At the first Desktop Summit in 2009 KDE’s aKademy and GNOME’s GUADEC conferences were just co-located. I cannot remember having had any interaction with non-GNOME folks (but I wasn’t around for the complete conference).
In 2011 it wasn’t just co-located but mixed tracks. Though I attended some KDE talks and had to realize that most were boring to me simply because I have nothing to do with that software stack I still had great and interesting conversations with some KDElers.
Mission
The Summit website states “The goal of this event was to share ideas and further collaboration between the two communities.” The section “Goals for this year” also lists “Collaborate on desktop software projects”. And everybody has different experience and opinions whether this actually happened or not.
Criticism
In both communities there are mixed feelings whether the concept of a Desktop Summit makes sense. I know that some GNOMErs expressed their opinion that Desktop Summits slow down GNOME development. You can draw two conclusions from this: Either to not repeat Desktop Summit because of that. Or to fix this for the next Desktop Summit.
While the technical stacks are mostly different, there is room for collaboration in less technical areas such as release management, bugsquads or documentation efforts – even if it’s only about exchanging experience or best practices. This also applies for some technical areas that are shared in our stacks via freedesktop.org, e.g. parts of the accessibility framework.
So?
If the fear is that planning and development in each environment slows down under the collaboration banner, and that a GUADEC-only conference is more helpful in pushing things forward in GNOME, why not have it both? Have one or two days of collaboration related sessions only and nothing GNOME or KDE (or LXDE) specific, followed by two co-located conferences that only have environment-specific sessions. Does that make sense?
Disclaimer
I left out the financial part on purpose as I have no clue about that, however I know that it has influence on the decision whether to continue Desktop Summits or not.
I just do not see the point in having a Desktop Summit instead of a GUADEC. I prefer a GUADEC. Working together with KDE is good, but I see not having a GUADEC as a no-go. Prefer every year a GUADEC and then possibly a Desktop Summit for those who want it.
I also feel GUADEC should happen every year and then we should host a special desktop summit (could be just after since most devs will be there already) for people working on cross desktop stuff. Those 2 conferences serve different purpose which cannot replace each others.
I’d like to register a different opinion. I feel it is very valuable to have cross-pollination between the desktops. And meeting people from other camps reduces tension and mistrust. It’s unfortunate that LXDE or XFCE weren’t present, and I think that instead of breaking up again, more effort should be spent to get those to come.
Inge: There is 0 tension/misrust among the GNOME developers that I know towards KDE. I think meeting people from other projects is really nice, but I fail to see why GUADEC has to be given up to to have that. I don’t see any of that changing much though; the goal of GUADEC is meet/plan/party. Desktop Summit has a totally different goal and atmosphere.
[WORDPRESS HASHCASH] The poster sent us ‘0 which is not a hashcash value.
I agree with Inge. Thinking that cross-desktop stuff is for other people is just negating the goal of all that. I’ve seen people realising that they were sharing 80% code, and re-implementing the exact same stuff one for GNOME, one for KDE. The e17 talk was also interesting, as it showed that some choices done in FreeDesktop may harm other desktops which don’t have the same requirements. You can’t realize that if you don’t meet people, if you don’t mix them. The parties also help to have more cross-desktop junctions, so people are more aware of who to contact, and easily find who is their counterpart in the other comunity.
This may be seen as slowing down GNOME, or seen as more bridges built between communities, which will help making more and more common code so people don’t waste their time. This may slow us down now, but I’m sure it will be really helpful in the long run.
My argument is more that I rather thave it done without having to give up GUADEC (especially so often).
Saying there is 0 mistrust or anything like that between K and G is crazy. There is. Not with everyone, but unfortunately there are a few influential people (I’d rather call them idiots) who are blocking cooperation and thereby hurting the Free Desktop quite a bit. MeeGo failed in big part due to a few idiots refusing to use any lib with a K in it (while all Qt/KDE devs were happily using Tracker, Gstreamer and other ‘G’ tech). And tell me, how likely do you think it is that GNOME gets a dependency on Qt? There’s no technical reason it can’t, just like many KDE apps require GTK and GNOME libs to work, but it just won’t happen. There is plenty of duplication still going on and I frequently hear horror stories from both sides about especially lower-stack developers ignoring or happily hurting ‘the other side’. And let me tell you, as long as that goes on, I won’t be supporting separate GUADEC’s and Akademy’s.
The thing I hate most about all this is that for me, not being a developer, it’s almost entirely hear-say. So I don’t want to mention any names or specific things as I’m sure I’ve only heard one side in most cases. But that’s just another reason why the two communities have to interact more: figure out WHY people act like they do. And you don’t find that out over a network, you need a face to face for that.
Oh and yes, I do believe Linux didn’t succeed on the desktop because of the people who ‘hate KDE’ and ‘hate GNOME’. The two desktop teams have different UI philosophies but the underlying technologies could have been shared which would have saved a huge amount of resources and entirely wasteful and silly duplication (in many cases even pointless replication of mistakes because the dev’s didn’t even LOOK at what the ‘other side’ has done).
I often jokingly say that GNOME Shell should’ve been build on KDE’s Plasma libraries (as they were explicitly designed for an usecase like that), but honestly, I bet it probably would have shaved off 1 year of development and resulted in a better GNOME Shell as well as a better Plasma Desktop.