds,

01:06:26   if I was into video podcasts, I would want an app

01:06:29   that was built around video podcasts,

01:06:30   'cause they're different enough.

01:06:32   And I suppose you could make one super app

01:06:34   that does both audio and video, but that's a tall order.

01:06:37   Like, I would be perfectly happy getting separate

01:06:39   applications for video and audio podcasts.

01:06:42   I mean, you know, and people use the YouTube application

01:06:45   if they're watching YouTube channels and stuff.

01:06:47   - Exactly.

01:06:48   So that's why no video, no streaming was harder

01:06:51   to take, basically.

01:06:53   The reason I don't have streaming is not because

01:06:57   it's some oversight that I just forgot to add it,

01:06:59   like as some emailers have assumed.

01:07:02   That's not why.

01:07:04   The reason I don't have streaming is that

01:07:06   all my audio processing stuff is done using raw,

01:07:09   low-level core audio APIs that don't inherently

01:07:12   automatically provide streaming support.

01:07:14   The other apps use AVPlayer.

01:07:17   And AVPlayer, the AVPlayer framework is higher level.

01:07:20   The downside of AVPlayer is that it removes

01:07:22   a lot of the control that you have.

01:07:24   There is a way to do voice boost.

01:07:27   Not quite as well, but you can do it.

01:07:30   It's a little more CPU intensive if you do it that way,

01:07:31   but it is possible to do voice boost with AVPlayer.

01:07:35   To the best of my knowledge,

01:07:36   it is not possible to do smart speed.

01:07:38   And I've thought about lots of, over the last few years,

01:07:41   I've thought about lots of different ways

01:07:42   to maybe attempt to hack smart speed into AVPlayer.

01:07:46   And I just could not come up with anything

01:07:48   that was remotely doable and reasonable

01:07:52   and not like a ridiculous, horrible pile

01:07:55   of terrible, fragile hacks that would break immediately.

01:07:58   So in order to make Smart Speed, I had to do raw core audio.

01:08:03   And I thought Smart Speed was a good enough feature

01:08:07   to make that worth adding months of development time

01:08:10   and making me have to re-implement things

01:08:13   that everyone else gets for free with AVPlayer,

01:08:15   including streaming.

01:08:16   The other thing is, when the other apps that do streaming,

01:08:19   they have to do it in a limited way

01:08:22   because AVPlayer, through reasons that I believe,

01:08:26   I tried asking Apple about this at WWDC

01:08:28   in the labs this year, no one really knew for sure,

01:08:31   or at least told me, but it seemed like the reason why

01:08:33   is because of HTTP live streaming and its DRM,

01:08:36   and its expectations thereof.

01:08:38   You can't save a stream, and you can't turn a download

01:08:42   into a stream.

01:08:44   You have to either stream something or download it.

01:08:46   You can't do both.

01:08:47   You can't just start playing a partial download,

01:08:50   and you definitely can't convert a stream into a download.

01:08:53   Like, you know, just stream it and just save it,

01:08:55   just save what you're streaming until it's done

01:08:57   and then save that as a file.

01:08:58   No, you can't do that.

01:08:59   What I'm going to do when I do add streaming,

01:09:03   which is, I'm gonna begin work on that shortly

01:09:05   once 1.0 stuff settles down,

01:09:07   I'm going to do it so that you can just start playing

01:09:09   a progressive download.

01:09:11   Doing that well in the background download system

01:09:14   requires new capabilities that are in iOS 8 only.

01:09:18   So it made sense to wait until iOS 8 came out

01:09:22   and do it then.

01:09:23   And so that's what I'm gonna do with streaming.

01:09:25   - So let me ask a genuine question.

01:09:27   Why is streaming such a big deal,

01:09:30   not from a development side, but from a user side?

01:09:32   Like there have been times I've wanted to listen to a show

01:09:35   and it hasn't been downloaded

01:09:36   and so I just wait for it to download.

01:09:39   Like what am I missing that makes this such a big deal

01:09:42   to so many people?

01:09:44   - Most of the benefit of streaming I think has been removed

01:09:46   with background downloading.

01:09:48   in iOS 8 or 7.

01:09:50   I really think that for the most part,

01:09:53   most people are going to have things downloaded

01:09:56   when they're at home or the office on WiFi

01:10:00   and they will never even notice it downloading

01:10:01   and then they'll go out and start listening

01:10:04   and they'll just listen to whatever they have downloaded.

01:10:05   But streaming is really useful

01:10:07   like when you're adding a podcast

01:10:09   and you wanna start playing it immediately.

01:10:12   And so for that instance,

01:10:14   like oh, I just discovered this episode or this show,

01:10:17   Let me start playing this right now.

01:10:19   - You can't, not you personally,

01:10:21   but you can't wait like literally 60 seconds.

01:10:23   Also the chat room is saying storage space.

01:10:25   I mean, I shouldn't be arguing with anyone.

01:10:28   I just, I didn't realize that people were running

01:10:30   within like 50 or 100 megs of the limits of their device.

01:10:35   I'm surprised to hear that.

01:10:36   - You gotta live gigs free.

01:10:38   I just tried to upgrade my iPhone to 712,

01:10:41   which I'd forgotten that I hadn't upgraded it to that.

01:10:43   And every time I try to upgrade iOS,

01:10:46   and that thing, it tells me I don't have enough space.

01:10:47   Only 1.4 gigs are available, so I gotta go delete stuff.

01:10:50   And you need a lot of room just for even small OS upgrades.

01:10:55   - Also, there's the issue of download speed also.

01:11:00   - Yeah, which is what the chat room is now telling me,

01:11:02   is that I'm spoiled by LTE/decent wifi.

01:11:06   - Well, and also, some podcasts are hosted

01:11:08   on servers or CDNs that themselves don't send the files

01:11:10   very quickly or can't send the files very quickly.

01:11:12   - Yeah, that is a big problem.

01:11:14   So my experience, obviously I don't have an iPhone,

01:11:16   but like with my iPod touches,

01:11:17   with the background downloads,

01:11:19   every time I pick up my iPod touch

01:11:22   and go to Overcast over the past month or so

01:11:24   I've been using it, everything's already downloaded

01:11:27   'cause it's just been sitting there in my wifi all day.

01:11:29   But on the occasion when I said,

01:11:32   oh, actually I wanna get like you just said,

01:11:33   oh, I just wanna get this episode,

01:11:35   I do the thing you just said, Casey,

01:11:36   and I go to download it and I look, it's like 1%, 2%.

01:11:40   And you're like, oh, no, it's gonna be,

01:11:42   and for me, I can't say,

01:11:44   well, I'm gonna get in the car and drive away anyway

01:11:45   Because as soon as I leave my house, the download stops.

01:11:48   It's an iPod Touch.

01:11:49   So I think there are two use cases.

01:11:52   One are the people-- when you say streaming,

01:11:54   you think of people just wandering around,

01:11:57   constantly pulling audio over cell data that

01:12:00   goes right into their ears.

01:12:01   But I think the other use case is I

01:12:05   don't want to have to wait for the download to finish,

01:12:07   and I want to be able to start listening as fast as I can.

01:12:10   Because presumably, the service can send audio data fast

01:12:13   enough that you can listen to it in more or less real time.

01:12:16   So that's what I would like out of the streaming.

01:12:19   That's like, I remember being frustrated

01:12:21   when Apple first added podcast to iTunes.

01:12:23   I got all excited when it's a podcast, got a subscription,

01:12:26   and then I clicked on an episode

01:12:28   and saw the downloading bar.

01:12:29   I'm like, why isn't it playing?

01:12:31   It's not gonna play until it finishes downloading.

01:12:33   This is ridiculous, you know,

01:12:34   because you're used to like everything, you know,

01:12:35   web browsers, if you go an audio file, a video,

01:12:37   like of course it starts playing

01:12:38   before the whole thing downloads.

01:12:40   It's not, you know, but podcasts weren't like that,

01:12:42   probably for framework reasons.

01:12:43   And by the way, a lot of people in the chat room

01:12:45   have mentioned lots of other podcast apps

01:12:47   that already do this.

01:12:48   I think what Marco was saying is that

01:12:50   the frameworks Apple provides

01:12:52   don't give you a convenient way to do that.

01:12:54   So if you want to do that, it is entirely possible

01:12:56   you have to do it yourself though.

01:12:58   And what Marco's saying is he wants to wait until iOS 8

01:13:00   because there are frameworks that make that easier.

01:13:02   All these other people with podcast apps,

01:13:04   like Marco said, got a big head start on him.

01:13:06   They've already implemented their own ways

01:13:07   to do all these things.

01:13:08   Essentially, I think what it comes down to for streaming is

01:13:11   You gotta pick what's gonna make it into 1.0,

01:13:13   and that one didn't make the cut.

01:13:14   I mean, if you had added streaming,

01:13:15   you'd be waiting many more months before the thing came out.

01:13:18   - Right, and specifically, what they've added in 8,

01:13:22   you know, this is no big top secret,

01:13:24   what they've added in 8 is the ability

01:13:26   to basically use streaming on the same download

01:13:31   that can become a background download if it has to.

01:13:34   In 7, if you wanted to do a background download

01:13:37   when the app wasn't running or whatever,

01:13:39   you kick off the request and it says,

01:13:42   all right, and then it can tell you progress updates,

01:13:45   but you can't get to the data until it's done.

01:13:49   So there's no way to stream that.

01:13:51   So if you were offering streaming before,

01:13:53   you had to offer streaming in a way that,

01:13:55   oh, and you also can't convert any other downloads

01:13:58   into a background download.

01:13:59   So you'd have to offer streaming in a way

01:14:01   that if you were streaming hit pause

01:14:04   and the app got terminated or got backgrounded,

01:14:07   that download would get canceled and fail

01:14:08   and you have to start over again with a background download

01:14:10   if you wanted that file.

01:14:11   Where what they've added in 8 is the ability to convert

01:14:16   a streaming data download into a background download.

01:14:20   So that that use case can continue and can do that well.

01:14:24   So that's what I'm gonna use.

01:14:26   - All right, what about the dev process

01:14:31   surprised you the most?

01:14:32   Was it just that there are that many screens

01:14:34   or was there something else?

01:14:36   - I was most surprised that all this audio stuff

01:14:38   worked on an iPhone.

01:14:40   - That's actually a pretty good answer.

01:14:41   - Really, because it was like, oh my god,

01:14:44   the amount of processing I do on every sample

01:14:49   that goes through 44,100 times per second,

01:14:54   or if you're a stereo podcast, 88,200 times per second,

01:14:58   the amount of just processing and math,

01:15:02   I look at almost every sample in some way

01:15:04   when smart speed is on, at least.

01:15:07   and even when smart speed is off,

01:15:09   to render the little peak meters animation,

01:15:11   the little EQ animation, I'm looking at every sample.

01:15:15   I'm doing an FFT on every sample.

01:15:19   That's crazy talk.

01:15:20   Thinking about that, just logically,

01:15:24   that shouldn't be so fast on an iPhone.

01:15:27   It runs at 30 frames a second on an iPhone 4

01:15:30   and without maxing out the CPU.

01:15:33   And to answer people in the chat

01:15:34   talking about battery impact and everything,

01:15:37   I was concerned about battery impact.

01:15:39   I've been using this for over a year now

01:15:42   and the battery impact is fine.

01:15:44   It's fine, like it's not.

01:15:47   What hits the battery the hardest

01:15:48   is playing shows faster than 1X.

01:15:52   And so to some degree, Smart Speed does that

01:15:54   because the reason why that's so hard

01:15:56   is that you're then asking the hardware

01:15:58   to process more than 44,100 frames per second.

01:16:02   You're asking it to, you know, if you're playing at 1.5X,

01:16:05   it's gonna process 60 whatever frames per second.

01:16:08   So it's going to, when you're playing faster than 1X,

01:16:12   it still has to decode the MP3, read all that data,

01:16:15   process all that data, so it's doing all that work

01:16:20   more per second.

01:16:21   That is what makes a big battery impact.

01:16:24   If you look at the CPU usage meter

01:16:26   as you're running one of these things,

01:16:28   but that's true of all the podcast apps, including Apple's.

01:16:31   They all have that problem,

01:16:32   because that's just the reality of playing data quickly

01:16:35   is that you have to then read and process it more quickly,

01:16:37   so you have to do more of it and it costs CPU time.

01:16:39   - You're not doing the equalizer animation

01:16:42   when the screen's off, right?

01:16:43   - Correct, yeah.

01:16:44   If there isn't one on screen, or the screen is off,

01:16:47   or the app is not foreground, then it does not do anything.

01:16:52   - All right, what's your favorite feature,

01:16:54   which may or may not be the one you're most proud of?

01:16:58   - I'm definitely, okay.

01:17:01   I think my favorite one is playlist reordering.

01:17:04   The way playlists are implemented is crazy

01:17:08   because I really wanted this particular feature,

01:17:12   which is I didn't want to have to distinguish

01:17:13   between a manually organized playlist and a smart playlist,

01:17:17   the way most iTunes does it that way.

01:17:20   I wanted all playlists to be smart and manually organizable.

01:17:24   So that's how I did in Overcast.

01:17:27   So you can set your criteria for your playlist

01:17:30   and what should be included,

01:17:32   But you can always just drag it around and reorder it.

01:17:35   And then when new things come in,

01:17:37   your order is preserved and it just tries to,

01:17:40   it tries to obey the filters in a reasonable way

01:17:43   and the priority settings and everything

01:17:45   to try to fit a new episode in where it should probably go

01:17:49   given your reordering.

01:17:50   - Yeah, you know, when you first told us about Overcast,

01:17:55   that was quite a while ago and you were talking about

01:17:57   things like smart speed and things that you

01:17:59   were trying to include in it.

01:18:01   And you had told us about your playlist ideas.

01:18:05   Possibly it was underscore and I down in South Carolina.

01:18:08   But regardless of when, I never used playlists ever in any podcast app I've ever used.

01:18:14   And so when you gave us the beta, I tried to do a playlist for things that just Aaron

01:18:20   and I listened to, which is basically just IRL talk, and then everything else.

01:18:25   And every time I started fiddling with the playlist settings

01:18:30   or perhaps reordering the playlist

01:18:32   like you were describing,

01:18:33   it was very quickly apparent to me

01:18:35   that this was something that I've always wanted in my life,

01:18:37   I just didn't know it yet.

01:18:39   And it's really, really well done.

01:18:41   And I am not surprised that you count that

01:18:43   as your favorite feature.

01:18:44   I think my favorite is probably Smart Speed

01:18:46   because I think it is so transparent,

01:18:48   but the playlists are a close second.

01:18:51   - Yeah, I mean, Smart Speed, I think,

01:18:53   is definitely what I'm most proud of.

01:18:54   because smart speed, it's a hard thing to do.

01:18:57   Like one of my formulas for my own happiness

01:19:01   and intellectual happiness as well as success

01:19:03   in the app world is it's nice if you do like one hard thing

01:19:08   and a lot of easy things in an app.

01:19:11   That's a really good balance to have

01:19:12   because if you do like one really hard thing,

01:19:14   that makes it harder for competitors to do

01:19:16   and to copy you feature by feature.

01:19:18   But then all the easy things help it be easier

01:19:21   to maintain for you.

01:19:23   And usually, a lot of things that are very hard to do

01:19:27   aren't really worth it.

01:19:27   Like they don't bring,

01:19:29   it isn't a must-have feature for customers.

01:19:31   So you gotta be careful what you pick.

01:19:33   But I talked in the last after show about how

01:19:36   I worked so hard on the Kindle feature for Instapaper.

01:19:40   And in reality, that probably wasn't worth

01:19:41   all the effort I put into it.

01:19:43   'Cause it was hard.

01:19:44   It's easier these days with some of their tools,

01:19:46   but it was hard.

01:19:47   Anyway, Smart Speed is a very hard feature

01:19:50   to implement well.

01:19:51   You can do it badly in a few easier ways,

01:19:55   but I wanted to do it right and I wanted to do it well,

01:19:57   and doing it well is not easy.

01:20:00   That said, there isn't anything stopping

01:20:02   the other podcast app developers from doing it.

01:20:04   So, you know, like just time and dedication

01:20:07   and possibly having to rewrite a lot of code,

01:20:10   you know, they could do it.

01:20:11   It's not like this is gonna be exclusive to me forever.

01:20:14   Like, I'll be lucky if it's exclusive to me for six months,

01:20:17   you know, but anyway.

01:20:20   And yes, I know RSS Radio already does it

01:20:22   to prevent any feedback like that.

01:20:24   So it is even exclusive today.

01:20:27   But yeah, so anyway, I'm very proud of that.

01:20:31   I'm also very, very proud that my EQ meters

01:20:34   animation thing actually works.

01:20:36   Because the whole reason I did that

01:20:38   is because Apple in their music app and iOS 7,

01:20:43   like the problem I was trying to solve with that

01:20:46   was how do you indicate in the list of episodes,

01:20:49   how do you indicate which one is currently playing?

01:20:52   And the way Apple handles this

01:20:55   is they have a little pink animation in the music screen

01:20:59   that is like a tiny version of this

01:21:02   in like a little circle on the side or something like that.

01:21:04   It's a tiny version of a peak meter's view.

01:21:06   What drives me crazy about it is that it's fake.

01:21:09   It's just a canned animation of some bars moving up and down

01:21:13   it is not actually reflecting the music that it's playing.

01:21:17   So me being a smart ass and being arrogant,

01:21:20   I thought, you know what,

01:21:21   I bet I could do that for real in my app.

01:21:24   And so I tried and I'm like, you know,

01:21:26   and at first I was like,

01:21:27   I'm probably gonna have to use OpenGL

01:21:28   and to make it really fast and everything.

01:21:30   It's kind of hard to use OpenGL

01:21:32   on a translucent blended layer

01:21:34   that's gonna be blended into the rest of UI kit.

01:21:35   Like that's not easy.

01:21:37   And I really didn't want the overhead of doing GL

01:21:39   'cause I didn't wanna be spending two weeks

01:21:40   on this one little feature.

01:21:42   - What made you think you knew

01:21:43   we're gonna need to use OpenGL for it?

01:21:45   just to get the animation fast enough,

01:21:48   or for battery reasons.

01:21:51   Maybe I could do it with QuickDraw or whatever,

01:21:53   but maybe it would take too long.

01:21:55   QuickDraw, right.

01:21:57   You mean the actual drawing process, though.

01:21:59   What you're worried about is not so much the core animation

01:22:02   will be run through OpenGL, but you

01:22:03   didn't want to have to draw the little lines with core graphics

01:22:07   thinking that you wouldn't be able to get 60 frames per second

01:22:09   out of that.

01:22:10   I was worried about that, and I was also

01:22:12   worried about the FFT that's running in the background

01:22:15   processing all the sound, like what is that going to do to battery life?

01:22:18   What do you think the GPU is sitting there waiting to run your DSP functions on it, right?

01:22:23   I've thought about that, but honestly the CPU versions of it work so incredibly well.

01:22:28   For the scale I'm running them on, I don't think the GPU will make a big difference.

01:22:33   The libraries, they have those, what are they called, the ARM extensions, the ARM SIMD extensions,

01:22:38   neon or something? Yeah, yeah. So like presumably Apple's libraries are using that behind the scenes

01:22:43   on the CPU's that support it.

01:22:44   So it's kind of like you get your own, you know,

01:22:46   it's like Ultivec all over again.

01:22:48   - Exactly, yeah.

01:22:49   So I made very heavy use of the accelerate framework,

01:22:52   the VDSB functions, which all do,

01:22:54   as far as I know they all use neon whenever they can.

01:22:58   And so, yeah, so I made this visualizer with that.

01:23:03   I realized how quickly it ran,

01:23:05   and it was originally only gonna be on,

01:23:09   like just overlaying the artwork on the table cells.

01:23:11   That's all I needed it for.

01:23:13   And then I realized, oh, this is really cool.

01:23:16   And I added it to the main playing screen,

01:23:20   first of all because it looked cool

01:23:21   and it seemed like a waste to have it

01:23:22   on the navigation screens and not on the now playing screen.

01:23:25   And also because I thought it might help,

01:23:29   if you're listening to a podcast on your phone,

01:23:32   but you're not doing something else,

01:23:34   if you're just sitting on the couch looking at your phone,

01:23:36   like nothing else is going on,

01:23:38   or you're sitting commuting on a plane

01:23:40   or something like that,

01:23:41   You wanted to signal to other people that you're listening?

01:23:44   - No, I wanted, well first of all,

01:23:45   it is nice that it has that feature

01:23:47   where like in a lot of other apps,

01:23:49   the only way you can tell that they're playing

01:23:51   is if the time is moving, basically.

01:23:54   And you know, or the symbol changes on the play button,

01:23:57   but like it's, these are pretty subtle signals.

01:23:59   With Overcast, it's very clear when it's playing something.

01:24:02   But also, I wanted to give people

01:24:04   who were just sitting there with their phone

01:24:05   who had nothing else to do, something to look at.

01:24:08   Like just something to like keep you

01:24:10   visually entertained a little bit,

01:24:11   so that maybe, just maybe, you'll,

01:24:13   for a few seconds more, not switch over to Twitter

01:24:16   and stop paying attention to what you're listening to.

01:24:19   So it was kind of like a political statement as well,

01:24:21   like, let me give you some reason to stay in this app

01:24:23   and something to just lock your visual attention

01:24:26   so that you can pay attention to the podcast,

01:24:28   rather than switching to something else

01:24:29   and zoning it out.

01:24:31   - So why did you mirror all the bars, Johnny Ive?

01:24:34   - It looked better.

01:24:35   Exactly, that's it, it looked better.

01:24:37   Simple as that.

01:24:39   There were a number of good reasons why the number of bars

01:24:43   had to be, it's 19, 18, 19, yeah.

01:24:48   There were a number of reasons the number had to be

01:24:51   no greater than 19, like the FFT window size.

01:24:54   It was like, it would have been really difficult

01:24:56   and much more complicated and much more CPU intensive

01:25:00   if I had more FFT buckets.

01:25:02   So I couldn't do more bars than that.

01:25:04   And so not only does it look better to have the shape

01:25:07   be symmetrical that the bars form,

01:25:10   rather than just being like a big slope

01:25:12   and you fall off the edge of the screen.

01:25:13   So not only is the shape nice and symmetrical,

01:25:16   but also the bars can be nice and thin

01:25:18   relative to like, you know, not being like these,

01:25:20   these, you know, big fat ones.

01:25:21   - You can always cheat and interpolate

01:25:23   between the bars, right?

01:25:24   - Yeah, but that's stupid.

01:25:25   Yeah, I couldn't do that.

01:25:27   - All right, so I've got two more questions

01:25:30   and then I'd genuinely like to hear John

01:25:32   destroy your user interface.

01:25:35   All right, firstly, what do you hope people appreciate?

01:25:39   And perhaps you feel like you've already answered that,

01:25:41   but an example of something we haven't really talked

01:25:43   about yet is the link to other ads.

01:25:46   - Do we, wait, before we made this entirely the me show,

01:25:49   do we have any other topics we wanted to get to?

01:25:51   'Cause I don't wanna like take over our show.

01:25:54   - Too late, we do have topics,

01:25:55   but they'll keep till next week.

01:25:57   - Yeah, and you can blame me

01:25:58   'cause I keep asking you questions, so it's my fault.

01:26:02   So yeah, so what do you hope people appreciate?

01:26:04   And what I was starting to say was,

01:26:05   an example of this that I've seen a lot of positive feedback

01:26:09   for from whatever was the linking

01:26:12   to the other independent podcast apps.

01:26:14   And I've seen a lot of comments fly by

01:26:16   about how that was a classy thing to do.

01:26:18   Maybe that's your answer, maybe it isn't,

01:26:19   but what do you hope users of the app really appreciate?

01:26:23   - To answer the question directly,

01:26:25   what I hope people really appreciate is the smart speed

01:26:27   and the voice boost, because these are the things

01:26:29   that just were the hardest.

01:26:31   They took the longest.

01:26:32   And for the time being, they set me apart.

01:26:35   These are features that they don't immediately scream,

01:26:39   oh my God, must have to a lot of people.

01:26:42   Some people, sure, but not to probably

01:26:44   the majority of people.

01:26:46   But once you try them, they're incredibly valuable

01:26:49   if you care about those things.

01:26:51   And some people don't.

01:26:52   Some people never play podcasts faster than 1X.

01:26:53   They don't want to, they don't see the point,

01:26:55   or they don't like the sound or whatever.

01:26:57   A lot of people couldn't possibly care

01:27:01   about the volume normalization and everything.

01:27:02   That's fine.

01:27:03   You know, you don't need everyone to love you

01:27:05   and you don't need everyone to appreciate

01:27:07   the amount of work you put into anything.

01:27:08   That's, you know, the amount of work you put into something

01:27:10   does not correlate to its value.

01:27:12   So that's fine.

01:27:14   But I hope people appreciate that stuff who care, you know?

01:27:19   And so that's the answer to that.

01:27:21   Now to get to the question that you kind of asked

01:27:23   about why I did the link to other apps in the,

01:27:27   I basically have a section in settings

01:27:29   that links to all my competitors for, well not all of them,

01:27:32   linked to some of my competitors,

01:27:34   probably the biggest ones.

01:27:35   The main reason why I did that,

01:27:38   it was kind of a what the hell moment.

01:27:41   It was like, you know what, I wanna do this thing,

01:27:44   and I think it would be pretty cool.

01:27:46   I was afraid though, it was a risk,

01:27:48   I saw it as a risk in two reasons,

01:27:51   one small and one big.

01:27:53   The small risk was that Apple would reject it

01:27:55   because they have pretty strict rules

01:27:57   about whether you can display and link to other apps

01:28:01   within your app and under what conditions.

01:28:03   And if you read the rules, it's kinda like,

01:28:05   well, this might qualify, this might not.

01:28:07   It was a little bit vague,

01:28:09   and I wasn't sure if they would approve it.

01:28:10   Turns out they don't care, not yet at least.

01:28:12   So we'll see future updates,

01:28:14   we'll see if this feature lasts.

01:28:16   So far they don't care.

01:28:17   The second concern, and that's a minor concern,

01:28:21   'cause if they say I can't do it, fine,

01:28:23   I take it out and I resubmit, no big deal.

01:28:25   The second reason, though, was I was afraid that it would,

01:28:30   My goal with doing this, with linking to my competitors,

01:28:35   was really well-meaning.

01:28:36   It was really to just like, you know,

01:28:39   some of these people are my friends.

01:28:41   This is a small business.

01:28:43   These are all indie developers like me.

01:28:45   We're not talking about,

01:28:46   like I'm not linking to like major corporations

01:28:48   with VC funding who were trying to crush everybody

01:28:51   and put everyone out of business.

01:28:52   Like that's not this business.

01:28:54   This is a business of small independent developers

01:28:57   who are very similar to me.

01:28:59   and I wanted to kind of show solidarity with it almost,

01:29:03   just kind of be friendly with the business

01:29:06   rather than being hostile towards my competitors.

01:29:09   And again, some of these people are friends of mine,

01:29:12   which made it even easier to do it.

01:29:14   I was afraid though that it might be perceived as

01:29:17   me looking down on these people,

01:29:19   like, oh, you don't like my app?

01:29:21   Well fine, just take one of these,

01:29:23   'cause I'm like the arrogant (beep)

01:29:24   who everyone hates.

01:29:26   I was worried about that perception happening,

01:29:30   but it didn't, and in fact, not only did it not happen,

01:29:33   I didn't see a single person who seemed to think that.

01:29:37   Having that feature in the app,

01:29:39   I think has gotten me more tweets, more comments,

01:29:43   more emails that are all positive

01:29:46   than anything else in the app,

01:29:47   including all my features, Smart Speed,

01:29:49   people like that more than everything else

01:29:54   I've done in the app,

01:29:55   according to what people are telling me about and what people are posting on Twitter. People

01:29:58   love it. They love it.

01:30:00   I was just going to bring up the one jerk who I saw who was yelling at you about that

01:30:04   on Hacker News. I guess you didn't see that one. But anyway, rest assured that the internet—

01:30:08   I haven't read through all of Hacker News yet.

01:30:09   Yeah, read the Hacker News story. There's one guy who gets all upset that you linked

01:30:13   to your comp— like, it takes a special kind of determination to dislike somebody. To figure

01:30:20   out a way that someone linking to his competitors in a respectful way from his application is

01:30:27   somehow a, you know, shows that he is a terrible person. But this person took their best shot

01:30:35   at it.

01:30:36   Yep. So my last question, which I think you just answered seconds ago, was what was your

01:30:40   most surprising response? And it sounds like the linking to other people's apps was that.

01:30:45   - Yeah, definitely.

01:30:46   That and a combination of both of how many people

01:30:50   immediately saw that and immediately cared about it

01:30:53   in a positive way, so that's number one.

01:30:55   Number two, I was really just overall

01:30:57   very pleasantly surprised at how positive

01:30:59   the reactions generally were.

01:31:01   I didn't know, like until this launch,

01:31:04   I had no idea whether anyone would care

01:31:07   about the new cool stuff I did

01:31:11   because I have the, 'cause I don't support

01:31:13   the iPad, the Mac, Android, I barely support the web.

01:31:18   I don't have streaming, I don't support videos.

01:31:22   Like there's all these limitations in Overcast.

01:31:25   Those are pretty big omissions to a lot of people.

01:31:28   But they're not very big to me.

01:31:29   I only listen on my iPhone.

01:31:31   I didn't really feel myself that all these other platform

01:31:36   clients were important to have on day one.

01:31:39   Not as important as launching a year earlier, say.

01:31:42   So there was that.

01:31:44   I also, you know, I never watch video podcasts

01:31:48   because I'm always listening to podcasts in contexts

01:31:51   where I can't look at a screen.

01:31:52   So I don't use video podcasts.

01:31:54   And streaming, as I said, I'm going to ad streaming.

01:31:57   I didn't make it into version one

01:31:58   because it wasn't so important to hold back the release

01:32:02   for another three months while I did it

01:32:04   and wait for iOS 8.

01:32:05   Like that, so that's why.

01:32:08   Overall though, I didn't know how many people out there

01:32:11   thought like me on this.

01:32:13   Like certain features I don't even have, big and small,

01:32:17   I didn't know how many people that would anger

01:32:20   and offend and turn off and who would just say,

01:32:23   "Useless, one star."

01:32:25   And it turns out there are a few, of course.

01:32:27   Yeah, there's always gonna be a few,

01:32:29   but it's a much smaller percentage of people

01:32:30   than I expected.

01:32:31   - Yeah, that's awesome.

01:32:34   And I should, to back up just a brief moment,

01:32:36   I should point out that I was curious

01:32:39   how you were going to handle ordering the podcast apps

01:32:44   that you were linking to,

01:32:45   because one could perhaps unreasonably construe that,

01:32:49   oh, whatever you put at the top

01:32:51   is clearly your most favorite of all your competitors.

01:32:54   And I don't remember if it was that you told me

01:32:57   or I noticed, but you found a way around that,

01:33:00   which I thought was very clever.

01:33:01   - I just randomized the list every time you open the screen.

01:33:04   - Right, which I thought that was a very clever idea

01:33:08   a really good touch as well.

01:33:09   - Yeah, I do the same thing for the same reason

01:33:12   with the directory categories.

01:33:13   Like the category list itself is ordered intentionally,

01:33:16   but the podcasts in each category are randomly ordered

01:33:20   every time you load it.

01:33:21   - Nice.

01:33:22   - Same reason, because it's like, you know,

01:33:23   I don't wanna like, I don't wanna have to rank these

01:33:26   like, you know, one through six in this category.

01:33:28   Like I don't care that much.

01:33:30   I don't want it to be like that.

01:33:31   - You can always do alphabetical for the directory, I mean.

01:33:34   - Yeah, but then you have like AAA Tech Show,

01:33:38   (laughs)

01:33:39   That you have the yellow pages problem.

01:33:41   - People are gonna be gaming your particular

01:33:43   iPod or podcast application.

01:33:45   - Well, but like your lessons to Casey.

01:33:47   It's like, you wanna add that security now or later?

01:33:49   (laughs)

01:33:50   It's better off to add it now.

01:33:52   (laughs)

01:33:53   So yeah, that's why I did that that way.

01:33:55   And also, by the way, another risk that I thought,

01:33:59   the only developer who was in that list

01:34:02   who knew about it beforehand was _DavidSmith,

01:34:04   'cause he's a friend of ours, he was on the beta.

01:34:07   So Pod Wrangler, by _DavidSmith,

01:34:09   that was the only app on there that knew about it beforehand.

01:34:12   I was also afraid, and this could still happen,

01:34:16   that one of those developers would get really pissed off

01:34:18   at me for including them on that list

01:34:19   and would demand removal.

01:34:20   If that happens, I can do it server side, it's no big deal.

01:34:24   But--

01:34:25   - Why would they demand removal?

01:34:27   - Maybe they really hate me as a competitor

01:34:30   and don't wanna be in my app at all, I don't know.

01:34:31   - They don't want you to link to their app?

01:34:34   - It was a risk.

01:34:36   I don't think it was a major risk, but I was worried,

01:34:39   like what if one of these competitors gets really upset

01:34:41   with me putting them in here?

01:34:42   - I think it's an insane thing to have to worry about,

01:34:45   but I concur with you, Marco, that that is something

01:34:47   I would worry about as well if I were you.

01:34:50   - I mean, keep in mind, I've been kind of overly sensitive

01:34:54   to what, of this thing I'm doing that I think

01:34:57   is well-meaning and harmless, what could people possibly

01:35:01   have a major problem with?

01:35:02   Because everything I've written on the internet

01:35:04   or said here has been attacked with that kind of standard

01:35:06   recently. So like, I have to be very careful.

01:35:08   Like, anyway, so just try to think like,

01:35:10   in what ways can this be totally misconstrued

01:35:14   that would cause problems for me or other people?

01:35:17   - So in the chat room, I thought of something

01:35:18   that I thought of when I first saw this.

01:35:20   It's like, if you had put things from Megacorps in there,

01:35:23   they'd be like, you can't use our trademark image

01:35:25   in your application.

01:35:27   - Technically maybe, however, it's their app icon,

01:35:31   which I'm loading from the app store,

01:35:32   using it to show their app page,

01:35:34   so that I'm not sure that would really hold water,

01:35:35   But if anybody complained, I would remove them.

01:35:38   It's no big deal.

01:35:38   Again, I said I could do it server side

01:35:40   and by commenting out one line in an array.

01:35:42   But still, I'd rather avoid conflicts

01:35:46   rather than invite them and then try to escape them.

01:35:49   But overall, based on the response

01:35:53   from that competitors list, I think it was the right move.

01:35:57   I'm totally blown away by how many people just love it.

01:36:02   - All right, so the chat room would like me to ask,

01:36:05   How do you handle the retired greats

01:36:07   and perhaps in general, could you give us just a quick blurb

01:36:11   on how you handled the directory overall?

01:36:14   - I whipped it together three days ago

01:36:17   and I intend to update it sometimes.

01:36:19   That's it.

01:36:22   - That's kind of what I thought.

01:36:22   I'm not sure what they're looking for you to say,

01:36:25   but at least we did.

01:36:26   - The directory part, like the categories though,

01:36:28   where it's like, oh, this is a tech podcast,

01:36:29   is that, how is that determined?

01:36:31   Is that metadata that's in the podcast?

01:36:33   is that you're just saying, oh, this is a tech podcast?

01:36:35   - These are all handpicked in my admin panel.

01:36:37   Like I can create categories

01:36:39   and put whatever I want into them.

01:36:40   - But so it's not exhaustive.

01:36:42   I haven't even browsed through it.

01:36:43   It's just like, here's a bunch of tech podcasts.

01:36:45   - Yeah, it's meant to be a starting point.

01:36:47   So there's, every category has in it

01:36:50   a subscribe to all button.

01:36:51   And the idea of adding this,

01:36:54   like one of my beta testers asked,

01:36:57   is this here because a lot of people need to do this?

01:37:00   Like, is this a common request?

01:37:02   Why is there a subscribe to all button

01:37:04   taking up a whole table row in the directory?

01:37:07   And the reason why it's there is not because anybody

01:37:09   has ever asked anybody for that feature as far as I know.

01:37:11   It's kind of an opportunistic, like,

01:37:13   well, let me give them a chance.

01:37:15   Maybe somebody will do this.

01:37:17   It's a way, in my opinion, it's a way to promote podcasts

01:37:21   and to give people a quick way,

01:37:23   if they have, if they subscribe to nothing,

01:37:25   if this is their first time ever using a podcast app

01:37:26   and they don't know what to subscribe to,

01:37:28   go pick a category, hit subscribe to all,

01:37:30   and you'll have enough content to last you

01:37:33   most of your time that you need to listen to something.

01:37:36   So that's basically it.

01:37:38   It was like a, these are like little mini collections,

01:37:40   like little mini curated collections of six to eight

01:37:44   podcasts in each category.

01:37:45   The idea was not to be exhaustive,

01:37:48   but to provide like a starter kit to people.

01:37:51   And this was, this is a hedge.

01:37:53   I don't know if the directory will prove to be

01:37:56   the most important part here,

01:37:58   in which case I should probably make it more robust.

01:38:01   Or if most people will use search and Twitter

01:38:04   to find their stuff.

01:38:06   So this is kind of a hedge where I have this,

01:38:08   I have like a medium strength Twitter feature,

01:38:12   a basic directory, and a good search.

01:38:15   And I think a good search you always have to have,

01:38:17   so that's out of the question, you have to have that.

01:38:19   But between the Twitter and the directory,

01:38:21   which one do I focus more time on in the future,

01:38:23   that will be determined by what people actually use.

01:38:27   Well, you can amp up your directory

01:38:29   as people start using the application,

01:38:31   because then you can do the stuff

01:38:32   that Apple does in its store.

01:38:33   I'm assuming you don't have access to Apple's metadata,

01:38:35   right?

01:38:36   Like, yeah, new and noteworthy, top rated or whatever,

01:38:39   but you can say it like popular.

01:38:41   Pretty soon you'll know what are the popular tech podcasts,

01:38:44   not just from your list, but in general.

01:38:46   If you are willing to categorize a large portion

01:38:49   or scrape categorizations from some other place,

01:38:51   a large portion of the podcast,

01:38:52   you will have data to know, oh, this one,

01:38:54   I mean, you could rank them by popularity,

01:38:56   like what is the most popular comedy podcast

01:38:59   and put that one up at the top.

01:39:00   - Well, the podcast put iTunes category metadata

01:39:04   in the feed usually,

01:39:06   'cause iTunes has this RSS spec

01:39:09   that almost every podcast feed adheres to

01:39:12   so they can be listed in iTunes.

01:39:13   And so many podcasts claim their own categories in the feed.

01:39:18   So if I wanted to do something like that, I could.

01:39:22   I'm not convinced of the value of that.

01:39:24   If you browse around the iTunes top list,

01:39:26   once you leave the editorial areas of iTunes,

01:39:30   like the editor's picks and stuff,

01:39:32   if you go to just the raw top things in this category list,

01:39:36   I don't know how useful that is.

01:39:37   I'm not sure, this is the thing, I don't know yet.

01:39:41   I don't know, do most people find their podcasts

01:39:43   by browsing the directory,

01:39:45   or by searching for something by name?

01:39:47   - The social one is probably the most reliable view,

01:39:49   even though it seems like the weirdest

01:39:50   and the weirdest thought,

01:39:51   "Oh, I gotta put my Twitter thing in," or whatever,

01:39:53   because recommendations from your friends

01:39:55   or people you follow, it's like the top list,

01:39:57   if you rank by popularity, just the same stuff always,

01:40:00   it's like this American life is gonna be the number one.

01:40:03   You know, spoiler, that's what's gonna be

01:40:05   the number one thing there, it's really super popular.

01:40:07   But who hasn't heard of This American Life?

01:40:09   And in some respects, it's like, well,

01:40:10   that's a good podcast, people, you know,

01:40:12   there's a high chance that just an average person

01:40:14   picked out of a hat will like this podcast

01:40:17   because lots of people like this.

01:40:18   But with podcasts, with a curated list,

01:40:22   then it's like, well, if your taste is like Marco's,

01:40:24   then you may like one of these picks

01:40:26   more than you would like this American Life

01:40:28   because it's quirky as your particular thing.

01:40:30   That's why I think the Social one is actually the best

01:40:32   because you have to find someone who has similar taste

01:40:36   to you and then follow their recommendations.

01:40:38   The Super Popular one only works if you are

01:40:42   not an individual but are rather the average human being.

01:40:46   Because then you're like, oh, yes, exactly.

01:40:48   I am the average human and I like the best things.

01:40:51   I find with podcasts, people are most excited when they find something that may not be particularly

01:40:56   pop—this show, for example—that may not be particularly popular in the grand scheme

01:40:59   of things on this American life type level, but exactly speaks to their particular interests

01:41:06   of like, "I like people who have one particular kind of model train, and they will be so excited

01:41:12   about a podcast just about that kind of model train, they will love that show more than

01:41:16   this American life."

01:41:17   - Yeah, and that's, I really think the social features

01:41:22   are much more useful than the directory for most people.

01:41:26   I think the role of a directory

01:41:28   is to provide a starting point, right?

01:41:30   And that's why I have my little starter kits

01:41:34   in each little categories, that's fine.

01:41:37   But once you get past give me a few things in X category,

01:41:42   the browsing experience is terrible,

01:41:43   the usefulness of the rankings is off.

01:41:46   That's why I think this whole social thing is the way to go.

01:41:49   I don't know.

01:41:52   - All right, Jon, do your worst.

01:41:55   - Thanks a lot to our three sponsors this week,

01:41:57   Need, Squarespace, and Fracture,

01:42:00   and we will see you next week.

01:42:02   (upbeat music)

01:42:04   ♪ Now the show is over ♪

01:42:07   ♪ They didn't even mean to begin ♪

01:42:09   ♪ 'Cause it was accidental ♪

01:42:11   ♪ Accidental ♪

01:42:12   ♪ Oh, it was accidental ♪

01:42:14   ♪ Accidental ♪

01:42:15   John didn't do any research, Marco and Casey wouldn't let him

01:42:20   'Cause it was accidental (it was accidental)

01:42:23   It was accidental (accidental)

01:42:25   And you can find the show notes at ATP.fm

01:42:31   And if you're into Twitter, you can follow them

01:42:35   @C-A-S-E-Y-L-I-S-S

01:42:39   So that's Casey List M-A-R-C-O-A-R-M

01:42:44   Anti-Marco Armin, S-I-R-A-C, USA, Syracuse. It's accidental. They didn't mean too

01:42:57   accidental. Tech podcast so long. Did you really just do that?

01:43:06   No, it should be in the after show. I was on the beta, so it's not like I'm

01:43:11   saying anything that like I should I should have been giving more feedback during the beta but

01:43:14   every time I went to the the beta feedback thing and I read it I read everyone else's feedback I'm

01:43:18   like yeah they're getting all this stuff like you know people would say things yeah it's good

01:43:23   someone said that are you outbound good something you had a lot of good beta testers so they they

01:43:27   covered everything more or less although yeah I mean I have a few things I could throw some of

01:43:32   these have already been said but they're both saying in the show what did they cover insufficiently

01:43:36   ah not insufficiently so I talked to you about this once in an after show before like things

01:43:40   are revealed. But let me just start with some... What I tweeted was like, "I'm using this app

01:43:47   instead of my iPod Shuffle." And I am. I'm off of my iPod Shuffle now. Part of that is

01:43:52   due to Bluetooth and not so much the application because I've switched over to wirelessly playing

01:43:57   it during my commute.

01:43:58   Just let me believe it's the app.

01:44:00   Yeah. But the other part is... Here's the other part. I'm one of those people who does

01:44:03   not care about smart speed. I don't listen to things at more than 1X. I don't need a

01:44:06   to volume booster.

01:44:07   So all those features you spent a long time ago

01:44:09   are meaningless to me.

01:44:11   But I would not be using your application at all

01:44:13   because I have other podcast apps that I purchased, right?

01:44:16   And I didn't use them for you for two

01:44:17   and again, part of that is Bluetooth.

01:44:18   But it's because what I was doing with my iPod shuffle,

01:44:21   besides cursing at it,

01:44:23   I was bringing it back to iTunes, plugging it in.

01:44:26   And at various times I would either,

01:44:28   with my old iPod shuffle,

01:44:29   I would manually drag individual episodes

01:44:33   onto the iPod shuffle.

01:44:35   like manually sync individual episodes, right?

01:44:38   And then when I got a new iPod Shuffle,

01:44:39   it didn't let me do it that way,

01:44:40   and I had to make a playlist,

01:44:41   but it was still just a manual playlist.

01:44:43   So I would drag individual episodes into a manual playlist,

01:44:46   and then synced them to the iPod Shuffle.

01:44:48   And what I was doing when I dragged them into the playlist,

01:44:50   I was like, all right, here's the order

01:44:52   that I wanna listen to,

01:44:52   because I didn't have a screen on the iPod Shuffle.

01:44:55   I just have to hit play when I get in the car.

01:44:56   I need them to play in exactly in the right order.

01:44:59   So I was prioritizing them.

01:45:01   And so I would say, oh, well, as soon as an episode of,

01:45:04   you know, whatever, Roderick on the Line comes out.

01:45:05   I wanna listen to that right away.

01:45:07   Even if I'm in the middle of an episode

01:45:09   of some other thing that I was listening to,

01:45:11   I want the Roderick on the Line to come on the top.

01:45:13   But, you know, I wanna do these three episodes

01:45:16   in this order and I wanna do them as a block.

01:45:18   It was like priority order but with manual adjustments.

01:45:20   And if that sounds familiar,

01:45:21   it's exactly what Overcast does.

01:45:23   (laughing)

01:45:24   And that is probably the,

01:45:26   maybe the one and only main reason I use this application

01:45:28   is because the playlists allow me to make an arrangement.

01:45:32   This is exactly what I was doing manually before.

01:45:34   The difference is now that, I mean, you know,

01:45:36   it's on a server somewhere, it's synchronized,

01:45:40   so like I installed Overcast on my iPad too,

01:45:42   not that I'm gonna listen from there.

01:45:44   It keeps track of where I am.

01:45:45   I don't have to have the problem of like,

01:45:47   switching off different environments

01:45:49   if I don't have the shuffle with me or whatever.

01:45:50   Like anything that's an iOS device

01:45:52   will know where I left off, what episode I was on,

01:45:54   what the play positions are, and all of those things.

01:45:55   And as I reorder things, that follows me around as well.

01:45:58   So this is an automation of the terrible manual process

01:46:02   I was doing before.

01:46:02   So that's why I used the application.

01:46:05   But-- all right, so this is supposed

01:46:07   to be me complaining.

01:46:08   I'm basically--

01:46:09   [LAUGHTER]

01:46:10   This is a great complaint.

01:46:12   What I'm trying to explain here is that when people said,

01:46:14   oh, well, you know, people on Twitter

01:46:16   were having-- there was this mini debate that spun off

01:46:18   from me saying that I used the application.

01:46:20   If, unless you care about exactly the same things

01:46:23   that I care about, that doesn't mean this application is

01:46:25   necessarily for you, right?

01:46:26   Because if you care about the smart speed stuff,

01:46:31   And I don't.

01:46:32   It's a whole different set of priorities.

01:46:34   And that's what you're getting at with all the people

01:46:35   with the feature requests.

01:46:36   If a podcast app is useless to you,

01:46:39   if it doesn't use video podcasts,

01:46:40   you're not gonna like this app

01:46:41   because it doesn't do video podcasts.

01:46:43   Everybody has something like that.

01:46:44   And so the same reason that I really like this application

01:46:48   is a reason that someone else might reject it entirely.

01:46:51   Or it's like the demands people have

01:46:54   of podcast applications are so different

01:46:57   that hearing just people like,

01:46:58   "Well, John recommends it.

01:46:59   It must be good."

01:47:00   No, maybe it's no good for you.

01:47:01   Maybe you want a feature that it doesn't have,

01:47:03   or in the case of video,

01:47:04   that it's probably never going to have.

01:47:06   Just because I recommend it doesn't mean

01:47:08   that oh, it's automatically good.

01:47:10   It just means it fits my needs appropriately.

01:47:13   And I think the Playlist feature in particular,

01:47:15   I can't be the only one who does,

01:47:17   maybe I'm the only one who did it manually

01:47:19   because it's just too much of a pain,

01:47:20   but once people can do it without the pain

01:47:22   that I was going through,

01:47:22   like they can just do it by just dragging

01:47:23   their thumb around,

01:47:25   I think a lot of people will find that feature,

01:47:28   like you were saying, your favorite feature,

01:47:29   I think that is the killer feature,

01:47:30   Even though it's like, oh, you let people reorder stuff

01:47:33   in a table view and did a little algorithm

01:47:34   to make stuff land in the right place, more or less.

01:47:37   That, I think, is the most important thing for me

01:47:40   for this application.

01:47:41   But my first complaint is related to that.

01:47:43   When I went to try to say, OK, well, Marco has this application.

01:47:46   He says I do his playlist stuff.

01:47:47   Let me-- first of all, I didn't have faith

01:47:49   that I was going to be able to do that.

01:47:50   I figured I'll just make a manual playlist,

01:47:51   and I'll just manually do stuff like I was doing before.

01:47:53   Even that would be an upgrade, because manually dragging

01:47:55   tracks onto the iPod Shuffle is a nightmare compared

01:47:57   to doing it on the screen.

01:47:59   As a previous iPod Shuffle owner,

01:48:01   using an iPod Shuffle in any way other than

01:48:04   playing a random selection of music tracks is horrible.

01:48:08   - Even random selection of music tracks is horrible

01:48:10   if you have other stuff on there.

01:48:11   (laughing)

01:48:12   The voiceover thing where it tells you

01:48:13   like if you happen to have music and podcasts,

01:48:16   you gotta make sure you're in the right realm

01:48:17   before you start playing.

01:48:18   - Oh wow, that was added after my Shuffle time.

01:48:21   I had the very first Shuffle and I think the third one.

01:48:24   - Yeah, I've had a series of them

01:48:26   and they're not my friends.

01:48:28   A lot of it has to do with iTunes being crappy too.

01:48:30   But anyway, when I was trying to make my first playlist,

01:48:33   I saw that it has this priority podcast thing.

01:48:36   And I'm like, oh, yeah, that's what I want.

01:48:38   And then I remembered you talking about the manual

01:48:39   reordering with the priorities.

01:48:41   I don't understand why priority podcasts as a thing

01:48:44   are a concept in the application.

01:48:45   Like, there is two kinds of podcasts.

01:48:47   There is a priority podcast, meaning

01:48:49   a podcast that can have a priority,

01:48:51   and then non-priority podcasts.

01:48:53   Why not just make all the pod-- because the first thing I do

01:48:55   when I make any playlists is I say select priority podcasts

01:48:59   and I select them all because they all have a priority.

01:49:02   Like there's no tail at the end where I'm like,

01:49:04   okay, this is my top five,

01:49:05   but everything else is just exactly equal.

01:49:08   Because what I find myself having to do

01:49:09   is go into the thing and say,

01:49:11   select priority podcasts, select, select, select,

01:49:13   select, select, and that step feels like

01:49:15   it doesn't need to exist.

01:49:16   Like that concept of a priority and a non-priority podcast

01:49:19   seems like a complication that could confuse people

01:49:21   into thinking they can't get the playlist they want,

01:49:23   when really they can, it's like,

01:49:24   "Oh no, you've got to mark those as priority podcasts first."

01:49:26   Now maybe I'm wrong and other people really do want to say, "I have three priority podcasts

01:49:29   and everything else I don't care, and it's just a big mush for me."

01:49:33   But I want all of them to be priority podcasts, because I know, I mean, there's not that many.

01:49:39   I mean, I read off my list, it's like maybe 10, 12 subscriptions.

01:49:43   I know which ones, what order I want them, all 12 of them.

01:49:47   It's not as if I haven't confused about any ones at the bottom.

01:49:49   Anyway, people in the chat room are saying they just want to prioritize a few of them.

01:49:51   But for me, that is an additional complication that confused me at first, and then it frustrated

01:49:57   me every time I had to go make a playlist, because I had to select the pri—and even

01:50:00   now when I add a subscription, because I still haven't even caught up, like I think I didn't

01:50:03   even put core intuition on here yet.

01:50:05   When I go add that, I'm going to have to remember to go into my playlists and mark

01:50:08   that as a priority podcast and sort it, rather than it just becoming the last priority podcast

01:50:14   or whatever, or the first one or something like that.

01:50:16   Now it's in the non-priority podcast category.

01:50:19   So I find that refreshing.

01:50:20   a couple of people have already complained to you about this,

01:50:22   but like this, I know I have a crappy A5 device,

01:50:25   but there's a pause sometimes when I tap on a thing

01:50:27   before it switches screens, before it starts playing,

01:50:29   I don't know what it's doing during that pause,

01:50:31   but it's not playing the audio.

01:50:32   It's for already downloaded stuff.

01:50:33   Do you have any idea what that?

01:50:35   - Yeah, no, that's real.

01:50:36   Yeah, that is like, I have to async that basically.

01:50:39   Like the API for reading audio files,

01:50:43   it's just like this old blocking C API.

01:50:48   And what I would love to do is have lower level access

01:50:52   to the files so I could like, whatever it's doing

01:50:54   for uncertain files where it has to like preload something

01:50:57   or preread something to get from zero to playing,

01:51:00   I would love to do that when it downloads

01:51:04   and then cache the result of that

01:51:06   so that way when you go to play it, it can play instantly.

01:51:09   But I don't have that kind of access in the APIs

01:51:12   and maybe I can go lower level somehow

01:51:14   but I haven't found a way.

01:51:15   - Some kind of like Amazon Fire predictive,

01:51:17   like if I've just added a podcast

01:51:19   or if I've just downloaded it,

01:51:20   chances are good that that's gonna be the one

01:51:22   that I'm next gonna play.

01:51:23   Or even if you wanna fake it out

01:51:25   and make it perceptually faster

01:51:27   by changing screens before the playing starts.

01:51:29   - Right, so that's the right answer

01:51:31   and that's what I have to do.

01:51:32   I just haven't done it yet.

01:51:34   The right answer is I have to like kick off the screen load

01:51:38   and then give the screen like a loading state

01:51:42   and then like give the now playing screen a loading state

01:51:46   and then kick off the load of the file asynchronously.

01:51:48   So I have to do that, it's on my to-do list

01:51:50   for 1.1 or whatever, I haven't gotten to it yet.

01:51:53   - Yeah, another thing that several people

01:51:55   have already complained about,

01:51:57   and I'm sure you know about,

01:51:58   but it's worth airing out for the listeners as well,

01:52:00   is the idea of going through

01:52:01   and cherry picking individual episodes,

01:52:03   having it boot you back out to whatever screen

01:52:04   it sends you back out to.

01:52:05   Like, it's not as easy to just go through,

01:52:07   oh, I want that episode of that,

01:52:08   oh, I want that episode of that,

01:52:09   you get sent back out to the home thing

01:52:11   and you gotta dig your way back in again,

01:52:12   and fine, I've been frustrated by that a few times.

01:52:15   In this one, I know you got to complain about this too, but I don't remember what your answer

01:52:21   for it was.

01:52:23   If you are unfortunate enough to hit the end of a podcast before you realize you've hit

01:52:27   the end of it, and it goes away, like, I don't know what the solution is, but you're saying

01:52:33   like maybe you could have a holding bin for it or whatever.

01:52:36   Several times I've either accidentally scrubbed to the end of the podcast, which is easy to

01:52:39   do with the fat thumb on the little scrubber thing or whatever, or it has run out when

01:52:43   and I have like, you know, it's been playing

01:52:45   and I took my earphones off and didn't notice

01:52:47   it was still playing or whatever and hit the end

01:52:48   and now the thing is gone and I have to re-download it.

01:52:50   What is, what are you doing about that one?

01:52:53   - That's tricky.

01:52:54   Most of the feature requests I've gotten so far today

01:52:56   or the people who are like, oh I cannot use this app

01:52:58   because I need to manage X.

01:53:01   Almost all of them are storage management of some sort.

01:53:06   Like, a lot of people are, like one request

01:53:08   that I keep getting which I never really considered

01:53:11   that people would want this, but they do,

01:53:13   I just didn't cross my mind, I guess,

01:53:15   is a different state, a state of things that are new

01:53:21   or that you haven't played and deleted,

01:53:23   so they're the active state of the item,

01:53:26   but that is not downloaded and won't be downloaded.

01:53:30   'Cause right now, my states are basically,

01:53:32   needs to be downloaded, downloaded, and deleted.

01:53:36   That's it, those are the three states an item could be in.

01:53:38   - Yeah, like a pre-download, like a will-be downloaded queue.

01:53:41   - No, but what people are asking for is

01:53:43   they want items that, like on certain devices,

01:53:47   just are never automatically downloaded.

01:53:49   - Right, like it's a queue that doesn't run.

01:53:51   - Like it's a new item, yeah, it's a new item

01:53:53   that I want to keep as new, like keep as unplayed,

01:53:56   but don't have the app try to download it.

01:53:59   And maybe I'll try to download it some other time.

01:54:01   You know, so there's, and that's a very common request

01:54:05   so far in the email.

01:54:06   Again, I was not expecting that at all.

01:54:07   In general, looking at the state of episodes, like when I'm going through any list, it's

01:54:12   not always easy to tell what state is that.

01:54:16   Some kind of iconography or visual elements to say...

01:54:19   Because you have those states in your head, like, "Does this need to be downloaded?

01:54:23   Has it already been downloaded?

01:54:24   Has it been downloaded and started?

01:54:26   How far progress in it, or is it one of those weird things you just described where it's

01:54:29   not supposed to be downloaded on this device?"

01:54:31   You've got the little "i" there to try to figure out what's what, but...

01:54:36   Instacast is the main other application that I've used, and I forget what its iconography,

01:54:39   it's like a little green corner thing when it's been downloaded, and maybe there's a

01:54:45   progress bar or something, I don't even know what the hell it uses, but some visual way

01:54:48   to look at a list, and when I look at these lists, I mentioned before my nervousness about

01:54:52   going on a plane flight to WVDC and saying, "I don't know if all the podcasts I want to

01:54:55   listen to are downloaded or not."

01:54:57   I can't glance at the app and tell that.

01:54:59   I could go into each individual thing and see if it's downloaded and try to start playing

01:55:02   to make sure that it's there, but like,

01:55:05   some sort of visual reassurance or acknowledgement

01:55:07   of the state of each episode.

01:55:09   - See, and that, this is gonna be a hard balance to strike.

01:55:13   You know, podcast apps, you know,

01:55:16   I think on one extreme, you have apps like Downcast,

01:55:20   I think iCatcher is like this,

01:55:22   where they just have tons of customizable settings

01:55:25   for everything.

01:55:26   That comes at a pretty big cost of interface complexity.

01:55:30   And this app is trying to be mass market.

01:55:35   I'm trying to, I'm really trying to compete

01:55:40   with Apple and Stitcher.

01:55:42   Like that's what I'm looking at.

01:55:43   I'm not trying to compete with Downcast

01:55:46   and Pocketcast and Instacast.

01:55:48   - But this isn't like a twitchy feature.

01:55:50   This is like a simple, like a visual,

01:55:53   I mean, I guess it's like visual clutter

01:55:55   you're trying to not have.

01:55:56   Like you want to, you need to convey

01:55:57   a certain amount of information

01:55:58   and you don't wanna make it clutter.

01:55:59   You don't want to have like you look at an episode list and it has like a million little

01:56:03   stickers and dials all over it telling you like obscure icons that you're like, what

01:56:06   does that mean?

01:56:07   It's like a lollipop and there's a red circle and there's a line through it and there's

01:56:09   a thing that looks like it could be a progress bar.

01:56:13   But I think that state information about which one was I just playing, which thing is currently

01:56:19   paused.

01:56:20   I mean, like you said with the big giant EQ meter over the things like that gives an indication

01:56:23   of what's currently playing.

01:56:25   But for all the other states, like has this been downloaded?

01:56:27   Is this going to be downloaded?

01:56:29   far am I in the episode, is it played or not? And by the way, someone in the chat room mentioned

01:56:32   a reasonable solution to what happens when the timer goes off the end. Have a recently

01:56:38   played list that you just pull off the end and delete from that after some period of

01:56:42   time. So if you accidentally scrub your way through an episode, you can go to the recently

01:56:44   played list. I don't know what the good solutions are. The reason I didn't put a lot of these

01:56:48   in the beta feedback is there was no-- I can't say, oh, obviously you should need to do whatever.

01:56:53   Because I would try to think, well, how would you solve that problem? They're not easy.

01:56:57   There's only so much room for buttons,

01:56:58   there's only so much room for visual clutter,

01:57:00   and you have a navigation hierarchy that you,

01:57:02   I mean, you don't wanna break out of that

01:57:04   and start going like three-dimensional chess

01:57:05   where you normally go right to left,

01:57:07   but you can also zoom in and out.

01:57:09   It gets confusing, 'cause like you said,

01:57:11   there's a lot of screens.

01:57:12   - Right, and the question with all these features,

01:57:15   you know, if, okay, I want a way to do X,

01:57:18   I want a way to toggle this state

01:57:19   without affecting this state,

01:57:20   so I want a new state of this,

01:57:21   or I want a new option to do this.

01:57:23   One of the big questions is,

01:57:24   well, where do you put all this stuff,

01:57:25   and what do you do in the interface?

01:57:27   Is that hidden behind a gesture?

01:57:30   If so, how does it explain to anybody?

01:57:33   Is it behind an info button?

01:57:36   Well, where do you put the info button?

01:57:38   Is it in the settings screen?

01:57:40   Well, then how big is the settings screen?

01:57:42   The apps I've mentioned, the ones that have a lot of options

01:57:44   like Downcast and iCast, they have six different pages

01:57:47   in their settings screens 'cause they have,

01:57:49   all these settings that people request have to go somewhere.

01:57:52   And the UIs for that can be pretty clunky and cumbersome

01:57:57   and can turn off a lot of people.

01:57:58   And so I'm trying to balance this,

01:58:01   I'm trying to balance between trying to give people

01:58:06   what they want, or rather, trying to give people

01:58:09   what they ask for, and trying to be a mass market appeal app.

01:58:14   I don't want to add a million features.

01:58:17   I'm never gonna win that race,

01:58:21   mostly 'cause I can't, but secondarily,

01:58:23   'cause I don't want to.

01:58:24   If you want tons of fine-grained controls

01:58:28   over all that stuff,

01:58:30   I think you'll be better off using downcast.

01:58:33   Just use downcast.

01:58:34   - You can just concentrate on adding,

01:58:36   making it, the features you already have,

01:58:38   the features that are already there,

01:58:39   making it clearer to the people using those features

01:58:42   what's going on, 'cause you have a certain amount of state

01:58:44   that you already support, right?

01:58:46   And you have certain features that you already support,

01:58:48   but a lot of them, like you said,

01:58:50   might be hidden.

01:58:52   Just to give another example, I'm looking at the app now.

01:58:54   On the little download icon, it has a badge that says five.

01:58:57   I tap-- first of all, I don't know what that badge means.

01:59:00   I assume that means it's five downloads.

01:59:02   Are they in progress or whatever?

01:59:03   I tap on the little five, I see a list of five episodes.

01:59:05   They all say download paused underneath them

01:59:07   in very light text.

01:59:09   It would be hard to read if I was even older.

01:59:11   I don't know why these are all paused.

01:59:14   And this is an example of the mysteries of Overcast.

01:59:19   Sometimes it does things that I don't understand.

01:59:21   It's best, it's best, working best

01:59:24   when it's like you're in the cycle.

01:59:25   - Now you see why I wrote three downloaders.

01:59:27   Go ahead.

01:59:28   - Well, it's best when you're in sort of the cycle

01:59:29   where like I was in a long period of time

01:59:31   where it's just like I didn't do anything to the application.

01:59:34   I had set up my subscriptions, I had set up my playlists.

01:59:37   I would wake up in the morning,

01:59:38   all my podcasts would be there

01:59:40   in exactly the order that I wanted them.

01:59:42   And they would just play.

01:59:43   Like that's, you know, that's how you want it to work.

01:59:45   It's like, I don't have to touch this application,

01:59:47   it does exactly what I want.

01:59:48   I never have to wait for it or whatever.

01:59:49   But when something does go wrong

01:59:51   or I wanna do something weird like,

01:59:52   oh, someone recommended,

01:59:53   this is before the Twitter recommendations,

01:59:55   but oh, someone recommended that I should listen to

01:59:57   that episode of Electric Shadow

01:59:59   where they talk about Apple TV.

02:00:00   So I go into search, find Electric Shadow,

02:00:01   find the episode, I'm like,

02:00:02   well, how do I get this into my thing?

02:00:04   Oh, now I gotta mark it as a priority podcast

02:00:05   or maybe I could just manually arrange that.

02:00:07   Am I subscribed to Electric Shadow now

02:00:09   or did I just download the one episode?

02:00:10   Is it going to download all the rest of the Electric Shadows

02:00:14   or does it know I just got that one?

02:00:15   Like these are features that already exist.

02:00:16   Like, you can do all of these things,

02:00:18   but I was unsure whether I was using the app in the right way

02:00:22   to do them.

02:00:23   You know what I mean?

02:00:23   And that, I think, rather than adding features other

02:00:26   than the ones you think you have to add,

02:00:27   is probably the place to concentrate, especially

02:00:30   for a mass market.

02:00:30   Because this app already does tons of stuff

02:00:32   and already has tons of features,

02:00:34   but it's not always clear to the inexperienced user

02:00:37   or the new user that those features exist

02:00:41   or that people are successfully using them.

02:00:44   Yeah, you're totally right.

02:00:45   I agree.

02:00:46   I'm not gonna argue with you there.

02:00:48   I agree, there's a lot of things in the app

02:00:50   that are not as clear as they should be

02:00:51   and that are not as intuitive as they should be.

02:00:54   And a lot of features that are kind of hidden

02:00:56   that probably shouldn't be.

02:01:00   It's 1.0 and I'm open to suggestions, so file a bug.

02:01:04   - Yeah, well that's the problem.

02:01:05   Like if I had all the good suggestions,

02:01:07   I'd be telling you exactly what to do,

02:01:09   but it's not an easy problem.

02:01:10   - That's exactly.

02:01:12   - It's not like, oh, it's just,

02:01:14   it's a lot of times it is.

02:01:14   You use an application and you're like,

02:01:16   This is obvious you should be doing this and it's up for some of the feature features are easy to do like oh marker

02:01:20   You should totally let me start listening immediately and then say that as a down like you know you need to do that

02:01:24   That's yeah technical thing you can tell the person exactly how you wanted to work, and it's fine

02:01:28   But the UI thinks it's like so what do you add? What buttons do you add? What buttons do you remove?

02:01:32   But you know you've already got swipe and tap where else you're gonna stick this other thing

02:01:36   That's why I keep thinking visually like I don't know what the design would look like

02:01:41   I don't think I think you could get something that fits in the theme

02:01:43   but that's the only thing I have a concrete suggestion for is like

02:01:46   some kind of visual indication of the states that already exists for episodes and such and the navigation thing of when you're trying to

02:01:54   Cherry-pick episodes not getting chalked back out

02:01:56   But again, you've heard most all this feedback before and I'm more or less just airing it on the podcast

02:02:00   So other people can know that you've heard this feedback

02:02:02   You should just make that you should just make the beta board public and say people have said all these things before take a look

02:02:07   To some degree it's a difference of philosophy and on how much of this stuff that people are asking for I should even offer

02:02:13   I would much rather favor good default behavior

02:02:18   than a preference if I can get away with it.

02:02:22   And there's sometimes when you can't get away with that.

02:02:24   Sometimes if people are really divided on something

02:02:26   and it's like 50/50, half people want this,

02:02:28   half people want that, neither of them is clearly

02:02:31   better than the other or just a difference of preference,

02:02:33   then you usually have to have a setting somewhere for that.

02:02:35   - Do you think you need a setting for the

02:02:37   hit the end of the episode and it immediately gets deleted

02:02:39   and you can't find it again?

02:02:41   Because the disk space people probably want that.

02:02:43   People who are like, I want it to be gone immediately.

02:02:45   But if you just put them into the recently played episodes

02:02:52   thing and got rid of them after five minutes,

02:02:54   the disk space people wouldn't care.

02:02:56   But the I accidentally went to the end people.

02:02:58   Did you mean to scroll to the top by hitting the status bar?

02:03:01   Or did you not really?

02:03:02   An Instapaper?

02:03:03   You need to have that feature, but sometimes people

02:03:06   do it accidentally.

02:03:07   This is a similar situation.

02:03:08   It's a frustrating situation when you hit-- especially

02:03:11   when I'm trying to scrub around, even on our own episode,

02:03:13   I was trying to listen to something we were saying

02:03:14   in an after show, and I was trying to scrub around

02:03:16   near the end of the thing to find, to go back,

02:03:19   'cause I had missed something, 'cause you know,

02:03:21   I've been distracted for a second,

02:03:22   and I scrubbed off the end of the episode, went away.

02:03:24   That's not good.

02:03:26   And like the disk space people, I'm like,

02:03:27   you can wait five minutes disk space people.

02:03:30   I'm one of the disk space people.

02:03:31   - Believe me, I will get those emails.

02:03:33   - I'm always running out of space on my devices too,

02:03:35   but it's much worse to accidentally try to, you know,

02:03:38   scrub back to hear something that Casey said,

02:03:40   accidentally go off the end and uh...

02:03:42   yeah, and people were asking for a skip to skip the ATP theme song and I'm like

02:03:46   "Oh, people like the theme song, right?" But I was at... I find myself, speaking of going off

02:03:50   the end, at the end of Back to Work, I don't like listening to that theme song. As soon as it starts

02:03:54   playing I want to be done. But to make the episode go away, I have to manually,

02:03:58   as far as I know, manually thumb the scrubber to the end to make that episode,

02:04:03   you know, be marked as played, where as far as I'm concerned, as soon as the little music

02:04:06   starts playing I'm done with the episode.

02:04:08   - Yeah, I just hit fast forward like three times,

02:04:11   which is enough to cover it,

02:04:11   and then I just delete the cell phone

02:04:13   and goes to the next episode.

02:04:14   - Yeah, I know, there's no good solution to that,

02:04:16   but what I'm saying is like the things that happen

02:04:18   near the borders at the end of the episode

02:04:21   and fighting with the little, you know,

02:04:23   the behavior of should it go away?

02:04:24   Sometimes I want it to go away immediately,

02:04:25   sometimes I don't want it to go away accidentally.

02:04:28   But I think just like the recycle bin or trash can

02:04:32   or whatever, having some sort of buffer zone

02:04:34   for things to stay in for a short period of time

02:04:36   without any settings related to that,

02:04:38   Just say that's the behavior,

02:04:39   the people who need the space will just, you know,

02:04:42   just tell them it's immediately freed

02:04:43   and then five minutes later it will be.

02:04:45   - Yeah, right.

02:04:47   Yeah, I mean that's worth considering,

02:04:49   but then I have to have a UI for,

02:04:52   okay, well, how do you get to these deleted,

02:04:55   do they show up in a special list somewhere?

02:04:57   - Yeah, I mean like you have playlists, right?

02:05:00   So I would just keep using that metaphor,

02:05:01   like these are smart playlists that are built in.

02:05:04   You know what I mean?

02:05:05   - Yeah, it's, I don't know.

02:05:07   there's enough cost to that feature

02:05:09   that I'm not sure it's worth it.

02:05:12   And then there's also, there's like the cognitive burden

02:05:14   that you put on all the users.

02:05:16   When an app could behave in two or three different ways,

02:05:20   then the user's trying to figure out,

02:05:22   okay, well wait, I just did this, then where is this,

02:05:25   or what happened, or do I have this file?

02:05:27   Like you said earlier, that you wanted to be sure

02:05:31   if you're getting on a plane or something,

02:05:32   you wanna know whether everything's downloaded or not.

02:05:35   One of the reasons why I've kept the storage model so simple

02:05:38   is to provide that assurance, to know like,

02:05:42   okay, if there is no number badge or frowny face

02:05:47   on the downloads icon, then I have everything.

02:05:49   That's it, like that's it.

02:05:51   If the downloads icon is not showing a status--

02:05:53   - I didn't know that.

02:05:54   That concept is clear to you,

02:05:56   but I didn't know that that should have,

02:05:57   I'm glad I now know that, so I will look at that

02:05:59   and if that's the model you're,

02:06:00   but that model, that's the, you know,

02:06:02   the programmer model, that's in your head,

02:06:03   but the user model is not, like there's nothing in the UI

02:06:06   that tells me that that should be the case.

02:06:09   The way I can know that I'm safe is I just look

02:06:11   to see if there's any badge on the download icon

02:06:13   and if there isn't, I'm safe.

02:06:14   - Right, and you got burned early in the beta

02:06:17   because I changed this halfway through the beta.

02:06:20   Earlier on when it first hit you,

02:06:21   when you first complained about it,

02:06:23   I was not validating the files I was getting.

02:06:26   If the download gave me a file

02:06:29   and the download completed successfully,

02:06:31   I said, "Okay, it succeeded, done," and I moved on.

02:06:34   And then you'd tap the file and you'd get the

02:06:36   that is unfortunate error.

02:06:37   So, and it would say, "Oh, I can't play this file,

02:06:41   "I don't know why."

02:06:41   That error means that opening the file

02:06:45   in the audio processor failed.

02:06:47   And often that was 'cause it was a bad download.

02:06:50   Sometimes it was like, it was not actually an audio file,

02:06:53   it was like somebody's 503 page

02:06:55   that when there was a server error.

02:06:56   (laughing)

02:06:58   So what I do now in the downloader,

02:07:00   after you made that complaint that was correct.

02:07:02   And so what I do now is as soon as the files download,

02:07:07   there's a brief pause where it seems to freeze at 99%.

02:07:11   And the reason why is 'cause it's loading it

02:07:13   into the audio processor right then

02:07:15   to see is this a real audio file and how long is it

02:07:18   and is there anything I need to get,

02:07:21   any information I need to get from this.

02:07:22   That way once it downloads,

02:07:24   it can give you an accurate duration.

02:07:26   And if it's not actually an audio file,

02:07:28   it can consider that a failed download.

02:07:30   and then show it in the UI and maybe try to retry it later

02:07:32   or something like that.

02:07:33   So now with that, you can be pretty sure

02:07:37   that what you're getting, as soon as it was downloaded,

02:07:40   it was tested in the thing that opens audio files.

02:07:43   So it will open.

02:07:45   - Well, I know you've got the accidental delete

02:07:47   by scrubbing complaint before,

02:07:48   and I think you'll get it again.

02:07:49   So maybe you got it from a couple of people in the beta.

02:07:53   Maybe you'll check your email

02:07:54   to see if it comes up more often.

02:07:55   but I still say that scrubbing around in a downloaded episode

02:08:00   should never cause that episode to be deleted.

02:08:03   Despite the fact that yes, I agree

02:08:04   that when you're done playing the episode,

02:08:06   having it be automatically deleted

02:08:07   is pretty much what people wanna do.

02:08:09   But me moving my thumb around

02:08:11   or even hitting the fast forward button

02:08:13   should never cause that episode to be deleted.

02:08:14   'Cause if I am on the plane, everything's downloaded,

02:08:16   and I'm listening to something,

02:08:18   and at the beginning of the episode,

02:08:20   they make a reference to something at the end,

02:08:21   I'm like, oh, I wanna skip to the end

02:08:23   and see what they're referring to,

02:08:24   and I go skip to the end, I guess delete it,

02:08:25   like I wanted to listen to that to our podcast,

02:08:27   now I'm in a plane with no wifi and I can't.

02:08:29   Like scrubbing, using the scrubber does not seem

02:08:32   like it should be something that causes data loss.

02:08:35   - Well, but because like--

02:08:37   - But I mean, you just wait, wait to see

02:08:39   what the feedback is.

02:08:40   Like maybe I'm the only one, there was like three,

02:08:41   two or three other people who had similar comments

02:08:43   in the beta.

02:08:44   - No, I've had it happen, I've had it happen.

02:08:46   - I mean, just, if it's a problem, you'll find out.

02:08:49   'Cause right now, like when the people download on day one,

02:08:50   maybe they don't notice.

02:08:51   But once they use it for a week or so,

02:08:53   you'll find out what the real percentage is.

02:08:55   - Yeah, I mean one thing I did about,

02:08:57   in like the second or third beta,

02:08:59   I changed the scrubber so that before it was continuous,

02:09:02   as soon as you would drag it around,

02:09:04   it would seek to that point.

02:09:05   And so if you just held your finger over

02:09:07   and dragged to the right,

02:09:09   it would eventually,

02:09:10   like as soon as you hit the far right edge,

02:09:13   bam, deleted the podcast.

02:09:14   The way it's released now,

02:09:15   and the way it was about a third of the way into the beta,

02:09:18   is it's momentary, or rather, you know,

02:09:20   so like you have to touch down, drag it around.

02:09:23   As you're dragging it around,

02:09:25   it doesn't actually seek until you release it.

02:09:28   And so in order to hit the end and delete the file

02:09:31   on the scrubber, you have to drag it to the end

02:09:33   and release it while it's at the end

02:09:35   to actually have that happen.

02:09:37   - Yeah, when you're trying to get to like the last 30 seconds

02:09:39   of a two hour podcast and your big thumb is covering

02:09:42   like the timestamp and you can't really, you know,

02:09:44   it's easy to accidentally even release,

02:09:47   even just to take a look over the timestamp

02:09:49   is not realize you've hit the end.

02:09:50   - Yeah, that's fair.

02:09:51   I mean, what you're identifying is definitely not ideal.

02:09:55   The question is whether the alternatives are worse.

02:09:59   Like whether the alternative complexity

02:10:01   is actually overall worse.

02:10:04   And I don't know the answer to that.

02:10:05   I mean, I'm sure there's ways I can improve it

02:10:07   that I haven't thought of yet.

02:10:08   But you know, like adding this whole different state

02:10:13   of episodes of like this like purgatory state,

02:10:17   which would be confusing.

02:10:19   A lot of people would never even realize

02:10:20   they could go back and get them,

02:10:22   and the story people would be very upset by it,

02:10:25   and it would be more work, and things would be

02:10:26   more complicated, and there'd be more weird bug

02:10:28   edge case potential, so like--

02:10:30   - Storage people would never know.

02:10:31   Anyway, you can always make it a setting in that case.

02:10:35   It's five minutes, you deleted in fact five minutes out.

02:10:39   The other weird thing that, I mean,

02:10:40   there's limited space for navigation here,

02:10:42   but I wanted to show my wife the directory today,

02:10:46   and I had to think, now where was the directory under?

02:10:48   'cause if I look at the top,

02:10:49   it's icon that really should be a gear,

02:10:51   but Marco's stubborn.

02:10:52   Download, add a playlist, and then a plus.

02:10:56   - I agree with the icon.

02:10:57   I don't like the gear.

02:10:58   Well, that's not fair.

02:10:59   I think that the icon works.

02:11:02   - It's a perfectly nice icon.

02:11:03   I'm just saying that gear is the symbol for settings.

02:11:06   Anyway, and I had to think like,

02:11:08   well, it's not gonna be under downloads

02:11:10   'cause it's a directory.

02:11:11   The little plus thing with the document,

02:11:13   I had to keep reinforcing to myself

02:11:15   that that means that's add playlist, right?

02:11:17   - Yeah.

02:11:18   And so the plus is like, well, by a process of elimination,

02:11:22   it has to be the plus to see the directory.

02:11:24   And that makes sense, because you're adding a podcast,

02:11:26   kind of.

02:11:26   But it's like, I expect that to be like the search or whatever,

02:11:29   but it's also the directory.

02:11:31   Anyway, my wife searched for ATP and didn't find this podcast.

02:11:33   So that's her bug report.

02:11:35   You already got that one, too, though.

02:11:36   Well, it's probably somewhere in my--

02:11:39   it's now my 590 emails?

02:11:41   I mean, someone tweeted.

02:11:43   My wife showed it to me, and I said, take a screenshot

02:11:45   and send it to Marco.

02:11:46   You can use Bugshot.

02:11:47   Put a big red arrow.

02:11:48   I think she actually did email you about it.

02:11:50   But anyway, that has to do with like,

02:11:51   we have to put ATP somewhere in our metadata

02:11:53   or description or something for that to work.

02:11:55   - I don't even know.

02:11:56   I got, I don't know.

02:11:58   - You're toast.

02:12:00   - Yeah, well, the other option is

02:12:02   when you launch the application,

02:12:03   just have a gigantic big subscribe to ATP button.

02:12:06   - Well, it's already in the directory.

02:12:08   It's in the tech category,

02:12:10   which is the top left category in the directories.

02:12:12   - Yeah, yeah, we're almost there.

02:12:14   Anyway, that's plenty for now.

02:12:17   - See, the important thing here is that I ship

02:12:20   before FastText's update for iOS 7.

02:12:22   - Yes, yes, yes.

02:12:24   - I cannot believe I beat you.

02:12:26   - To be honest, I've worked on FastText

02:12:27   like once, maybe twice since iOS 7 came out,

02:12:30   but that doesn't negate your point,

02:12:32   which is that you wrote, what,

02:12:34   three quarters of an application in the time

02:12:36   that I couldn't basically just recompile for iOS 7.

02:12:40   (laughing)

02:12:42   - Oh, that's the best.

02:12:43   I'm so happy.

02:12:44   When you first joked about that,

02:12:47   It was probably six months ago.

02:12:49   - Oh, probably.

02:12:49   - And I thought for sure, oh yeah, I'm never gonna--

02:12:51   - I thought for sure I'd make it,

02:12:52   but I just haven't found the time.

02:12:55   - Oh, that's amazing.

02:12:56   - Yeah.

02:12:57   - That makes me so happy.

02:12:59   - I did it to myself, I can't even be upset.

02:13:01   Well, certainly not at you anyway,

02:13:02   'cause I did it to myself.

02:13:03   [