Tuesday, March 4. 2008The Genesis of a Plasma patchAt one day, there was a patch. The patch was small and in early development, and it was posted to the Plasma review board. A couple of minutes later, the patch was rejected. Many reasons were given: The patch didn't have the right intention, it would cause bitrod, and it broke the coding style. Not a big thing, this happens every day in a Free Software Environment. Neither something to get excited about nor something to loose hair over it. Thats the end of the story? Not quite, in the bible it took 7 days as well, right?. So lets look at it a bit more closely. Day one: The review said that the patch contained a coding style error. Of course the review didn't mention where the coding style error was, or even better, on how to fix that (didn't the review board have this awesome patch annotation feature?). It is a quite obvious way to motivate a possible new contributor by giving him constructive feedback on how to improve his patch. Not a big deal? Not a big deal. However, as it was mentioned as the last point in a long rambling did make it smell a little bit like "Even if you manage to fix all the flaws mentioned before, have fun figuring out how you make me like the arrangement of spaces in your patch!". There was Day, and there was night. 2nd. Day. Now.. where was I? Ah right, light!.. eh Plasma! So, the patch had its intention to hide the Plasma toolbox from the KDE 4.0.x desktop. The KDE 4.0 desktop was released in January in its first stable release, intended for early adopters. It was expressed quite clear that it is not finished yet ,and the adequate set of features will come with KDE 4.1. So what would be more obvious to highlight the pure beauty and new shininess of KDE 4.1 by introducing a feature-complete plasma toolbox with KDE 4.1? I spend my second day thinking about this. There was Night, and there was Day. 3rd Day. Lets enjoy some quote from the review: you can, of course, put this patch in your downstream builds. you'll be doing the project a disservice, however. or are we not actually developing software in the open anymore but only delivering a final product and therefore completely avoiding the entire feedback cycle? A couple of things to notice: The patch wasn't posted by a distribution, it was posted by a KDE developer. It does sound strange that a normal KDE developer is suddenly downstream and only Plasma developers are upstream, but this is how it was communicated. Ah, you're saying that I'm ironic and that I know that in my other live I'm also a distro guy? Fair enough, I should have known that. Obviously I cannot be two persons at the same time. And of course I'm doing a disservice to the project by posting a patch to upstream (!) instead of just patching it downstream. In addition my disservice to the project is trying to polish a stable code branch (nowhere this patch was mentioned to be intended for the KDE 4.1 development tree). I'll spend the rest of the day thinking about if we want to alienate our userbase even more by making them swallow down an unfinished product with rough edges and give them the feeling that whatever code branch, name it stable and released or unreleased and in development, they get the same amount of bleeding. Be free, and bleed freely. There was Night, and there was Day, 4th Day. So, it turns out, I wasn't the first person to notice that the toolbox has only two features:
While it is arguably true, that the "Add plasmoid" (it is actually called "Add widgets" btw) has some value (the same value the context menu on the background, the panel, the kickoff menu and probably a couple of other places have as well). Note that I also don't want to mention how KDE 3 users have managed to find the hidden context menu in the past to add an applet to their panel, not because it is unimportant, just that I only have one week and I'm already at day 4. So day four: I've asked my girlfriend what she thinks a widget is. And even though she lives with me for several years already, she couldn't describe it properly, nor explain me what she would expect to happen when she presses the button. Obviously its a translation thing, right? A "Widget" is the same as "I don't know the name of this thing" in German. Ah, so it adds a little "I don't know what its name is?" to "I don't know where" ? That makes perfect sense for this feature! I spent the rest of the day being depressed that I didn't take a picture of her face. I would have seen value in adding that picture to my desktop background.So there was Night, and there was Day. 5th Day. You're expecting me to follow up on the "Break your desktop" ? Okay, well.. its the general reaction you get when you let new people that pop up on a fair where you demo the desktop try it. They go like "uhh.. it turns white.. I broke it!" and jump away from the computer, being frightened that they made a change that they cannot undo. but there is undo, right? Well, only 3 levels of undo, to be honest: if you zoom out more than 3 times, you cannot zoom back in, because you can't actually press the button anymore. The only way to undo that is to delete your plasma configuration or, for the average user, to reinstall your computer. Oh, another thing you learn from being a distro guy: whenever something happens that average users don't expect and they can't undo: they either call support (be it the 24 hour child/mother support hotline or some other) or if they have a little bit more knowledge, reinstall their computer. Or they ask somewhere on the net and get the recommendation to rm -rf ~/.kde. I spent the rest of the day enjoying the other blog about the "break your desktop" button and the followup, which was particularly funny. Unfortunately I cannot link that one anymore because the blog posting was pulled from the net. you can still find it in the Google cache if you try really hard. In essence it said that whenever one has a certain issue with Plasma, one should contact the developer mailing list instead of publically blogging about it. Which I did, btw. Or didn't I? Did I only blog because somebody else blogged about my blog which I'm writing now? Damn, I wish I'd had a day left to think about that. So there was Night, and there was Day. 6th Day. I think after collecting the pieces, it is time to finish up. Or in my case, understand the big picture. Perhaps today "it also will grow into something that has clearer purpose". But wait, isn't that admitting that the previous days didn't have a clear purpose? The big picture for me is that somewhen we've apparently forgotten what it takes to be a good community: Be friendly, be constructive, allow different opinions to co-exist and clearly distinguish between technical issues and emotionally-overloaded sandbox-thinking. An additional bonus is to provide leadership by actually inspiring others and give them a clear vision, a vision that they can implement, that they can bugfix and polish and that they can understand well enough to share it with others. So there was Night, and there was Day. 7th Day. And today, I need a bit of rest from all the blogging. Doesn't it piss you off as well that you can ping as much as you like, be open and constructive for a discussion as you like and still only end up being able to write up your own thoughts up in a post without actually being able to establish any constructive mind-sharing? So, thats it. Or as downstream puts it: Have a lot of fun. Trackbacks
Trackback specific URI for this entry
No Trackbacks
Comments
Display comments as
(Linear | Threaded)
"..somewhen we've apparently forgotten what it takes to be a good community".
A good community doesn't air its dirty laundry on a public blog. Blogs are incredibly poor ways of resolving developer disputes. Discuss this on the Plasma mailing list or IRC, posting it on your personal blog will do nothing but inflame the issue.
I respect you a lot and greatly respect your work in the community but this blog post was a bad call. Not to say others weren't or your justification isn't there but "two wrongs don't make a right".
Blogs are great for some types of communication. This is not one of them.
Mike,
Thanks for your comment. I think it is important once in a while to step back and see how we interact with each other. This has nothing to do with dirty laundry - it is just a voice of a personal opinion - do not forget that.
Step back for a moment and see if you can see anything you read there in your daily interaction, and see if you can perhaps learn something. If you don't, you're doing everything all right, and you have every right to call this blog post to be totally misplaced.
I have learnt from your comment and I'm far from perfect myself. However you must know that the Plasma team will read this (as it is on the Planet) and that perhaps they will find this more antagonistic than an email or IRC conversation on the topic.
You may say this is about "us as a community" examining our interactions but in reality this post is about (or at the very least will seem to be about) telling the Plasma team that they need to treat contributers better. Right or wrong they may be, they don't deserve to have it communicated in this manner.
So what's your thesis here? Remove the toolbox from plasma would make the 4.0 series suitable for the average user? I don't think so... I think that the toolbox there is useful to make people familiar with it. Many people can't live with such a small change in the desktop paradigm (really can't understand why..) and I see that giving them an extra 6 months to familiarize could be useful.
I'm not talking about technical aspects here - I've finished that on Day one. (I could have mentioned how I dislike it because it doesn't have whatsthis-help or a tooltip or how it doesn't have a "go away" button, but I didn't, because those issue do not belong into a blog). I'm just focusing on the bigger picture, the social aspect behind it.
We are creatures of our habit and as such each try to reeducate users that may come with 20+n years of desktop-experience will fail horrible just like it will fail for computer-newbies to know there own root-password
So...I kinda liked this subject being put up. Especially since A.Seigo is enraging everytime someone mention that the desktop toolbox got to go...Don't get me wrong, that guy is marvelous, but he is a real dick when it comes to this particular subject, blabbering on an on about how he does it all and everyone should be grateful. This overlord behavior/attitude wiped all my plans of contributing as it is not exactly the first time I read stories about patches being denied for stupid egocentric reasons.
I think the crux in this argument is the purpose of the 4.0.0 releases. From my long and painful discussions with Aaron, it definitely seems like 4.0.0 to 4.0.x < 4.1.0 is about one step away from running SVN. Features can exist, even if their sole purpose is to help in debugging and to advance the development of the software. That's fine, but it really has to be communicated very clearly for it not to backfire, because in all regular software releases (yes, even .0.0 ones in the OSS world), features must have at least one tangible benefit for the users.
And at some point, that attitude has to change, because although it does lead to the software getting better in the future, it also leads to the software being worse now. Like you said Dirk, right now the zoom button is just a "break my desktop" button for users, which has a detrimental effect on their impression of the software.
Hopefully these are just symptoms of the .0.low releases. Like I said, I'm happy to test features like that and report bugs, but I'm far from a typical computer user.
it does lead to the software getting better in the future, it also leads to the software being worse now
Yeah, that's the central point and where those that are working on long time goals are confronted with those who try to get things in shape for the next release.
Probably it's Plasma where this is most observable atm cause while Plasma made huge tetra-steps, it's still heavy under progress, one of the direct visible corners of KDE4 and as such there is just a lot of pressure to move even faster forward and with each new step even more feedback and opinions arrive.
Not that easy at all to find there a balance between now and tomorrow.
Hi,
i liked this blog entry very much as a counter part of Aarons blog. They both show two reasonable points.
Personally I like your point better but also can understand Aarons point.
Quite awkward situation.
Nevertheless it will not be better if every damn stupid leading-nowhere discussion that ran out of reasoning is dicussed on blogs. It does not help the community at all if the message of every 20th blog entry is to fuckin' stop reporting obvious[tm] issues and issues already decided somewhere [tm] not to be implemented.
So, I like this blog entry to criticise that in general.
Thanks for bringing up an non-schildish-like opinion.
cheers
Given that Aaron said he would not remove this feature are you really that surprised that a patch to remove it was not received with open arms?
Trying to dress this up as a community communications issue or as a disincentive to new developers issue because your patch was not well rejected seems pretty disingenuous when it has been quite broadly aired that this feature is not going to removed. Why spend 7 days on this when the heart of the original answer was 'this feature will not be implemented'?
I agree with Dirk. KDE is plagued with this kind of politics resulting in KDE 4.0 being able to ship in the state it did (esp. Plasma) and what's worse, its own developers praising this mess and telling critics that "they've never been involved in building something great". This is one of the main reasons why I left KDE.
I fully agree with this posting, Dirk. If we cannot get past throwing attitudes around and seeing things from the user's side (developers are users too), we've already failed our mission. The crux of it is that in FOSS, we don't have the traditional customer/vendor approach, which enforces the aspect of at least attempting to do the right thing, since if we don't the customer goes away.
In the case of FOSS, it's community that drives us forward, and without listening (which lately, I'm seeing less and less of
I don't see the point to this jabbering really. Aaron made it clear that he did not want to remove the toolbox from the desktop, because he sees a great future for it. Fine. He who codes, decides. However, he also explained that the desktop is just another plugin, and that you can replace that plugin with any other plugin you like. So, what's to stop you from copying all of the desktop plugin, apply your patch to ditch the toolbox, and put it on playground (or somewhere else, who cares) labelled "DirksDesktop" instead of "defaultDesktop" or something like this? Blog about it, put up a link to it, put it in a wiki, and you'll get users for it in no time.
This way, Aaron can maryly go on in his own visionary way (or whatever you'd like to call it), and you and any other critics are no longer bothered by this toolbox thing that seems to haunt your dreams.
About the toolbox you're right. But I wasn't really caring that much about whether or not my patch was rejected but *how* it was rejected. if you're saying that forking whatever you don't like is fine then thats fine. However as part of the review I'm referring to I was told that even trying to change something in plasma would be a disservice to the project. I don't think thats a way to attract new Plasma developers - and I care about that because Plasma is central to KDE 4.
So KDE is back to the mantra "From developers for developers"? Since this is what Aaron basically said: "If you don't like code it yourself, we've given you all the necessary tools". Nevermind that most people do not have the necessary programming skills to accomplish such task and it's leaving the impression that my opinion as an end user is worth nothing because nothing must come in the way of a vagualy specified vision.
I really hope that distributions will add a remove option on their own and end this ego trip. If the toolbox has any value to the user at all he'll keep it despite the option to remove it. Instead of shoving it down people's throat he should finish the implementation first and then try to communicate its usefulness to the user (Plasma tutorials anybody?).
> my opinion as an end user is worth nothing
The opposite is the case. Developers just see things at another light and therefore for some things they are just blind. That's why feedback is such an important thing, why it was important to get 4.0 out asap, why things like e.g. the traditional menu or the option to configure the panel size got done and even backported, why we have things like Bugzilla and the great commit-digest and so on. Well, at the end it's all about the user
I totally quote borker!!
The author does not allow comments to this entry
|
QuicksearchSyndicate This BlogCommentsAlbert Astals Cid about KDE SVN: New Coverity builds available Mon, 13.10.2008 20:08 Cool thanks for the work Diederik van der Boor about KDE SVN: New Coverity builds available Mon, 13.10.2008 18:09 Nice stuff anonymous about Post Akademy action: bugs.kde.org switched to new version Tue, 19.08.2008 07:09 2 student: It's not open-sou rced yet. And it's hosting is on notkde.org server. student about Post Akademy action: bugs.kde.org switched to new version Mon, 18.08.2008 22:50 What will this mean for a norm al kde user? and why not use l aunchpad since its free and fr ee? Adult Ühler about openSUSE KDE Community IRC team meeting: Wednesday 25th, 16:00 UTC Mon, 30.06.2008 14:29 Not fair indeed. Bille about openSUSE KDE Community IRC team meeting: Wednesday 25th, 16:00 UTC Tue, 24.06.2008 18:41 Dude, you partied while I was on holiday? Not fair! Simon Edwards about PolicyKit fun Sun, 13.04.2008 18:58 Cool. I was looking at PolicyK it just yesterday, trying to f igure out what it did and what I was going to do about [...] Jayson Rowe about weekly KDE 4.1 snapshot (4.0.67), KDE 4.0.3 testers wanted Fri, 28.03.2008 20:19 Just FYI, I'm running openSUSE w/ KDE 4 off of the KDE:KDE4: STABLE: repos, and it pulled d own 4.0.3 for me last ni [...] www.kde4.de about weekly KDE 4.1 snapshot (4.0.67), KDE 4.0.3 testers wanted Fri, 28.03.2008 19:27 Good news im waitung for the s napshot packages for testing ; -) Christophe about weekly KDE 4.1 snapshot (4.0.67), KDE 4.0.3 testers wanted Fri, 28.03.2008 15:21 I would love to try kde4.1 sap shots on kubuntu. Is there any one who has compiled them? I regularly submit bugrepo [...] Dirk about KDE4 port of openSUSE Updater, KDE 4.1 snapshot (4.0.66), FOSDEM talk Mon, 17.03.2008 12:28 should be fixed meanwhile. it was a clash with KDE3's opensu se-updater-kde mimoune djouallah about KDE4 port of openSUSE Updater, KDE 4.1 snapshot (4.0.66), FOSDEM talk Sat, 15.03.2008 10:58 yeah actually kde 4.0.66 is mu ch better then the last snapsh ot, just one complain, give us back kplato in the buil [...] Erlend Hamberg about KDE4 port of openSUSE Updater, KDE 4.1 snapshot (4.0.66), FOSDEM talk Sat, 15.03.2008 10:05 I installed kde4-opensuse-upda ter-0.7.0-2.5 from KDE:KDE4:ST ABLE:Extra-Apps, but when I tr y to start it I get this [...] Sebastian Sauer about The Genesis of a Plasma patch Wed, 05.03.2008 22:44 > my opinion as an end user is worth nothing The opposite is the case. Developers just see things at another li [...] Categories |