21 November, 2008

talking about Nokia and KDAB

Yesterday I gave a talk at a seminar at Mediaplaza about the cooperation between a community and commercial ventures. To prepare, I conducted an interesting interview with Till Adams from KDAB, contacted Qt Software and read up on some stuff. I must say I enjoyed taking time to really prepare this talk and dive a bit into the subject.

The way KDAB and Nokia/Qt Software work with us is very interesting - they strongly believe in close collaboration, and act accordingly. Great examples of good FOSS citizenship.

I did have a hard time complying to the 30 minutes schedule. I normally talk about KDE for 45 minutes. Now I had to introduce KDE in <15 min to this rather 'green' audience, and then talk about the cooperation between company/community. Interesting challenge, but I think I did a decent job at it.

The amazing pictures created by the KDE India community were incredibly tempting, but I had to resist. Talking about the pillars themselves would take too much time. Would've been too technical as well, I suppose. I did mention Nepomuk and Phonon, but that's because they are cool examples of cooperation of KDE and other organisations. I will use those for my future slides, though, as they're just stunning.

BTW the talk I gave last week at Linuxworld in Utrecht was nice, as I got some intestesting feedback. Still have to think more about that, and integrate it in my future talks. One good point somebody made: yeah, KDE created this great architecture to be able to move faster & incorporate more user input. But currently, many users don't feel KDE is listening. Others agreed with him, so this isn't something we can cast aside. Of course we know why this is - we already KNOW (mostly) where the issues are, and we ARE listening - we're just not always replying to their questions, bugreports and comments...

The distributions shiping stuff we don't think is fully ready for the enduser (everything after KDE 3.5.x and before 4.2) didn't help us much either. They had good reasons for that, of course. Many users wanted to see KDE 4.0, and distributions had to spend a lot of resources on shipping both 3.5 and 4. The unfortunate effect was that many users lost precious features compared to their previous desktop, and the switch didn't go smooth either. But seeing how much time for example the Plasma team has spend on doing what the users requested from them (delaying the cool stuff they'd probably rather be doing) makes clear this is an issue with how we are percieved. Not how we are.

I wouldn't know how to solve this in our communication. Maybe we should just let KDE 4.2 talk for itself...

Anyway. I'll try to get the presentation in KDE SVN as soon as I can. Just in case anybody wants to see it asap, send me an email... firstnamelastname at kde.org should do the trick.


  1. Please reconsider your general statement "the distributions shipping stuff that is not ready (everything >3.5.9 and <4.2) didn't help us much either".

    Most distributions have taken great care to ship KDE 3 and KDE 4 releases simultaneously to fulfil user demand and provide essential testing/new developer for KDE 4.

    Saying that making KDE 4 available on these terms "didn't help", as an excuse for user dissatisfaction during the 3->4 transition, is a gross oversimplification of how Free Software development lives and breathes, and is pretty ungrateful to those who contribute to KDE's lifecycle by distributing it.

  2. @Bille: you're right, I've clarified my statement.

  3. The big problem was lack of communication in advance of how you perceived KDE 4.0 and even 4.1 not to be "ready for the end user". Saying it after the release or even a few days before the release is too late, distributions have already made their decisions at that point! You need to tell us months beforehand (in clear terms) that a release is not intended to be distributed. We have schedules too... Realizing "oops, this is not as good as we thought" at the moment of the release does not help us at all, we cannot just back out all the work we did on packaging the current version.

    For Fedora 9, our decision to ship a 4.0.x with as many fixes as possible in the release and then an update to 4.1.x as soon as possible was made in the planning phase for Fedora 9, which was around November 2007. Back then, the word (at least the information we were aware of - it may have been a communications issue) was that 4.0 would be the release for early adopters and 4.1 for end users. The warnings that 4.1 is targeted for early adopters only were posted in the announcements of KDE 4.1, around July 2008. That was about 8 months too late!

    Another issue is that not all developers and marketers agree on these things, some say 4.1 is perfectly fine, others like you say 4.2 will be "the" release. (Consider that one of the people involved with KDE PR is also the primary developer of a major distribution which just released with KDE 4.1.2. But there are more people thinking that 4.1 is shippable. For the record, this obviously includes us Fedora packagers.) Without a clear message, whom are we supposed to believe?

    In addition, the communication to end users is also lacking, there are a lot of newsitems, blog posts etc. about the great new features of KDE 4, almost none of them saying that 4.0 and 4.1 "aren't intended for the end user" (which not everyone even agrees on, but I already mentioned that). So users ask us to ship 4.x because they read about all the great new features, and then they're really disappointed with what they get.

    And then we try to make them happy by backporting some of the obvious missing features (e.g. panel autohide) and we get flamed for that too. No matter what we do, we're always the bad guys. :-/

    And finally, in the particular case of Fedora, we are primarily a distribution for early adopters (just read our objectives) which is exactly the target audience 4.1 is described as being for.

  4. @kevin:
    For 4.0 we sure didn't communicate clearly enough, imho the release announcement was esp lacking in that regard. You're right in that the messaging about 4.1 was mixed - simply because the feelings and opinions of those involved were mixed. 4.1 would be a bad choice for somebody used to the power and configurability of KDE 3.5. But it would be a reasonable choice for a first-time linux user or an ex-gnome/XFCE/etc user, as it offers enough for them. At least that's how I see it.

    BTW if a distribution is wondering about whether to ship or not to - they could ask. On the KDE-promo mailinglist, or on a blog.

    And lastly, in case of Fedora - you guys want to be bleeding edge, with the intention to advance the Free Desktop. Shipping 4.0 is, even with such a goal, quite something, but shipping 4.1 totally made sense imho.


Say something smart and be polite please!