Driving desktop forward

You’re right, the “14.04” Ubuntu Software Center did indeed have paid app support. It wasn’t a tremendous success for a number of reasons.

  • Minimum payments per application was too high (~4 GBP)
  • Developers had to build debs - which are hard, especially for people who are just trying to bring their application to Linux, and don’t want / need to learn all the intricacies of debian packaging
  • All package updates were gated by a (very small, and massively overloaded) team of human reviewers, which led to delays, and constant back-and-forth between review team and developers to refine the debian packaging
  • Software needed to be re-built for each release of Ubuntu. So if a developer depended on a library which changed name or version in a new release, they had to rev the packaging going through all the hoops again, above
  • There was no option for In App Purchases, Downloadable Content, Pay What You Want or crypto payment

Developers found it hard to publish software, they had to set prices people wouldn’t pay, and were then signed up to maintain multiple revisions of their application for different Ubuntu releases. For these, and a bunch of other reasons, the USC paid apps support was not a success.

Cut forward to Ubuntu Phone. We had a paid app store model there too. A small number of applications were published as “Click” packages (the spiritual forerunner of Snap packages) but had many of the same issues as the desktop store. In addition, the market was very small, given the thousands of devices sold, limited markets, and new development platform, it wasn’t a success.

So now lets look at today. We have solved many of the issues above. Making a snap is easier than making debs, they are installable on multiple releases, and most of the time require no initial human review, and no review for each upload. These are compelling advantages over the old desktop store model, moving us towards being able to have paid apps in the Snap Store.

There are still some issues however. Fundamentally, the majority of software these days isn’t actually sold in one-off purchases. Users are typically more inclined to buy subscription services. Think of Spotify, Slack, Skype - some of the most popular applications in the store, none are gated by a one-time payment. The obvious exception to this is games, which often have a single payment, with optional DLC or IAP. Games have plenty of places to be seen though, with Steam, GoG, Itch, Epic and others.

While there is some popularity among indie developers for PWYW (Pay What You Want), it’s by no means universal, and I’ve seen no evidence it’s tremendously successful except for a few specific examples. There are indeed other distributions which have PWYW models for their developers, and perhaps due to the small audience, or lack of wide app availability, despite the press, they’re not super financially successful.

Crypto currencies are an option for uber nerds, but not for the average joe, typically.

There’s also the overhead of running the thing, dealing with refunds, complaints, failed purchases etc. No online store seems to do this particularly well, and we’d need to really do it in a way that’s profitable. There’s no point us having paid apps in the store, if only a tiny percentage of them actually result in completes sales, which generates revenue for the developer, and us.

There’s plenty of other issues along the way, making this hard. I certainly haven’t covered them all.

Maybe one day we will implement one or some ways to do it, I’d certainly like that to be the case, and do personally add it to the internal wishlist/backlog when I get a chance. Implementing one-time purchase, PWYW and crpto seems like a panacea, but it’s a lot of work, for potentially little profit. So we’ll have to wait until it’s determined to be worthwhile, I think.

3 Likes