I’m : a programmer, writer, podcaster, geek, and coffee enthusiast.

New Territory

After last year’s WWDC, I argued in Fertile Ground that iOS 7 was a huge opportunity for developers: with so much change required for established apps to remain competitive, anyone making a new app had a big advantage and a great chance of establishing a foothold. Established markets were plowed over and shaken up, leaving opportunity.

This year, the opportunity is different, but even bigger. With iOS 8’s new Extensions, entire categories of apps that were previously impossible are now possible. Rather than shaking up the existing apps, Apple has created vast new markets that are currently empty.

We’ll see the easy stuff on day one, of course. Apps like Instapaper can stop using clunky Javascript bookmarklets and URL schemes to shuttle data out of Safari or between apps. Photo apps can now operate on the photo library as first-class citizens, just like the built-in Photos app. Notification Center can now contain status message widgets, shortcuts for launchers, quick-entry fields for task managers or note-taking apps, and almost anything else an app wants to put there (with user permission, of course). Almost every existing app can now do something very useful that it couldn’t do before.

But existing apps were all written by people like me who have been solidifying the assumption in our minds over our entire iOS careers that these features simply aren’t and never will be possible on iOS. We never thought we’d get these. It’s going to take a while before we internalize these new abilities and forget the old restrictions.

The first amazing, forehead-smacking innovations with iOS 8 won’t come from us: they’ll come from people who are coming to iOS development from this point forward, never having known a world with the old restrictions.

It’s a great time to be an iOS developer.

FCModel and Swift

After years of writing Objective-C, I was finally confident enough in my skills and understanding of the language to write and open-source a general-purpose model layer, FCModel, as a lightweight Core Data alternative that’s closer to SQL. It has attracted a small but solid audience, with many invaluable contributions and bugfixes.

Apparently, it somewhat works with Swift. But I don’t plan to port it in the near future.

Swift is a new language with new behaviors and, most importantly, new standards and expectations. I expect to accumulate enough Swift knowledge and experience to be able to make something like FCModel for it, but:

By that time, SwiftModel (or whatever) may not need to exist. Core Data in iOS 8 and 10.10 matched some of its major advantages, so maybe we’ll all just use that. Or maybe someone else will come up with a better SQLite model implementation for Swift by then.

I just wrote a huge app with FCModel, so I’m sticking with it for a while in Objective-C, but I don’t intend to port it to Swift soon, or in a recognizable or compatible form.

Don’t Go To College

In The Talk Show Live at WWDC 2014, I joked about college not being necessary if you thought you didn’t need it.

Attempts at humor are often missed. In this case, a lot of people missed it, which was my fault. To clarify, I was joking.

Go to college if you’re fortunate enough to have the opportunity. I did, I learned a lot (both academically and socially), and I wouldn’t trade that experience for the world. Not everyone needs college, but you should go if you can.

My philosophy about being a C student and not needing to do 80% of the work should also be taken lightly. That strategy works well if you want to follow a path like mine after college: working for small companies that care less about your GPA, or that you can convince to hire you by other means (showing impressive personal projects, advanced skills, etc.). But my GPA was so bad that no big tech company — not Microsoft, not Apple, not Amazon, and definitely not Google — would even consider hiring me. (I tried.)

I was a C student because I was (and am) a slacker and lacked the self-discipline to do better, not because it’s the smartest path to take. Performing better opens more doors.

Podcast Networks Are The Wrong Model

Last month, I wrote, “The glory days of podcast networks are behind us.”

A lot of people have asked me about that since then. Many have argued that I’m wrong because some podcast networks are still good or useful, but I hadn’t argued otherwise. I meant exactly what I wrote: that podcast networks, as we know them today, have peaked, much like blog networks did years ago. Like blog networks, they’ll probably always be around as long as podcasts exist, but the era has ended in which being a member of a podcast network is the dominant, default, stupid-if-you-don’t-do-it method of having a popular, well-produced, profitable podcast.

Podcast networks are like record labels: they promise exposure, tools, distribution, and money. But as the medium and infrastructure mature, their services are often unnecessary, outdated, and a bad deal for publishers.

Writers don’t need blog networks to be successful today because of two major shifts since blogging began:

Networks primarily make sense before those conditions are reached. Podcasting has made major progress in both areas in the last few years, and will likely progress even further in the near future.

Some roles provided by networks are useful to podcasters, but there’s no reason that these roles must be tied to their membership in a network — podcasters are better off having services available that can be hired and fired at will without having to move or lose their shows, the same way web hosts, designers, and salespeople are hired.

Podcasters should own their shows and have contract services available to help, and their choices of such services should be neither apparent to their listeners nor part of their shows’ identities.

Today, ATP’s website is hosted by Squarespace. Its files are hosted by Libsyn. Its ads are sold by Standard. Its domain is registered at Gandi. It’s edited in Logic. If you listen, how many of these were you aware of? Do you really need to care?

None of these names appear in the artwork or domain name. Any of these can change at any time without bothering the listeners. We’ve owned the show, relationships, and audience access from the start, and we intend to keep it that way.

And I’ll do what I can to help and promote this as the way forward for everyone.

Update: About 5by5.

The Elephant In the Podcast Studio

Since last night’s post, I’ve had about a hundred people on Twitter accuse me of implicitly dismissing or denying the role of 5by5 in making our podcast popular, since two of us previously had shows on 5by5.

First, I didn’t mention 5by5 in the post because it wasn’t relevant to the point I was making: that joining a podcast network, as we know them today, no longer needs to be the dominant way to make a successful, profitable podcast. I don’t know how many times I need to repeat this, so here’s one more try.

5by5 will also be fine. Not only are they one of the few midsized networks that I think will still be around in five years, but Dan’s a smart guy, and he sees the same shifts that I’m seeing. That’s probably why he split his ad-sales operation into a standalone, non-network service.

People want to find secret drama in the end of my 5by5 show, but the truth simply isn’t that interesting. I decided to end the show for the same reasons I said at the time: it was getting repetitive and declining in quality, and I didn’t want to keep doing it until it really sucked.

When we got together and did Neutral, we set it up independently because it simply wasn’t very hard. I was already selling sponsorships for my blog, and we hosted it on Squarespace with very little effort. When we accidentally pivoted into a tech show, we kept the same setup. It wasn’t an explicit decision at the time not to join a network — the need just didn’t arise.

5by5 offered something that I needed badly in 2010: a podcasting setup where the administrative work was done for me. I prepared an outline of topics, I showed up on Skype, Dan showed up a few minutes late, and we talked through my outline for an hour. Afterward, I added the relevant links to the show notes, and then my job was done.

It might have sounded like “my” show, but it wasn’t — 5by5 owned and controlled everything about it. It was another 5by5 show, but with me as the co-host.1 I was paid a small flat fee every week, but the main value to me in doing the show was exposure. I was never told how many listeners the show had or how much money it earned.

5by5’s work has a lot of value, especially if you don’t want to do any of that work. It has even more value if Dan’s the host of your show, since then you need to do even less work, and you get the value of having a great host.

But everyone’s better off with more options to publish in a medium. If podcast networks continue to dominate the creation and distribution of good podcasts, we’ll have very few good podcasts. Networks don’t scale. And people are better off choosing their web host and ad salespeople on their merits, rather than because they’re stuck with the ones they started their show with. Today, podcasters have far more network-less options for services like these than we did in 2010.

A significant portion of my current audience came from my 5by5 show, and some portion of them certainly found the show first because it was on 5by5. But I also had an audience from here, Tumblr, and Instapaper for years before that. I don’t know how many people listened to my show because they knew me already or because they found it on 5by5, but I do know that being on a network doesn’t boost “discoverability” as much as people think it does.

Discoverability is overrated. The real way to get more listeners is to make a great, relevant show. The best content tends to be found, but it takes hard work and dedication. Each episode of ATP takes about 10–12 person-hours of work,2 and we’ve been doing it every week for the last 70 weeks. Our show is successful primarily because we put a lot of effort into it and have chosen a topic that fits our existing audience and the current podcast market well.

Having a “built-in audience” from two 5by5 shows didn’t sustain Neutral. It peaked at about 1,000 listeners — when we ended Neutral, the audience for the weeks-old ATP was already 15 times larger and skyrocketing.

Having a “built-in audience” from a 5by5 show, Tumblr, Instapaper, and didn’t sustain The Magazine. I sold it because it was cratering under my ownership and losing subscribers alarmingly quickly. I was about to shut it down, but Glenn wanted to try running it, so I sold it to him for much less than you’d probably assume.

Having a “built-in audience” from a 5by5 show, Tumblr, Instapaper, and gave Bugshot a respectable launch, but the initial spike fell to nothing, and even making it free a few months later didn’t help long-term.

Neutral simply wasn’t as good as ATP and wasn’t as relevant to the audience. The Magazine under my leadership was subjectless, unfocused, and irrelevant to most of my audience. Bugshot was only useful to a few people, and I didn’t put much time into it. All of these had the benefits of a “built-in audience” to give them an initial spike, but none succeeded because they simply weren’t good enough.

Podcast networks don’t guarantee success or profit. The best networks can sometimes give you a head start and help you get found sooner, but the quality and relevance of your show is what sustains it, not the brand name on the artwork.

5by5 certainly played a significant role in growing my audience, but it was far from the only factor.

  1. The running “It’s your show” joke isn’t entirely a joke.

    If you’ve been following me for long, it shouldn’t be a huge surprise that for my future ventures, I tried independent options first. It was nothing against Dan — I just like owning what I do. ↩︎

  2. Recording needs about 2.5 hours from each of the 3 hosts, plus 3–5 hours of editing, sponsor preparation, show notes, and publishing. ↩︎