GNOME as a platform

In the previous post, I discussed platforms and their relationship to “projects” and “products”. While I was writing it, I had in mind an old blog post from Havoc. It took me a while to find it…can’t believe it’s been 6 years. Anyways, you should go and read that post before continuing. Here’s the link again.

What I’d like to argue – and most of you probably agree – is that GNOME shouldn’t explicitly take the “building block” or “platform” approach. There are multiple reasons for this, but the strongest one I think is that if we focus just on making a Free Software desktop that doesn’t suck, by side effect we will produce a platform. And in fact – that’s exactly what has happened. Think NetworkManager for example. Getting a network experience (particularly with wireless) that was remotely competitive with Windows XP required us to invent a new networking system.

If we just said “we’re a bucket of parts”, and not the ones actually out in front trying to make a networking user interface, basically there would be no obvious driver for a networking API (besides toys/tests), so it wouldn’t be tested, and in practice it wouldn’t really work. Or at least, there would be some immense lag between some third party engineer telling us problems with the API and getting them fixed.

Will third parties take the code and do things with it? Of course. And that’s allowed by the fact that GNOME is Free Software, and we want to “support” that for some values of “support”.

One thing bears mentioning – of course GNOME should be a platform for application authors. That’s in fact an important part of our place in the ecosystem. But as far as being a collection of parts versus something more, here’s the way I think of it: if you can walk up to a computer and say “Oh that’s running GNOME”, i.e. we have a coherent design and visual identity, then we’re succeeding.

GNOME is not unique in being an “end-user” focused Free Software project debating the platform versus project/product issue. See also the Mozilla platform versus Firefox. The role and relationship of those two has been a subject of (sometimes very contentious) debate in that community. And that’s fine – debating the line is good. As long as you keep producing something that doesn’t suck while debating =)


  1. I generally agree with you that GNOME should not -just- be a bucket of parts, I do believe that any project which is suffeciently complex should be structured as a bucket of parts however — effectively GNOME should be the first downstream sink of those bucket of parts (meaning that there would be a GNOME-parts project and a GNOME-desktop project).

  2. But while you make this awesome thing from which a platform appears, you do need to support the people who use said platform to develop things that are not the awesome thing. Not sure how you do that while continuing to focus on the awesome thing.

  3. I, on the contrary, argue that Gnome SHOULD be a platform first, since it already acts like one.

    When I started running Cinnamon, I finally saw the potential of Gnome-Shell start to materialize – because it was made by people who cared so much about users they actually LISTENED to their feedback, as opposed to working like a cabal in IRC covens and disregarding any opinion that did not come from (or, but that difference is academic at best)

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s