Andre Klapper was a busy man at the Gnome Asia summit. Maybe it was the aura of the hackfest that got to him. Or maybe it was the spicy Indian food. Whatever it might be, Andre was unstoppable, giving one talk after another. We caught him off stage to do some more, um, talking.
Mayank Sharma: So what's your role in the MeeGo project?
Andre Klapper: In MeeGo I am part of the error management team, and before in Maemo work I was the bugmaster. Nokia at some point set up a public bugtracker, bugzilla, while having an internal bugtracker too without having a plan on how to sync them or how to actually manage the user feedback they get on the public one. So I am a bit of a proxy in between all the good and bad sides.
If you want ice cream, then file good bug reports.
In MeeGo now, upstream MeeGo is completely 100% open both the code and the processes, though this is a bit of a work in progress, of course because the stake holders in the companies that are involved quite often kind of keep their corporate culture without realizing that they should discuss things in the public.
Currently in error management we have some triage teams depending on the UX, like handsets or netbooks or the core of MeeGo which is shared among all of them. We do usual triaging, helping developers and users to provide good information. I am looking at hopefully making the processes more transparent.
MS: In that respect do you think Nokia backing out of MeeGo had any positive effect on the project?
AK: Yes, it has. But as far as I understand they don't completely back out, it's more like as far as I interpret the press release, of course Windows Phone 7 becomes the first platform but they still keep MeeGo going as a research and development project.
MS: Right, right....
AK: Yeah, and everybody is trying to interpret what that means.
So yeah, I think in some kind of way it has a positive effect because the MeeGo project and the community around it has to now invest way more effort into finding more stakeholders, like to put the entire project on a broader base when it comes to companies supporting it.
Plus I also have the feeling personally that after this announcement in February by Nokia, that people are way more open or direct in the discussions, for example, about architecture decisions. It feels a bit to me that earlier people were friendlier with each other but not really discussing the problems when it came to the direction of the project. Well now it seems like there's way more movement into making those decisions and discussing them in a very open way.
MS: You gave a presentation on Gnome translations. What's the status with that?
AK: Gnome translations in general are in a good state. We have a pretty good infrastructure that makes it quite easy for translators to contribute.
For example, if you really just translate you don't have to learn how to use the code-base like Git, or how to check out from the repository. Just go to l10n.gnome.org, download a file, translate it with whatever tool you like, upload it back again, and then the coordinator or the committer takes care of the technical stuff to get it into the repository.
Also when it comes to coverage we have always been having like 50-55 languages that have more than 80% of the user interface translated. So it's a quite strong base. Plus many more that have translated 50-80% of the user interface.
Maybe a little disappointing thing that I sometimes think is that this number is not growing because I think nearly all translators are volunteers. So you have some fluctuations and maintainers AWOL and sometimes it's a bit hard for new contributors to find a way on how to take over, or how to get it going again.
(from left) Lionel Dricot, Andre Klapper, Frederic Peters
MS: From a factual point of view, does Gnome 3 make any part of the translations easier?
AK: I am wondering if anything in specific becomes easier...I don't think so.
One change we had a couple of months ago was when it comes to the amount of translations on the web platform that we use to manage the translations in Gnome. We decided to have a reduced set of strings that don't any more include the schema. Strings like GSettings, dconf, gconf, which normal users never see but sometimes can take upto 20% or so in a settings intensive application. Those are really like low priority.
So translators were always asking "how can I set priorities" and this is just another help for them to concentrate on real user interface instead of wasting too much time on settings strings you basically never see.
MS: How was the hackfest? What did you break?
AK: It was fun as usual but it was very intensive. It's like you spend the entire day with your friends, or the people who work on Gnome, and you also normally spend the evenings in the hotel rooms together...it's very intensive.
But I am happy with the output, and I see it in a good way. We got quite some stuff achieved and if you are physically together it's also easy to help each other. Just looking over the shoulder and calling someone and asking "How do I do this or that".
MS: So in pretty good shape for the release?
AK: It's been tough. The last few weeks have been pretty tough especially when it comes to the time invested in it. I think the last couple of weeks I have spent a lot of time on Gnome, and I am fairly confident that we've got a fairly good product.