Battling Misconceptions: What is KDE?
A while ago, Justin and I received an email from Dr Tony Young who recently wrote for LXer about his experiences moving from KDE 3.5 to a Plasma Desktop and our Platform 4 apps. If you want to understand some of the frustrations people have experienced in making the switch, then those articles are well worth a read.
However, Tony’s email was prompted by a discussion on LXer about why KDE does not listen to its users (that was the point of view of many of the people making comments, you can make your own judgement about whether or not that is a fair criticism). On the suggestion of an LXer reader, Tony brought his concerns to us. We in turn discussed them with some members of KDE’s marketing team and Tony and I had an exchange of emails, now published by LXer.
What is KDE?
Now, Tony is an experienced user of free software as well as a scientist and – from his emails – clearly an intelligent and inquisitive man. So it was a real surprise to me that he and other LXer readers were so ignorant about how KDE works. I do not mean ‘ignorant’ in any kind of offensive sense here, merely that both Tony and the other LXer readers really did not seem to understand how KDE operates or who we are. Here are a few questions (paraphrased) that really took my by surprise:
- Who controls KDE?
- Who funds KDE?
- Can we contact KDE?
These show a few things to me. First, some (many?) people think of us as having a hierarchy like a company, as if we have a leader or set of leaders who tell everyone else what to do. Maybe these are the people who pay us and if it is possible to get in contact with those leaders then they might be persuaded to redirect the efforts of all the code monkeys.
Of course, KDE is not like that. We don’t have leaders. We have prominent community members, but they tend to operate within their own areas of expertise. I would not tell Aaron how to develop Plasma, or Frank how to design ownCloud and neither would they tell me how to set the editorial direction of the Dot. Even within teams, those who do the work tend to decide or a consensus is reached. If you wanted to change our software significantly then you would have to contact key people in every team and convince them of your vision. Then, when some of them got busy and new contributors came in, you would have to contact them and convince them too.
Challenges in talking to ‘KDE’
This of course raises some interesting questions. How do we keep things consistent? Well, to some extent we do not. Our applications can look different and handle things in different ways, but thanks to building everything on the KDE Platform, which provides easy ways of coding common tasks such as open and save dialogues, there is a strong degree of consistency in KDE software.
There’s also the question of who you need to talk to if, as an outsider, you need a particular feature or want to discuss something. Tony came to Justin and I more or less by chance, probably because we are mentioned on the MWG page. He could just have easily have contacted the KDE Promo team, but sometimes people prefer a named contact and keeping discussions small can lead to quicker results and a more personal feel to the communication.
It is probably similar for a company wanting to work with KDE. Say you want to use some of our software in your device, who do you contact? Strictly, you do not need to talk to anyone as long as you respect the free license terms of the software, but talking – particularly if you can identify or offer improvements – is good. Probably our press contacts and the e.V. board are often points of contact in this case. Perhaps we should think about having more – so that for example there is an ‘external contact’ for each large team in KDE.
For people unhappy with Plasma and Platform 4 software in general, the way to influence things is to engage. Turn up in mailing lists and offer assistance. Or even just get stuck in and code things your way – that is how things like Rekonq came about. Write the code and see if people agree with your way of doing things.
Weird is Good
KDE and free software communities in general are quite odd things. Lots of volunteers working mostly on things because they interest them, but still managing to communicate enough to create something that is consistent and better than most of the big budget proprietary competition.
KDE is a frustrating, infuriating, remarkable inspiring mass of individuals that find enough of a common goal and philosophy to unite them to create wonderful things. It shouldn’t work, but time and again, it does.
And, if the current set of KDE software doesn’t work for you, in the worst case you are free to do your own thing.
Hi,
for me KDE is clearly lacking a transparent structure where people can involve. For most parts you will tell me to fill bug reports but as often as those get discarded and given the non intuitive bugs.kde.org and the vastly outdated and self contradicting wiki meant for developers only I see why there is no hierarchy. Let’s face a simple example where KDE could have done better from the beginning: KDEPIM based completely on akonadi is not released but was promised to included in 4.5.0. At the moment I don’t expect it will included in KDE4.6.2 and from user perspective there are no advantages.
On the other hand I wished KDE would drop that SC nonsense and declare that it is a community project not driven by paid developers which promote N900 running KDE.
Best regards,
Otto Kranz
Ok, so we have problems – that is clear. How d we solve them. What is lacking at, for example, kde.org/community/getinvolved/ ? There is a web sprint coming up so maybe we can work on those pages.
Re bug reporting – well bugs do not get discarded, sometimes they do not get fixed because there are too many bugs and too few people. And yeah, bug tracking tools are horrible, but we do not have something better.
Akonadi is already bringing more backends for different services, but of course the delay is not good. Though it is better than releasing something early that breaks your email, no?
‘SC’ is almost entirely dropped in official KDE communication now.
I hereby declare that KDE is a community project (or rather, a community) that is not driven by paid developers to promote KDE software running on N900.
Great responses to Tony and wonderful blog post. I confess that over the last month or so a couple of events left me somewhat dispirited as a member of the KDE community. But this was really quite a pleasant and uplifting post that injected a bit of that happy, chaotic, creative spirit that has come to define being a part of the community.
Thanks!
Thanks 🙂 Yeah, it is easy to get dispirited but every now and again I find the time to stop and take a look at what is good in KDE and it always helps
FWIW, I think KDE developers do a great job at listening to users. Of course it’s not possible to cater to every single user, so there will always be people unhappy. That’s normal. There will be other options for them.
But KDE does a good job at listening to wishes. Anyone who was a KDE user in 3.5 times, and kept being one during the 4.0 transition and afterwards, until today, watching the new features being developed, the bugs getting fixed, the missing functionality being restored… can see this is true. As much as the developers possibly can, of course.
Since 4.0 I’ve seen a lot of criticism, sure. But most of it were basically rants, regarding:
– Some feature was gone (and very probably came back later, as announced).
– Something now was done in a different way (probably better for most people, and configurable to the old way, anyway).
– There were unstabilities (to be expected with such a rewrite, with such great goals).
This kind of rants did nothing to help, and much to destroy.
Sadly, as you said, happy users don’t complain, so unhappy users complaining make a lot more noise.
Anyway, this is getting long 😛
I just wanted to say the first line: in my opinion, KDE does listen, and does more than could be paid for.
Thank you all.
It’s nice to hear from happy people 🙂 Clearly 4.0 had issues – we admitted so much at the time – and of course there is still more to do. There always is 😉
The spirit is broken because there is no open discussion culture anymore like in – say 2003 – but control of language like in a dictatorship. No open debate about KDE4 was permitted and announcements did not hold expectations. That is a certain annoying governance style. I can’t see how criticism, even ranting harms a project, because it fires up the do-cracy. For instance renaming KDE “Plasma desktop” is just a stupid marketing move and press releases to this end are simply annoying. An open debate is needed how to improve stability and reliability and open debate about shortcomings. Look how the quality.kde.org project was wrecked by aggressive denial. When all criticism and user experience is described as FUD that certainly fails the popper criteria. This is not a war, so why communicate like in military propaganda? How can someone trust an announcement when past announcements turned out to be lies? Why can’t we have a true and fair view?
Ok, specifics please:
– What has changed to make you think there is no open discussion any more? We’ve gained the forums, the Dot has opne comments, most mailing lists are open, Bugzilla is.
– What was KDE is not renamed as Plasma Desktop, only one of the new workspaces is called that
– quality.kde.org (the EBN) still exists, no?
– Which announcements turned out to be lies? How so?
It’s far too easy to make vague sweeping statements, but they do nothing to improve matters because we have no idea what you are talking about.
First of all, no one has to justify himself when he says KDE4 is not ready for him.
– stability, reliability: developers do not seek hard figures but basically brush those off who say they are disappointed. It would be possible for KDE to implement oops style feedback mechanism but no one is interested in finding out what the actual situation looks like for their users because the official doctrine is that disappointed users are trolls.
– Because KDE traditionally does not feel responsible for the build process, most distributions provide delayed outdated KDE versions after a release has been made, fixes are not sent upstream, so with the next KDE release things would break again. When KDE 4.6 is out there is really no point in using KDE 4.4 with all the bugs, I am not speaking of Debian stable users.
– slowness in connection with certain common graphic cards and drivers still unresolved. This makes the Plasma desktop unusable for *many* users and you are not even told by KDE in advance that you may face these problems when using this or that card. They can’t tell you what graphic card to buy to run KDE or where it was tested.
– While KDE3 was suitable for desktop migration of “public administration” KDE4 is not simply because you still cannot contain risks that it breaks. That the Desktop regularly crashes is usually a killer.
– The Desktop is now officially called “Plasma”
– KDE development admits no mistakes in the development process of KDE4.
– Some parts of KDE development do not scale well, for instance the translation process, a low entry aspect of KDE development. You don’t get releases with full translations for major languages, they might get provided for the secondary releases. All criticism with regard to lack of developer capacity gets countered by an alleged massive growth of developers interested in the new KDE4 technologies.
– EBN hints which parts of KDE are not well maintained. Ironically you can’t find that EBN results in a decline of issues. Even a developer-centric application like KDevelop is full of unresolved issues.
– Lie: KDE4 was supposed to be cross-platform. In reality KDE for Win and Mac is “experimental” software which also lags behind the main release.
Open discussion means: you focus on problems, discuss bottlenecks and problems with the aim to find a way to jointly resolve them. There is really no point in praising what works well. When no one admits there are challenges, it gets difficult. Challenges attract Knights willing to slay the dragon.
Yes, you can criticise without justification. However, if you want to actually change things you need to explain what you mean and – ideally – help to fix them. So thanks for coming back with more comments.
So:
– Stability. Does the automated crash handler not do enough here? There is also bugzilla that invites reports of problems or suggestions for improvements. Developers do brush of those who cannot be bothered to report issues in bugzilla.
– Builds. Sorry, we really just do not have the people to make binaries for every distribution. Many distros do however make them available, if unofficially, but that is their choice. What can KDE do here it improve things?
– Some kind of graphics hardware friendliness to KDE software matrix would indeed be useful. Want to start one? Unfortunately, the KWin devs do not have every graphics card or enough computers to run them in to test every combination. Or the time.
– There are, I believe, public deployments of Plasma Desktop so it seems some public administrators disagree with you (I don’t recall the example at the moment, it is possible I am wrong)
– One particular workspace is now called ‘Plasma Desktop’. So wait: the name of the software stops you fro using it?
– Mistakes? Of course there were mistakes. SC 4.0 became default in distros too early (or at all). The ZUI for Plasma Activities never really worked and was dropped. The panel menus have been changed and tweaked. KNetworkManager wasn’t so great and was swapped for a Plasma Widget. Seriously, I don’t know how many thousands of bugs were fixed since SC 4.0, but every single one is an acknowledge mistake.
– Translations. KDE is fairly impressive here imho, but maybe I get a biased view as an English speaker – I’m surprised though that En-GB is always immediately available. How do you suggest we attract more translators, which seems to be the problem?
– EBN – I don’t really know enough about how it works or understand your point to argue here. But all our software is full of unresolved issues, or unrealised enhancements. And not just our software…
– Win and Mac for sure lag and have more bugs/stability issues, but the Platform 4 software is a hell of a lot more cross platform than KDE 3 ever was. Statement: KDE software will tun of Windows and Mac. Reality: it does.
If you really think there is not open discussion of problems, read the mailing lists. If you want to do something about it, join them. If you want to raise problems on the outside, that’s fine, but unless your shopping list of issues inspires more people to come and fix them, it will have no effect.
I have no problem with a negative development or immaturity. But it is really important to openly talk about shortcomings and challenges.
“Because KDE traditionally does not feel responsible for the build process, most distributions provide delayed outdated KDE versions after a release has been made, fixes are not sent upstream, so with the next KDE release things would break again.”
I am not sure I got this one.
All release announcement I can remember included information about available packages for numerous distributions.
For example take 4.6.1: http://kde.org/info/4.6.1.php#binary
As for fixes not ending upstream: do you have the impression that this is caused by downstream not sending them upstream or umstream rejecting them?