Markets as Silos

Today, I have seen an announcement about yet another Android application market. This one has a slick Web site, a wide range of features, and apparently little regard for copyright. More relevant to this post, though, is that this market, like most (if not all) of the others, thinks that it is an island unto itself.

I can understand why developers get frustrated.

Tactically, all these markets are unsupportable by developers. Developers cannot manage dozens of legal agreements and dozens of registration forms and dozens of update processes. Hence, developers pick and choose, with the vast majority electing to distribute solely on the Android Market. Those that elect to use other markets make their choices based on perceived reach, personal affiliations, and whatnot.

Strategically, markets are silos:

  • Want reviews? The market has its own!
  • Want information about related apps? The market has its own!
  • Want ways to contact support from the app developer? The market has its own!

Even the data that the developer supplies tends to be a one-way channel. Paraphrasing the old US “Roach Motel” tagline, data goes in, but it never comes out. At most, a market might have an RSS feed of newly arrived app titles and descriptions. Everything else is locked up behind terms of service, even for the data the developers supplied themselves. This is not unusual — most “Web 2.0” sites operate on this same “digital sharecropper” model — but it is unfortunate and, hopefully, avoidable.

Overall, this situation is untenable for the long haul for all parties, developers in particular.

A week and change ago, I attended Droidcon UK, a truly frabjous event. The first day was set up to be barcamp-style, with presentation pitches and the like. Since I had gotten an email that morning from a developer ranting about market-related stuff, I figured I’d pitch a topic on some of the issues I raised earlier in this post. I thought that this would be in a small room and I’d get a dozen or so people, and we could just brainstorm a bit.

Well…that’s not quite how it went down.

It turns out that the facility was aimed more for the Friday round of classic conference presentations, with a large projection screen and theatre-style seating. The room I was to use for my barcamp session could have held hundreds of developers of people, and I was to speak at a podium, flanked by a fairly massive screen. And, as it turned out, I had a fairly substantial audience, perhaps a 100, maybe a bit less.

Not quite what I had in mind.

So, that morning, I quickly tossed together a set of slides, and rather than having the brainstorming session I wanted, I threw out some ideas and got a bit of feedback. A bunch of people found my ideas fascinating and wanted to subscribe to my newsletter, so I set up a Google Group to try to continue the discussions.

Meanwhile, I’ve been trying to figure out how I got myself into this fine mess… :-)

I will be writing up some of my thoughts here as blog posts over the next several weeks, extending on the specific points in those slides. Meanwhile, in and amongst 18,000 other to-do items, I will try to put together some strawman data formats and technology to allow developers to publish their own app feed, which these markets can tap into, to try to minimize redundancy on several levels. If you find any of these ideas to be fascinating, join the aforementioned Google Group and chime in.

The AndroidX Tech site contains source code, transitive dependency details, and much more for Google’s androidx artifacts!