This post came up totally random. We were talking 3-4 days ago with a Gnome dev, and we were sharing some concerns about the progress of Clutter Toolkit. Clutter is vital to Gnome Project, since Shell is written on it. Lately the development progress of the library isn’t pretty good.
So I thought to send an email to Emmanuele Bassi (EBassi) and ask him couple of things. At the beginning I didn’t ask him for publishing anything, it was just a private email. Next, I considered what he said quite important, so I asked him if he can give some more info and re-print those.
A little background in case you don’t know. EBassi is the author and main developer of Clutter Toolkit. EBassi joined few months ago Endless Mobile. In Endless Mobile they building an OS for mobile devices based on GNOME. They haven’t publish any more info yet.
I go directly to my questions and his answers.
1. Since you left for Endless Mobile, Clutter Development has been slowed down. How this will impact GNOME?
I don’t think that’s a fair assessment; in all fairness, I probably have committed more to Clutter while at Endless than when I was working at Mozilla. as I said at GUADEC, though, I think Clutter 1.x is pretty much “done”.
While I might add some API that I’d like to experiment and iterate on, I don’t foresee any further changes (radical or otherwise) for the 1.x API series. right now.
I’m really working in integrating Clutter (under that name, or using a different namespace) inside GTK, something that I hope to have ready as a preview by the time January rolls around.
In the meantime, I’m still fixing bugs in Clutter, and reviewing patches coming from various contributors, including the GNOME Shell developers.
2. What are you building in Endless? In your speech in GUADEC you talked about something that it will use Shell. But since you don’t do Clutter anymore that doesn’t make sense. Furthermore you (Endless) ask Java/Gtk/Python developers.
We are building an operating system based on the GNOME stack, including the Shell and various core applications. We modified various things, and shipping our own applications and SDK, but it’s still very much GNOME (a mix of 3.8 and 3.10, depending on stability).
Prior knowledge of both Python and Java is not rejected, though.
3. Concerns about Clutter, and Google Students involvement on this project
Clutter is, sadly, not a great place to work on in terms of Google summer of code or code it; it’s a code base with baggage, which is currently the shell base toolkit.
It deals with parts of the platform that are pretty complex (graphics drivers, GL, etc.) and it is not an API that strives to add complexity to itself.
It’s bare-boned, simple, and I’m in the process of deprecating everything that is not a building block for something else. I’m not surprised at all that a student has issues working on it. Working *with* it is a different matter.
I usually am responsive on IRC and on Bugzilla, and sometimes even on the mailing list; if there are issues, though, I’d like to know.
4. Clutter will be embedded to Gtk 3.12? GTK Team will maintain Clutter?
3.12 is completely out of the question: we already have a pretty clear picture of the features of 3.12 at the moment. 3.14 is probably more realistic, but it’ll need time to get reviewed and integrated, so I’m not going to give timelines.
I’m going to maintain the GTK scene graph API, but part of the integration story is that I won’t be the only one working on it any more, yes.
What I think should be done with Clutter is pretty much matter of public record; as I said, I even had a talk about it at GUADEC 2013, and long discussions before and after that.
You can watch EBassi talk in GUADEC
You can watch all talks in GUADEC ;)