If you were, you know, living your lives, you’ve probably missed it, but old fires are burning brightly once again: there’s somewhat of a falling-out going on between KDE and GNOME, with Canonical siding squarely with… KDE. The issue seems to revolve around GNOME’s lack of collaboration, as explained by KDE’s Aaron Seigo.
Before we get to Seigo though, I want to focus on Mark Shuttleworth. In a lengthy post on his blog, he explains that while Canonical set out to create Unity under the umbrella of a GNOME, as a shell for GNOME. They wanted to use GNOME-friendly technologies, and if they were to choose to create something on their own, it would require “substantial review”. Canonical wanted to create competition within GNOME, instead of competing with GNOME.
“We’ve failed,” Shuttleworth concludes bluntly, “Much of the language, and much of the decision making I’ve observed within Gnome, is based on the idea that Unity is competition WITH Gnome, rather than WITHIN Gnome.”
A prime example of this, according to Shuttleworth (and Seigo, but we’ll get to that in a minute) is the work on status notifiers (appindicators in Canonical-speak), which is now a FreeDesktop.org standard, and has been adopted by both Canonical’s own Unity as well as KDE. GNOME, however, while first being open to this idea, later rejected it.
The reason behind the rejection is, according to Shuttleworth at least, that GNOME sees Unity as being in competition with GNOME, and hence, technologies that sprout from it – no matter how potentially useful – are a threat, and not a gift. “It doesn’t integrate with gnome-shell” is the reason given by GNOME, which is an odd reason given the API wasn’t being pushed as an integral part of GNOME; no, it was merely proposed as an external dependency so that GNOME applications running on other platforms could make us of it the API was present.
“So the premier reason given for the rejection of these APIs is a reason that, as best we can tell, has never been used against an external dependency proposal before: ‘it’s different to Gnome’. At the heart of this statement is something deeper: ‘it’s competition with an idea someone in Gnome wants to pursue’,” Shuttleworth argues.
The underlying goal, then, seems to be to discourage GNOME application developers from supporting this API – this has failed, according to Shuttleworth, because developers have picked up on the API anyway, whether GNOME likes it or not.
“What made this single statement heartbreaking for me to see was that it spoke clearly to the end of one of Gnome’s core values: code talks,” Shuttleworth further clarifies, “Here we had APIs which were real, tested code, with patches to many Gnome apps available, that implemented a spec that had been extensively discussed on FreeDesktop.org. This was real code. Yet it was blocked because someone ^aEUR“ a Gnome Shell designer ^aEUR“ wanted to explore other ideas, ideas which at the time were not working code at all.”
It goes deeper than that, though. The work on this API, and Canonical’s vision for it, were shown to GNOME Shell designers at the 2008 UX hackfest. “[Jon] McCann denies knowledge today, but it was a clear decision on our part to talk about this work with him at the time, it was reported to me that the conversation had happened, and that we’d received the assurance that such work would be ‘a valued contribution to the shell’,” Shuttleworth recalls, “Clearly, by the time it was delivered, McCann had decided that such assurances were not binding, and that his interest in an alternative panel story trumped both that assurance and the now-extant FreeDesktop.org discussions and spec.”
KDE did do the work to work well with the status notifiers, and as such, KDE applications ‘just work’ in Unity, all because the API has become a standard and has been adopted by KDE.
According to Shuttleworth, there’s major problems with the way GNOME is currently run. While the long tail of contributors and developers are just fine and dandy, it’s leadership where things to wrong. “I’ll state plainly that I feel the long tail of good-hearted contributors to Gnome and Gnome applications are being let down by a decision-making process that has let competitive dynamics diminish the scope of Gnome itself,” he states, “Ideas that are not generated ‘at the core’ have to fight incredibly and unnecessarily hard to get oxygen. Ask the Zeitgeist team.”
“This is no way to lead a project. This is a recipe for a project that loses great people to environments that are more open to different ways of seeing the world. Elementary. Unity,” he adds.
As already mentioned, KDE’s Aaron Seigo has also come to the conclusion that GNOME is not collaborating with the greater Free software community as much as it used to. “When I got to the point in Dave’s blog where he justified GNOME not picking up the appindicator work I cringed and then cringed some more,” Seigo explains, “For you see, the appindicator story is not so much about GNOME and Canonical as it is about GNOME and the rest of the free software desktop ecosystem and the regressive behavior being demonstrated there.”
Once, Seigo recalls, there was a lot of effort being put from all parties into closing the several interoperability gaps present in the Free software stack, with FreeDesktop.org being an important piece of that puzzle. More recently, however, it seems like GNOME is eschewing this idea.
“Those days seem to be receding into the past however as GNOME seems increasingly content to ‘do their own thing’,” Seigo believes, “It’s like they have lost sight of the goal of a coherent experience for users of Free software regardless of the applications they choose to use.”
“It is indeed not incumbent upon any project, be it GNOME, KDE or any other to adopt every single technology on offer,” Seigo continues, “However, it is in all our best interests to work together more rather than less and to share what makes sense, even if sometimes it means some short term compromise. KDE has made that decision many times even in recent times as we adopted various technologies such as D-Bus, shared mimetypes, shared icon theme definitions, etc. It’s taken a lot of effort on our part and at times we’ve had to make some compromises, but our users have been reaping the benefits.”
While I generally prefer GNOME to KDE, there’s no denying this last point. KDE has always worked a lot harder on integrating well with GNOME than the other way around, and it seems like this pattern is intensifying instead of diminishing. That’s a very bad thing, since cooperation, especially on lower-level elements, is key in the Free software world. Call it a blessing or a curse, but fact is there are multiple desktop environments, and they need to work together as much as possible.
We’ll have to see what this all leads to, but I’m hoping this ruffles a lot of feather and causes a great stink, since a re-evaluation of what kind of cooperation is important could be just what the doctor ordered.
You can’t blame Gnome for ‘not co-operating’. They don’t want to co-operate, it’s their right to not co-operate.
Although my opinion is that there even shouldn’t be two or more different desktop environments for Linux, I don’t think Gnome devs can be blamed for the path they have chosen.
That’s the price of freedom. We may not agree with the others, but we’ve got to respect their position.
Its not so much everyone is saying Gnome is wrong.
This is meant to focus attention what is going on and to understand why things are happening the way they are. Its the equivalent of asking someone why they don’t want to be your friend anymore.
If Gnome flat out said we just don’t want to collaborate anymore – then yeah its their prerogative and everyone would just have to accept that but that doesn’t mean people can’t ask why they decided that.
True they can choose a path. In the past they chose cooperation and that was a good decision.
I am so glad things like fonts, copy and paste, and other things that I am not in the know about, just work today.
Yes, it’s their right to choose the path they want.
That does however not mean that they can’t or shouldn’t be critiqued for their choice.
I for one am very happy to have a wide range of DE/WM choices in GNU/Linux, and by extension in BSD-based OSes. If it were just Gnome or just KDE or just the two of them, projects such as Tiny Core Linux, CrunchBang, Puppy, DSL, GParted LiveCD and countless system recovery and other quite useful distros wouldn’t exist. GNU/Linux doesn’t just exist to serve the needs of the very few desktop OS-oriented users. It exists for everyone and as such, there are a myriad of ways to get information from the system to the screen.
Canonical can choose KDE or Gnome or whatever it wants to, meanwhile all of the above distros plus some (GNUStep, Zenwalk, Wolvix…) will remain in my CD wallet for whatever I need them for on a particular day.
It certainly is Gnome’s prerogative not to use what the developers don’t want to use, but what has been really distasteful over the years is when Gnome has come up with it’s own solutions to things, shoved them on Free Desktop and portrayed them as FD ‘standards’ (when in reality they serve their own agenda) with no consultation or work with anyone else. Mostly they just ouright pass the work of others as their own. The Tango icon spec many moons ago was a good example. It was actually a KDE contributed spec that was dressed by by other people as something different.
Gnome simply starts from a default position of “If you’re compatible with us and use our libraries then you’re cross desktop friendly”, and that’s the way things are constantly portrayed.
I’ve certainly been pointing out this for years but it all gets portrayed as trolling and those around Gnome have all tried to weasel out of it by portraying it all as a ‘misunderstanding’ and that it’s somehow all everyone’s fault. I suppose it’s taken this long for it all to come to a head.
There is something deeply endemic and rotten in the Gnome project because the faces have steadily changed over the years but the attitudes have remained oddly familiar. It’s an attitude of being deeply protectionist, close minded and scared of anything that might be better than what they’ve come up with. However, the disturbing thing is that they routinely try and weasel out of that by telling people they have misunderstood, things haven’t been discussed when they have as Jon McCann did and just plain sticking to a story that isn’t true.
Gnome is a project with an endemic mental disease, put simply.
Mark Shuttleworth will probably be wishing he had just gone with KDE all those years ago and had a positive effect and relationship there. It might be too late for him and his company now.
Edited 2011-03-11 22:57 UTC
+1
I could say more but why bother, that statements stands on its own and any addition from me will most likely spoil it.
Well, at least that is at last being acknowledged now. I’ve simply got sick of people portraying Freedesktop specs as somehow all being Gnome originated and this terrible double standard of where if you interoperate with Gnome somehow you’re being ‘cross desktop’ friendly. The reverse never seems to be expected.
It’s something disturbing that has gone on for the best part of the last decade, and I’d say it has materially harmed, perhaps irretrievably, the perception of Linux and free software on the desktop and ensured that we always seem to be in a political mess.
They won’t do what we want! *cries* I agree with the poster above, that’s what happens in an open ecosystem like this. Maybe Canonical and KDE think the appindicators and Unity are a good idea, but maybe the GNOME folks don’t. That’s the way it is. Shuttleworth needs to quit whining when others decide not to follow him. This isn’t iOS where everything needs approval from someone else, this is an open system. If Shuttleworth doesn’t like the way GNOME is handling things, he’s free to switch Ubuntu to KDE if he wishes. Quite honestly, I don’t really care what he does since I use Arch anyway, but he sure seems to complain too much.
Funny thing is, my GNOME installation on ARch is way more stable *without* Canonical’s additions. Maybe that has something to do with GNOME’s reluctance?
True, but if Gnome leadership first indicates “great idea, let’s work together” and later changes plans, then others will lose trust in such leadership.
Actually, Shuttleworth created Unity for precisely this reason. His only disagreement is with GNOMEshell. WRT whining, he’s getting regularly attacked by people in the GNOME camp for not using GNOMEshell, so restating the reasons for not using GNOMEshell is appropriate.
Personally, I haven’t tried either Unity or GNOMEshell, but from what I’ve seen I don’t look forward to either. I am glad there are alternatives. Let the best shell win. I am willing to give both a try, but if neither work for me, KDE4 doesn’t either so I’ll go for XFCE since it has always “just worked”.
XFCE is indeed a very good alternative to GNOME. GNOME apps work nicely in XFCE, and XFCE offers a stable and familiar desktop environment.
Pretty much the only thing I don’t like about XFCE is its window manager, which is almost as lame as metacity. Openbox is a much better alternative, and you can easily change the WM in XFCE to openbox by typing in a terminal window this command:
killall xfwm4 ; openbox & exit
Then exit XFCE, saving the session. The next time you start XFCE, it will use openbox as its WM.
Having used OpenBox, Xfwm4 and Metacity, I can’t imagine why you’d think Xfwm4 has more in common with the latter, rather than the former. Xfwm4 is a lot more flexible and capable than Metacity, if nothing else.
Indeed, I’m itching to give the newest release a spin, though it means compiling for a few hours. Spare time is spare indeed in my current work-work-sleep-work schedule, but I may give it a go later this weekend.
If GNOME removes the minimize buttons everyone criticizes them for it (despite the fact that it was a sensible choice) but when they refuse to cooperate, making things actually worse for the users – it is their right to do that?!?
Of course it’s their right to do that, as it is your right in Japan to go Whale fishing and I bet there are countries where seal clubbing is legal too. Doesn’t mean it’s the right thing to do!
I rather criticize them for both decisions I will definitely not like having those buttons removed and several other “features” being introduced in GNOME3, and even though I haven’t bothered reading all the replies in those mailing list threads about this current topic I can’t help but feel slightly disappointed in GNOME devs.
disclaimer: I am not for Ubuntu or any distro, this is only my opinion based on my reading of facts.
Whose whining first? The folks who accuse Canonical for not contributing back to the Kernel, and now morphed into Banshee then to appindicator and Unity.
Please follow the relevant blogs, I see that the GNOME folks were at fault first for not collaborating on the open(freedesktop.org) despite of their _DENIAL_.
You gave KDE and Canonical’s perspective, but you didn’t give any GNOME perspective.
Unfortunately, in this case their perspective is wrong. They were initially open to the idea, made requests to change some things in the spec, many of which were done in order to accommodate gnome. Some things were not changed but were justified technically by kde, whereupon the gnome devs stopped replying…
Then later said no we don’t want this, it does not fit with our vision. It could have been made to fit but they chose to ignore it.
So say Canonical and KDE. The point is, we don’t have any GNOME perspective linked here. Quoting back Shuttleworth and Saigo doesn’t equate to having the GNOME side of the story, and the one thing we must always keep in mind in these situations is that no story is ever one-sided.
Read the comments in Aarons blog post…
Or here;
http://psankar.blogspot.com/2011/03/gnome-vs-canonical-freedesktopo…
That’s all anecdotal. Where is the evidence? Where are mails on the mailing list? What GNOME developers explicitly or implicitly rejected the proposals? And why? Hearsay doesn’t count.
Here you go:
http://lists.freedesktop.org/archives/xdg/2010-January/011228.html
I think that thread should provide ample proof…
I read through that whole thread, starting with that message. Frankly, Aaron does not come off well here. The Gnome folks raised valid concerns about the vagueness of the specification and some protocol issues, and Aaron’s response was to obfuscate or say the issues had been raised elsewhere (but certainly not addressed in the specification).
If you think this discussion indicates a lack of willingness to cooperate on the part of the Gnome developers, then you must not read many mailing lists or participate in specification development. This is an absolutely normal conversation, except that one of the participants is not really responding to the issues raised.
Unfortunately the KDE and Unity developers seem willing to put out a specification that’s not quite fully baked and let dominant implementation practice define expected behavior instead of capturing it in the document. When that happens, what results is a mess of backwards compatibility hacks and nearly-compliant implementations that fall apart at the edges. It’s why you can’t just write a CSS + HTML document and have it render the same on all browsers.
Doing specifications right isn’t the expedient path, but it is the right path, and I’m disappointed that the Gnome developers’ insistence on it is being used as evidence of their uncooperative nature. If any proprietary software developer acted this way regarding development of an open specification, they’d rightly be excoriated for it here.
I seem to have a different perspective. I just don’t think they were being particularly helpful. The gnome developers were focussing on silly details like api naming that everyone agreed would have to be changed.
They did not provide use cases or solid justification for their stated positions. Nor did they at any point say “look this simply does not work for us, how about doing it this way…”
The whole point of the discussion was to flesh out the specification and get others to join in. AFAIK that is what canonical did when the spec was first created.
Edited 2011-03-10 16:27 UTC
Well, on the other hand if you read the last email from Aaron, he made it abundantly clear that no amount of use cases and other stuff could make him change his mind about fixing the vagueness of the spec, which is really what bothered Dan Winship and Mattias Clasen.
What he’s arguing now is that the GNOME developers should just have eaten what they got served in the name of cross-desktop compatibility. And that’s a fair point, no doubt about that.
Well, if they had been more proactive they would have gotten to write parts of the menu. I’m much more inclined to take the side of kde as recent history shows kde being far more proactive on cross desktop than gnome.
Sure. I actually think he had a fair point in that collaboration means you may have to eat stuff you don’t like, wasn’t trying to be sarcastic.
It’s just that he pointed fingers, failing to say that he himself in the particular case he mentions hadn’t been overly receptive to the kind of collaboration where you try very hard to see the point from the other side.
To be honest, I think the big mistake in this whole mess is that way people are arguing over it. If you really want to change the way people go about their business, engaging them privately in a manner where egos and tempers won’t self-inflate immediately just seems much more effective. I just hope the whole thing manages to push people into starting those conversations…
They didn’t have to “eat what they got served”. They could have proposed something else. No, what happened is they stopped communicating, went off and implemented a different systray/notifier mechanism and barely even bothered mention it to the people on fd.o who were obviously interested. This isn’t about egos. This is about not fragmenting the FOSS app/DE ecosystem even more than it already is. Collaboration requires communication. You don’t like what was proposed? Fine. Propose something else or leave things as they are. Don’t go off creating something entirely different and expect folks to not get upset when you didn’t care enough to tell them about it. As an app developer, I now have to decide which systray/notification mechanism to support or just go back to using crappy XEmbed. That’s not progress…
As far as the merits of the Status Notifier spec, what a pleasure it is to run KDE apps in Ubuntu and see consistent GTK menus and tooltips in the GNOME systray instead of an out-of-place Qt/KDE menu or tooltip. How nice to see Plasma tooltips for Rythmbox instead of some out-of-place GTK tooltip on my KDE desktop. You would never think there were *any* practical value to the specs just reading their objections and, unfortunately, their objections are all we have as a testament to their efforts at cross-desktop collaboration on this particular issue. Again, this is not progress…
Agreed; from reading that thread, there was a major gap between Aaron’s thinking, and theirs. Aaron was insistent that the spec cover only the communication channel and leave the presentation entirely up to the visualisation component. And that was completely unacceptable to Dan and Mattias, who felt that a spec that didn’t cover the presentation was not usable.
Given those two uncompromising viewpoints, I don’t think there was ever any chance of the two sides agreeing.
Except for the fact that the spec was usable and was, in fact, in extensive use by that point. The whole claim that the spec could not work was demonstrably false.
Further, despite repeated requests they refused to provide any reasons or use-cases to show how it wouldn’t work. Aaron provided extensive reasoning and use-cases for why he thought it needed to be the way it did. The others refused to do so, they just stated the claim and expected to to be accepted as unquestionable fact.
So on one side we have several working implementations that show it does work, detailed explanations about the reasons for the structure, and specific use-cases that support this decision. On the other hand we have a claim that it is impossible to implement (despite the fact that it already has been implemented), no reason for this claim, no use-cases for this claim, and no response to the other sides reasons or use-cases.
No, Aaron makes clear that the aim of the spec is to FIX the current situation, not duplicate it in another way. They all agreed the current situation sucked with every app basically having full control of the systray (hence the systray container having no control at all).
In short, they were being stubborn and (probably on purpose) overly negative and shortsighted. If you have the vaguest idea of what is wrong with the systray (just read up a bit on Aaron’s blogs over the past 5 years) then you can only say that they are being idiots in every way. And again, I think you can blame them for that as unlike most of the readers here they actually know what Aaron is talking about.
That’s not a lack of willingness of the part of GNOME developers, those are all valid concerns and very constructive criticisms not unlike what you’ll find in any software engineering project.
No, it’s not. Read again. For example, this message by Aaron: http://lists.freedesktop.org/archives/xdg/2010-January/011236.html
Read it fully…
And the response to that by Dan (a GNOME developer).
http://lists.freedesktop.org/archives/xdg/2010-January/011238.html
Read it fully…
Yes, and Aaron responded to his mail as well of course. Dan got exposed for his silly arguments and of course kept going and going. I guess he and Aaron should spend some time in a room to fight it out – he refuses to understand what it is all about. And as I said before, if a non-technical person like me can get that, and if there are already 3 entirely independent implementations of the standard, Dan’s stubborn criticism can only be seen as dishonest.
Well, I’m a technical person and I disagree with your non-technical perspective.
So lets get technical. You can’t write a specification to an implementation or write an implementation then call it a specification. A specification is a detailed unambiguous protocol. It’s not an API. That’s the crux of Dan’s argument. And that argument is “technically” sound.
This is all software engineering jargon and I don’t know your background. However, I’d refrain from making wild accusations on Dan’s judgement given that some technical knowledge is needed to make any sensible remark on the issue.
As aaron mentioned later in the discussion, the spec being discussed here was suddenly held to a MUCH higher standard than ANY other spec that has made it to FD.o. Sure, a spec is what you say it is – but most work done on FD.o is of a ‘slightly lower standard’ (eg see this comment: http://aseigo.blogspot.com/2011/03/collaborations-demise.html?showC… )
Uh? Are you sure you read this piece?
Yeah, I did. I don’t see any insights from GNOME developers.
A gnome developers perspective:
http://nathaniel.themccallums.org/2011/03/10/i-hate-to-say-i-told-y…
(Not the project’s perspective, but that of an individual.)
Good job. It’s good to see a balanced perspective. It seems this post by osnews is skewed to vilify GNOME developers.
Oh go boo-hooing somewhere else.
Cry me a fricking river.
Well you provided quotes from Aaron (kde) and Mark (ubuntu) and none from any gnome developers involved in this issue. How can you pretend this is balanced?
Totally agree with you Mystilleef. It’s sad to see some of your comments voted down when the points you made were perfectly valid. The STFU response from Thom was also pretty dissapointing.
This piece can only be described as opinion. It comes down hard on Gnome, and definitely does not present their side of the story in any way.
While I happen to agree that Gnome is in the wrong on this issue, authors have to decide when writing a piece like this whether they are going to provide opinion (which is fine), or balanced news.
The distinction between the two is rarely, if ever, made on OSAlert.
Juvenile journalism, that’s what osnews has become. Only at osnews does an editor attack its readership. It shows the magnitude of your maturity and professionalism and respect for your audience Thom. I remember when osnews used to be populated by constructive debates (Eugenia’s time). That’s why I hardly contribute anymore.
I think all parties share a blame in this issue. They’re all playing politics.
Ubuntu can’t spend years developing projects in private, code dump it on GNOME, demand they retain copyright assignments and expect GNOME to agree to all that. It just doesn’t work that way. Yes, Ubuntu doesn’t get any special privileges. And they need to stop all these __private__ messages. Leave everything in the public (mailing list, bugzilla, forums, etc) where everyone can see it.
KDE can’t let their egos come into play when GNOME developers provide constructive criticisms. If GNOME rejects something, ask what can be done to achieve the goal of desktop interoperability. Same goes for GNOME.
GNOME needs to understand that community development is a give and take process. Sometimes you have to compromise to keep the peace. It can’t be GNOME’s way or the highway. They’ll alienate the community. GNOME needs to work on a process that make contributing to GNOME easy and __communicate__ better with its community. The perception that GNOME is an elitist community needs to be addressed by GNOME.
Yeah yeah and if the article had chosen KDE’s side you wouldn’t have said a thing.
You just don’t like it that I’m not siding with GNOME in my article – which is my prerogative. It’s fine if you dislike that, but don’t try to wrap it in a hoity-toity shroud of “journalistic standards blah blah”. We’re a blog, and we’ve never pretended otherwise. Don’t try to make us into something we don’t pretend to be just to claim some moral high ground.
You’re not called OSBlog.
Well, they’re not called “AlwaysObjectiveNews” either.
I don’t mind your opinions. However, I don’t come here to read your opinions. I’d subscribe to your blog or tweets if I was interested in your opinions. I come here for news. Is it too much to expect news to be balanced and objective?
Skewing the news to fit your opinions is unprofessional and tarnishes your credibility as a news outlet. Provide the evidence and facts and let your audience judge for themselves.
I’m not saying you shouldn’t give an editorial opinion, but make it known via a disclaimer that you’re making one.
I don’t get a commission for who you side with. So I don’t really care whose side you’re on.
OSAlert has been this way for years, and only now, when the opinion is not to your liking, do you complain? That’s typical.
There is no such thing as balanced news, and never did we state we aspire to it. You are just saying that because you want to discredit me, because you can’t stand the fact that I’m not on your side. That’s fine, but then be a man about it and just say so. This hoity-toity behaviour is silly and isn’t fooling anyone.
Edited 2011-03-11 15:29 UTC
I’m not discrediting you. I’m just giving you feedback as an osnews reader. I think your reporting in this instance is pretty one-sided. I’d like to see more balanced reporting from you in the future. Good job so far and good luck.
He is giving an opinion. Excuse him for thinking that the osnews readers are intelligent enough to form their own opinion. There is after all a very active comments section…
I’m not bothered by his opinion. I’m disappointed by his deliberately skewed reporting. He doesn’t seem to be denying it either.
I don’t have to deny anything. Reporting is never not skewed, and we here at OSAlert never pretend we are not skewed either. We have been like this for god knows how long, and only now, when the opinion doesn’t suit you, you’re throwing a fit.
And, as you can see, people just don’t buy that. You’re complaining the ice cream I sell you is strawberry instead of chocolate, even though we don’t sell chocolate ice cream and never have, never advertised as having chocolate ice cream, and clearly state we are selling you strawberry ice cream.
Well journalistic integrity holds that reporting __should__ be as objective as possible. It’s okay to have an opinion.
It’s okay to say, “my opinion is that GNOME developers are elitist and never collaborate.” That’s fine. But what you’re doing is clever. You’re saying, “Aaron (kde) says GNOME developers are elitist. Mark (ubuntu) says GNOME developers are elitist. Thus, GNOME developers are elitist.” You’re masking your opinion piece as a news report. When you’re called out on this deception then you used the “opinion” cup out. Mind you no where in this article did you explicitly state this is an opinion piece. Now we can conclude osnews reports GNOME developers are elitist.
This is why journalistic integrity holds that reporting __should__ be objective. An opinionated reporter has very little credibility. Your opinions are not more important than facts and evidence. Limit your opinions to your blog. On a “news” site focus on facts and try to be as __objective__ as possible when “reporting” it. Of course give your opinion, but state explicitly that it is so. Otherwise, this news reads that osnews claims that GNOME developers are elitist.
I’ve been an osnews member since 2005, Eugenia and Adam gave a lot of controversial opinion pieces. The difference is that they always explicitly stated they were opinion pieces by placing disclaimers everywhere.
Yes, yes, yes. Everyone hates you. Have a biscuit.
This discussion is going nowhere. If you’re not smart enough to set an opinion apart from fact without being taken by the hand and labelling everything, then you’re simply not cut out for reading the web.
I’m sorry.
You’re not sorry. You’re being arrogant and defensive. I guess osnews is your blog. You can do with it as you please.
1. Dave Neary’s blog posting is in both those links.
2. They don’t want to respond for obvious reasons.
3. They have made their feelings clear in the comments on those blog entries – they claim that things that others had said had been agreed never took place and that’s all. Have a read.
4. They hope this will all kind of go away by giving vague answers.
I’m a lot more interested in responses by developers involved in these discussions on the mailing lists, forums and bug reports.
Comments in blogs are usually opinionated noise.
Mark claims that libappindicator was proposed to GNOME and rejected. I can’t find any entries for this in the gnome mailing list.
Aaron’s discussion with the GNOME developers about the notification specification has already been posted somewhere in this thread. It’s a far cry from GNOME developers being elitist. As a matter of fact, very real software engineering issues were raised.
However, I’m tired of all this. It’s a waste of my time.
I’m afraid you’re just trying to get this to dissipate via some silly divide and conquer type commentary probably by claiming that there is all some misunderstanding going on here. That’s what usually happens.
The blog posts from various people, including those from a Gnome perspective are there as are the links to mailing list activity and people whove’ gone through them. If you feel one side has been treated unfairly then do the research, redress the balance and post here. Until you do that you’re not going to be successful in arguing anything.
I think that answers your query then. Was there any point?
Who’s arguing? It’s sad I’m getting a more balanced perspective from blogs than I am from osnews. For those who are interested in thoughtful GNOME perspective read the links below.
http://blogs.gnome.org/bolsh/2011/03/11/lessons-learned/
http://blog.fishsoup.net/2011/03/11/what-does-the-user-see/
I guess I expected too much from osnews.
Edited 2011-03-12 02:01 UTC
You might want to look at the timestamps of those two posts, mister smartypants.
This has nothing to do with timestamps. I got a more balanced perspective from these blogs than I did from your one-side slant. Of course if you had did your homework, as opposed to prematurely and prejudicially reporting events you clearly had not researched, we won’t be having this discussion. However, you’ve insisted I shouldn’t hold you to high standards of reporting. So carry on.
And mister smartypants? I don’t know you like that. Lets keep it mature, or you’ll be ignored.
No, this post was good and that response was quite short sighted.
Surely working upstream is good. I would never say Canonical did it right. But let’s stay realistic: GNOME would never accept something developed somewhere else, yes? Like Nautilus, maybe? or Evolution?
Both sides made mistakes and Holwerda is very right in pointing that out.
Where did he point that out? It seemed to me he concluded GNOME was at fault and sided with KDE and Ubuntu in the matter.
Well, GNOME IS at fault here and they refuse to acknowledge that. Canonical is also at fault and Mark has acknowledged that in his blog post – at least to some extend. There is a lot more they have to do (their policy of doing all development on Launchpad, copyright assignment etc is quite harmful) but let’s see how that works out…
I can’t figure out anything KDE did wrong in this whole thing – except maybe put in too much effort in cooperating in the face of sure rejection (again) or not exposing what was going on earlier.
lol, you just made another side to the point.
as a spectator after reading everything i can only say this
damn… i wish these people were hot babes fighting in the mud located few feet away from me while i am lying on the couch and drinking beer.
one wants to be steve jobs of FOSS and actually believes he knows the road to heaven. ones have one and the only truth. and one is just ranting to put more oil on the fire between the first two.
boohoo, cry me a river. make competing desktops where everyone supports as many distros as possible and let users decide which one is the best. then… if you lose, admit and support the winner by contributing to his project instead of yours
on personal note… seeing unity, i was disappointed. last thing i want is broken design. and even more so osx like menu. i also hate osx like pinned dialogs in g-s, but they are easy to disable. currently, on my desktop, g-s is winning, while kde never had even the slightest chance
Do you really love those icons in the dash when you have 20 windows all over your workspaces in GNOME Shell???? Please explain, those icons will get smaller and smaller, explain please why GS is better.
Well, you might want to read the comments. Jeff Waughn came in guns blazing and called Mark a liar without any specifics but a promise to dig up some dirt more later to prove his point. Then he went on to drive the discussion on Aaron’s blog offtopic by blathering about timelines, posts from 3 years ago as well as insisting on discussing who said what at a conference in 2008. Finally, he managed to degenerate the discussion into a flamewar. If Jeff’s antics are any indication of the way GNOME will deal with the issue, I think we may have to give up any hope of getting them working with the larger FLOSS community (and freedesktop.org).
I’d not blame Jeff exclusively on that – by my reading, half the participants (including both Jeff and Mark) were behaving like squabbling kids in the kindergarten playground. None of them setting a particularly good example of leadership…
Yes I saw that. It was a little funny if nothing else. Where the hell did that cancerous influence come from? He hasn’t done anything in, well ever, and he bulldozes his way in trying to talk for Gnome, but then sticking in disclaimers to tell people that he doesn’t actually speak for Canonical or Gnome.
We all know what the Gnome perspective is, because we’ve experienced it over the last decade.
They will weasel out of directly answering anything, portray it all as a misunderstanding and that it is all somehow everyone’s fault. They will tell us that certain things haven’t been discussed when they have. Jon McCann as well as others have done this already. Notice it’s everyone elses’ word against his and nothing can be proved or disproved? They then keep things going round in this cycle and no one gets anything out of them until the whole thing kind of goes away.
Nothing will then be solved and things will not move forwards. We’ll then waste another decade.
All I want to see are the correspondences in the mailing lists, forums and bug reports. Hearsay, anecdotal accounts, conflated events and emotional tantrums are irrelevant.
They’re in those links, including Dave Neary’s blog posting.
I’m afraid you’re not going to be able to get around this by claiming that Gnome’s ‘position’ hasn’t been represented and also claiming that everyone else is having a tantrum.
None of this is hearsay either. Read. Don’t post meaningless crap. READ.
http://blogs.gnome.org/bolsh/2011/03/11/lessons-learned/
I think this could be considered “GNOME’s perspective”, although from reading it a bit, I get the feeling that there is no “GNOME group” to give their perspective.
Yeap, I posted it above too. That’s a very well written entry.
Let’s see, when Canonical tries something different in doors is ok because is their right, and when GNOME tries something different in doors is because is “not cooperating”, go figure.
Canonical is one company making one distribution. If you don’t like what they do you can switch. GNOME is much bigger and much more influential and not under the dictatorial control of one person. If Shuttleworth wants his personal distro to go in a direction he has merely to point and that’s all there is to it. If you want to direct GNOME somewhere you need consensus and support from a majority of developers, at least, if not users.
The KDE Boss who told his user base to P**s off (You don’t need no steenkin’ desktop icons, among many other issues) when he introduced the worthlessness that was/is KDE 4x is lecturing somebody about working for the benefit of the users.
Seriously, the current status of linux desktops is they are building more and more baroque constructs on less and less solid foundations (not speaking of the Kernel here, but the underpinnings of KDE and Gnome). More bling, more regressions, random stuff not working that was working before. Every one wants to be a star, to introduce the next big thing, nobody wants to do the dirty work.
Oh no, not again this. You can have a regular desktop in KDE 4.x too, just switch to classic desktop (folder view) mode. It’s been here from the beginning, as far as I remember. (okay, it was probably not that usable in the very beginning)
Please, let’s keep the discussion constructive & civil.
Gnome removes features for who knows how long due to them being “confusing” or “user-unfriendly”. And now… Gnome refuses to add *new* features due to good ol’ Not Invented Here syndrome. Gnome, seriously, get your shit together. Quit acting like a proprietary, for-profit software vendor, a.k.a. Microsoft/Apple. I thought the whole f***ing point of free and open source software was to share ideas? I’ve been using Gnome for a while now, but it looks like the time to ditch it is drawing nearer and nearer every day.
Edited 2011-03-10 16:14 UTC
Wasn’t Miguel De Icaza’s rationale for using Mono in GNOME that pragmatism in open-source world is needed? [I don’t think he *actually* said that, but that’s the message I get]
Very much his attitude, yes. People constantly bash him as “pro Microsoft”, but in practice his attitude is simply that MS have produced some good ideas that shouldn’t be ignored just because of where they came from.
The problem is that if you point out some obvious problems with that (Microsoft don’t like you using their inventions for nothing) and you don’t like his approach you either get silence from him, hoping you’ll go away, and then you get a deluge of nonsense if you don’t.
The OOXML promotion nonsense was another classic, popcorn eating event.
What’s Aaron going to say next when Canonical do a dconf code drop he is not doing to like? GNOME have collaborated with KDE but just because they don’t like so called standards (not actual standards) just a place where so called standards are added, doesn’t mean they have to.
I think Canonical have sweetened the pot by using Qt, yet ironically Aaron doesn’t agree with Canonical CLA, which was drawn up without any “collaboration”.
It’s not that ironic – Aaron is quite consistent in his views on cooperation. Basically his view comes down to this: discuss common goals, then work together to achieve those goals – the goals being low level interoperability, while leaving UI design decisions alone.
So in the first case, Aaron criticized Canonical’s efforts because they came out with a ready-made solution with a take it or leave it attitude. Put it simply, his problem was that they missed the first few steps of normal cooperation, that is, discuss, code, and implement it together. That is what working with the community is about, not presenting code developed in-house without any chance for KDE to participate, than expecting them to adapt it right away.
In this particular case, however, KDE was just one participant in drafting a specification (Ubuntu/Canonical, Compiz, etc. being others). GNOME developers signaled a willingness to participate, than rejected any idea for obviously political reasons (Mark as well as Seigo points out that the reasons GNOME devs gave for rejecting this spec are plain nonsense).
So there is nothing ironic in his position – Aaron (and KDE in general), for the past few years at least, has consistently supported efforts to collaborate via freedesktop.org on specs and standards that would enhance interoperability between various apps in the F/LOSS stack. GNOME, however, showed a consistent refusal to work on this common goal, diminishing the importance of fd.o through their efforts. This is bad for users and developers alike, that’s basically his main point. The latter has nothing to do with his previous critique of Canonical’s QT decision. These are different issues (although in both cases, Aaron’s point was more or less the same: work together in implementing common specs and framework to improve interoperability between various software stacks.
Oddly, this has largely become Gnome’s default position over the last decade when it comes to awkward home truths about Freedesktop. I think it’s also been established that they haven’t collaborated as much as they like to portray to everyone.
You’re also missing the point here. It’s certainly their prerogative to do that, but when you talk with people and find common ground and come to some agreement about something, and then turn around, do the opposite and then claim that talk and agreement never happened then you can expect to be called out on it.
This crap has been happening for a decade now and it’s hurt all of us using free desktop software.
Aaron has been consistently critical of Canonical over some of the things they have done, including this, over many years. There is nothing ‘ironic’ about this nor is there any double standards going on here as you’re probably trying to imply.
Edited 2011-03-11 23:38 UTC
It’s been over 12 years now since the Gnome-KDE “war” began and fractured the Linux desktop. And Linux on the Desktop is still virtually no where in the market place – measly, meaningless usage rates.
Yawn. Shuttleworth can keep linux for his servers (which is probably the only place he’s making money), with a KDE shell if necessary. Then put his resources behind Haiku for a client desktop. I don’t see anything else at this point that has a chance of gaining meaningful market share from Windows or Apple.
The linux desktop is still a mess after more than a decade.
No Thanks! Keep Shuttleworth and Ubuntu devs well away from Haiku. The last thing we need is Haiku been broken too!
“Then put his resources behind Haiku for a client desktop. I don’t see anything else at this point that has a chance of gaining meaningful market share from Windows or Apple.”
ROFL – that was a good one …
I’m willing to bet that Haiku never, ever, will be as complete or usable a desktop environment as Ubuntu of today.
Is it? And if it is, “with a little help from my friends” (Debian).
Even if that’d mean setting a really low goal for themselves.
Don’t drink a lot. Are you sure Haiku supports all those massive hardware out there that Linux supports? You want Mark to use the Haiku kernel and be again crippled on all hardware fronts?
I say, Canonical was taking the right steps = Shell(debatable if GS or Unity) but its the right decision + wayland on the GUI, then support QT framework to get a massive ISV support. Application content is the key to success, then Linux on the Kernel, no not from BSD or Haiku since it will create a massive headaches and will almost require a rewrite of everything just to get it work.
Here’s the thing.
If -I- and my co-workers can use KDE (or GNOME) and go about daily lives without using Windows – it’s a sold proof that Linux has come a -long- way in the past 10 years.
I switched from Windows 2K to RedHat 7.2 as my primary OS back at 2002 and change is simply amazing.
Back in 2002, KDE 2.x and GNOME 1.x were miles behind Windows 2K and Windows XP UI while on the other hand, in many respects KDE 4.6 makes Windows 7 DE look bare.
A couple of examples:
Real virtual desktops with different icons and widget per virtual desktop (via activities), class / application name based windows grouping and virtual desktop placement, network [ssh/smb/X11/etc] transparency across the board, far better file management (explorer hasn’t aged very well), far, far, far better console (The Win32 console API should be table behind the shed and killed), and I can continue…
Now, Linux may never replacing Windows as -the- desktop Operating system, and frankly, I can’t say that I really care. (And given the rise of Pad’s and smart phones, I’m not sure the market ever cares…)
As long as -I- and the people around me have the tools that we require, I’m perfectly happy in using a OS that only has 1-2% market share.
– Gilboa
Is there anyone here sick and tired of GNOME/Canonical/KDE war? Let GNOME guys shoot themselves in the face, for goodness sake. Can’t we just have something different from KDE or GNOME that will just overthrow these fights and render this desktop war over once for all? Why XFCE, Fluxbox, Enlightenment, etc aren’t just paying attention to this and come with a breakthrough to save us.
Because they haven’t got the usage share that Gnome and KDE have. Instead, what you see is the little guys trying to be compatible with Free Desktop specs, while allowing their users to add features from Gnome or KDE (mostly Gnome) to fill in any desktop gaps.
They’re also desktops where people tip their hat to them…….until they realise that they don’t have the functionality of the two main desktops and they are called ‘lightweight’ for a reason.
One thing I’m missing from these discussions is how exactly does the current Gnome Shell implement app notifications.
Does it have them at all? If they are implemented, what features do they use that couldn’t be done if the KDE/Canonical/Freedesktop spec was being used? If the Freedesktop spec is lacking in functionality, then would it be possible to change it to add whatever the Gnome Shell needs to work?
Instead of technical issues like these being discussed, there just seems to be a lot of naming calling. Just because cooperation failed for whatever reason a year or two ago, I don’t see any reason for the parties involved to just give up.
I don’t think it’s notifications – as far as I could tell it’s about the little grey-scale icons in the panel that control stuff like volume, networking, user-switching, etc.
I read the discussion on the freedesktop list about ‘appindicators’ and understood it to be an attempt to separate the info that an app might want to send to a desktop shell about its current state, from how the desktop shell might present the data. So ‘little grey-scale icons’ as you describe could be one way of presenting the data, but the spec didn’t mandate that.
The Gnome devs appeared to be saying that it would be impossible to separate the data sent by the app, from how it is presented. They said the freedesktop org spec couldn’t possibly work, which might well be a reasonable opinion for all I know.
But surely now we have three actual implementations of ‘little grey-scale icon things’ or similar, it should be possible to objectively describe the technical issues, and evaluate the opinions that were expressed a year or two ago?
The truth is we had then and even more now many many examples of applications that quite successfully separate data from presentation in the systray. The vast majority of these aren’t new fandangled apps, but legacy apps that were easily updated to support the features made available by the spec. In fact for apps that weren’t updated, all current implementations of the spec fall back to the old XEmbed behaviour – they behave as they always have. The actual implemenations of the spec (KDE, Unity and Cairo-Dock) demonstrate quite clearly that, not only can it be done but it works quite well.
Who do you mean by ‘we’ here – is that the KDE or Unity developers or the Gnome Shell ones?
There have only been complaints that people like Aaron Seigo and Mark Shuttleworth are being unreasonable, and I’m still not clear how the Gnome Shell handles this functionality. I don’t understand why the Gnome Shell devs aren’t explaining how their tech works, as that would appear to be the best way to resolve why they didn’t go with the freedesktop spec.
The KDE folks did and Canonical had examples of apps patched to use libappindicator. Gnome Shell folks appeared to have objections then went effectively incommunicado.
I won’t dwell on the negative though. This particular problem is likely fixable. It is just a symptom of what has been a growing problem of inadequate or ineffective collaboration within the FOSS DE community, GNOME included. That problem is rather more important to address.
[q]They said the freedesktop org spec couldn’t possibly work, which might well be a reasonable opinion for all I know.[q/]
Since KDE and Unity is actually using it the spec obviously works. That’s not saying it works in the context of the gnome-shell though but it would seem like a design problem to me if data and presentation can’t be separated.
The GNOME developers are the one who made GNOME with their hard work. They are not perfect and they have their ego but without them there would be no GNOME at all and Qt would still be proprietary. GNOME with less interoperability is better than no GNOME at all. I believe we should acknowledge that they are doing good work before we criticize. “Thank you for the good work, would it be possible to work with freedesktop.org on this pretty please” is way better than “God damn developers, you suck for not implementing that feature, losers”
Edited 2011-03-10 20:31 UTC
No, really?
Sorry to rain on your parade but there’s no cause-effect relationship here.
It seems that all work on AppIndicator would have the copyright assigned to Canonical, whereas GNOME, being a GNU project, wants GNU to have the copyright.
Not quite right. libappindicator is a library that is one implementation (Canonical/Unity) of the proposed spec. Apps that use the library would no more be required to assign copyright that any other library.
More importantly GNOME didn’t need to accept libappindicator to implement the spec. They could have rejected libappindicator on copyright assignment or whatever grounds – as they have – and still implemented the spec on their own in GNOME shell. They didn’t. They implemented their own app/systray mechanism in GNOME Shell while effectively ignoring the fact that on fd.o it was known that folks were trying to implement a cross-desktop systray/notification spec. Now apps built to use the GNOME Shell systray/notification area may not work with KDE’s Plasma systray/notification area. And guess which party will probably adjust in the name of practicality? KDE. This is not the only instance of this kind of behavior and I can understand why KDE folks tire of it. I’m sure the KDE community and Canonical has areas for improvement as well. But, to Aaron’s ultimate point, we can’t keep pretending everything’s peaches and sunshine when there is real damage occurring. As a FOSS community, we can do better than this…
so… basically… they are doomed because they didn’t accept the spec implementation based on licensing they couldn’t agree with? or are they doomed because they didn’t put work into creating spec implementation which they don’t need?
ffs, isn’t FOSS exactly about having choice? in your words, they haven’t got one. wtf is wrong with the fact you deny developer right to his choice?
if you haven’t noticed, g-s was delayed few times now, meaning they have no time to work on spare work they don’t really need in order to put g-s out. some features even got cut out for 3.2. and still you demand work not needed to be done. hey, i also have some lawn that needs mowing. so, any gnome developer reading this, after you finish that spec implementation, contact me for further instructions
This has nothing to do with accepting someone else’s code or not liking the license. It’s about working with a cross-desktop spec. They would have been free to implement it however they wanted.
And yes, every project should be free to ignore whatever specs they don’t like. The point is not that ignoring the appindicators is terrible, but that Gnome is doing the same thing to practically everything that is proposed on fd.o. Essentially, the whole point behind fd.o is dead, because one side is ignoring it. Appindicators is just 1 hot topic that demonstrates this – the rule rather than the exception.
Oh please, that’s meaningless strawman argumentation. No one’s trying to deny any developer choice. The issue is quite simply the communication of those choices to others in the community who were demonstrably interested in improving the systray/notification area without break cross-desktop compatibility and while still allowing as much flexibility as possible. And no, no one can force them to communicate or implement anything they do not want to. But folks can certainly point out the accumulating consequences of this lack of communication and of the choices made.
The consequences are not just academic. In this specific case, there are real, practical consequences to app developers in the entire FOSS ecosystem. Many apps already designed to work with the spec won’t work properly with GNOME shell’s systray/notfication area. Apps designed to work with GNOME’s systray/notification area won’t work properly with other DEs systray/notification area. So sure, they can choose to do whatever they want and everyone else can just choose to suck it up and adjust once again to those choices in the name of practicality. However, I can hardly blame some folks for finally saying that they’re a little tired of it.
Feel free to carry on huffing and puffing about your lawn though…
meaning this is either gonna be the downfall or rise g-s. one way to weed out which one is better. but i doubt they would postpone release twice if they would have spare time and i also doubt there is any sense in postponing again every time something new shows up.
not that i like that fact, i can simply understand it as i am developer my self. i can also understand that sometimes i simply have to ignore or put on hold and follow my plan.
and to be truthful with slight sense of egoism, i don’t really care as i ONLY use gnome, anything not gnome is always removed from my computer. gnome has always served me 100%. beside the fact, i also really like g-s and i’m kinda pissed on waiting for g-s to finally come. but yes, this probably sucks for someone who is not using pure desktop.
I also prefer to use GNOME. But GNOME Shell in a test of it several months, mostly love and hate. I am also disappointed by the ugly icons in the dash and the default ugly icons that served GNOME all these years. Can you please tell them to hire some ICON/theme designers?
We already past the Windows 95 ERA ICONS! So please.
No such thing – GNOME is only loosely associated with GNU (despite the name), and certainly doesn’t require or even encourage copyright assignment to GNU (or anyone else).
I can’t see how Shuttleworth thinks bringing division within GNOME will help it.
Sounds like too me he is trying to divide GNOME and destroy it.
I am also not impressed with Ubuntu’s contributions to the whole Linux desktop thing. We need technical solutions to the desktop problems of LINUX in the areas of device drivers badly, and much less in the “marketing” area.
LINUX desktop’s are no longer an area where people have never heard of them. The problem is it is not entirely feasible for everyone to use LINUX and that comes down to poor hardware support.
I would like to see the guy enter through LINUX support at freedesktop.org with a 50 Million dollar grant to resolve the ludicrous display architecture issues we have with video cards and OpenGL/CL support for ATI cards.
You could hire 30 people for 5 years to work 24 by seven on the issue with 50 million bucks and the problem would get solved much sooner than that and provide a nice open source display architecture for LINUX and full 3D support on the fastest graphics GPU’s available.
-Hack
It may well require a slight touch of paranoia to suggest that GNOME Shell folks were afraid of competition so they deliberately implemented an incompatible systray/notification area. They may simply have been so caught up in getting things done they forgot or ignored the fact that there were other folks trying to implement a cross-desktop systray/notification spec. It was, at minimum, careless.
The hope now is that the community can start working together to fix this particular issue and fix/prevent other issues.
Oh, I think it’s deliberate enough – there’s not much doubt of that if you read the mailing list posts. It seems pretty clear that the Gnome people simply weren’t happy with cross-desktop spec that was being put together, concluded they couldn’t change it enough to be workable for them, and opted not to use it.
So yeah, for better or worse, it was certainly a deliberate design decision, not a case of forgetting or not knowing about other efforts.
I’m personally inclined to agree with them – Ubuntu have produced decent docs for their AppIndicators implementation of the spec, but the spec itself is extremely vague in terms of what an app that uses the interface can expect to happen. I appreciate Aaron’s point that this should be determined by the implementation, but the spec still should provide some guidelines.
This is a lot of trouble for two shells that (I believe) are not that attractive.
To both sides:
“My desktop computer is NOT a small touchscreen tablet or Netbook.”
I don’t like Canonical, Unity, Appindicators or KDE and I don’t care about anything they do.
In the end the desktop choices made by GNOME have suited me better than what everyone else has to offer.
same here
Me too…
me too, too!
Which has nothing to do with the issue at hand. UI design decisions (what you describe as “desktop choices” are NOT included in the spec. The only goal here is to make it easier for developers of ANY platform (QT, GTK, etc.) and every application to integrate into the YOUR DE of choice. Canonical/Unity, Compiz-dock and KDE adopted this spec (and GNOME was invited to do so) so that an app written in GTK would easily integrate into KDE, or a KDE app would behave like a native GNOME app (when it comes to notifications) on your GNOME desktop. It makes a lot of sense to cooperate on low-level APIs while keeping the presentation separate. A good example for such cooperation is KDE’s adaptation of DBUS, even though their own DCOP framework was more advanced/mature at the time. Thanks to that, we get more consistent behaviour across applications written in various toolkits. This kind of collaboration is good for the overall F/LOSS ecosystem – it makes users as well as developers happy. The issue here is that GNOME dev’s chronic NIH syndrome in the past few years sabotaged such attempts, diminishing the importance of freedesktop.org, for example.
Again, this is not about design choices of individual desktops, this is about consistency among applications. Consistency here does not mean that applications will behave the same way across all desktops. We don’t really want that now do we? You like GNOME and the way the GNOME desktop is designed, yes? Consistency here is within a particular desktop regardless of the framework used to develop the application. In case you ever need something written in QT or KDE, you don’t want it to be an eyesore on your beautiful GNOME desktop – you want it to behave like any other GNOME application. That is precisely the goal of these low-level specs – make it easier for developers to separate data from its presentation, allowing better integration of apps written different frameworks.
Edited 2011-03-11 04:47 UTC
You know what happens to dictators…
The linux desktop has long suffered with the GNOME vs KDE split, and here comes Mr Shuttleworth trying to create a 3rd desktop to divide the community even further. And what’s funny is he calls it UNITY!!!
Who cares about your freakin indicators….
Hey mamma look! We changed the maximize / minimize buttons to the left!!
How about trying to participate in GNOME from the inside, in a democratic way, like everyone else! But hey, you’ve already said that UBUNTU is not a democracy!
Guess what? GNOME is not your bitch… Get over it!
I present to you:
The attitude problem of the GNOME community.
most open source projects would really benefit from some criticism, abuse, and perhaps a few flying coffee mugs from a tyrannical designer with attitude.
good design is about attitude (among other things)! i use gnome because of their ‘attitude’, as you put it. it’s a design attitude that prevents gnome from releasing the kind of junk ui/ux that you see in most open source efforts. they take the process of design seriously. i can’t think of any other open source effort that has as much design intelligence and design integrity as Gnome. KDE and Ubuntu come up with lots of clever engineering and widgety things but their design just sucks.
You actually _prefer_ their design? D:
I have nothing more to say, as it would just be wasting my time.
you might have more to say if you had less chintz on your screen
Calico fabric over one’s screen _would_ impede vision…
Not that your post made the slightest bit of sense…
Does using the GNOME desktop all these years will include me in the GNOME community?
You might be saying “GNOME Developers” instead.
Peer moderation has been incredibly biased in this topic. If your opinion was “mainstream” your post would be modded up do the sky. If your opinion was against mainstream, your post would be modded down to hell.
That is why I didn’t post till now, and why my overall post count is much lower than it could be.
This is an old, known grudge of mine. Even if I have been told, time and again, that negative moderation doesn’t affect my Average Comment Score, I like to express my opinions freely, not under the pressure of moral blackmail.
Well, that was random-ish.
KDE did do the work to work well with the status notifiers, and as such, KDE applications ‘just work’ in Unity, all because the API has become a standard and has been adopted by KDE.
The status notifier spec was created by KDE Plasma developers Marco Martin and Aaron Seigo, then adopted by Canonical (I should know, I pushed within Canonical to adopt this spec instead of coming up with our own)
If you read the freedesktop mailing-list archive, you will see that when the spec was proposed, GNOME devs were interested but asked for some changes, which were refused…
Sigh…….
Alas, I think we can guess why. Gnome developers asked for changing thinking they probably wouldn’t get done, and when they were they simply didn’t accept what they hoped wouldn’t be coded in the first place.
I am not happy at all with the status of desktop Linux.
I categorically refuse to use KDE 4.
KDE 3.5, which was the best DE ever, of any OS, is hardly available.
I am compelled to use Gnome, but I am not 100% satisfied either.
Way to ignore user needs and to kill desktop Linux.
Well done, people.
And that is probably why proprietary OSes thrive. They can’t afford to say: “who cares about users”
Edited 2011-03-11 23:16 UTC
Right click on a KDE4 desktop, click “Desktop Settings”, click “Unlock Widgets”, on Layout choose “Folder View” and you have the same desktop you had on KDE 3.5 and more.
It’s not killing a desktop, it’s called moving on and advancing the technology. Get over it and adapt yourself.
No, sadly, “desktop as folder view” is not a good replacement for the functionality of KDE 3.5.x. That’s propaganda.
In KDE 3.5, I drag an icon to a USB device, I get the file. In KDE 4, I get the shortcut. It’s a kludge (with all respect to Aaron Siego’s coding skills)– it’s obvious that the plasma desktop was never designed to be a file space. Maybe that’s a good thing– Personally, I’m too used to the old way. We went from a “zoomable” interface to “activities”, which are understood well by asiego, but not so much by a large percentage of the userbase.
I could insert a standard rant about “work folders”, and how no desktop has gotten it right since OS/2, but it’s an old, tired argument.
I could complain about the fact that Plasma and Oxygen are two incompatible paradigms on the same desktop (why must I resize a dolphin window one way, and a folder view another way?!?), but that ship has also sailed. I can’t put one wallpaper across two monitors, which although minor, has been a personal irritant since KDE 4.0.
KDE 4.6 is finally getting somewhere, although the contortions akonadi puts me through make me grit my teeth.
All in all, both Gnome and KDE seem to have turned their backs on what the average user wants, and are determined to ram “their way” down the consumer’s keyboard, whether we want it or not. KDE has spent the last two years improving that original stance, whereas Gnome is still impaling themselves on their sword– so of course KDE comes across as more reasonable.
Hey Aaron, if you’re reading this… what’s the use case for rotating a plasma widget? I haven’t come up with one that isn’t a gimmick. And I’d still like to be able to shrink/iconify a folder view and/or activity to a single widget.
Exactly. That is what I say to myself: KDE 3.5 was a desktop for users, KDE 4 is a desktop for geeks.
You have managed to explain very well what I was trying to say. However that doesn’t explain why I was modded down twice.
That’s because it was basically a Windows clone. KDE4 is what you get when a bunch of hackers with no UI experience or taste try to innovate
I am not sure about that. What I am sure of is that open source developers should never forget their users.
The average computer user is not a geek. And even geeks often prefer to get something easy, so that they can be as productive as possible, without effort (from personal experience, from many friends who abandoned Gentoo or Slackware for something more user friendly.
In my experience, non-geeks like KDE4 just fine. It’s the hard core KDE3 lovers who can’t stand it. They definitely tend to fall in the geek camp, more often than not. And of course, lots of Gnomers – some of them are geeks and some not, but their defining trait is more about loving Gnome than anything else. It wouldn’t matter what KDE did, they wouldn’t be satisfied.
Guilty. I still don’t have KDE4 on my machine
I’m not Aaron but I believe that’s a functionality they added for form factors like touch displays, etc. So you can move widgets as you wish with your hands, and rotate them, etc.
I agree to an extent: have just this weekend returned to KDE after several years of using GNOME at home (I still use it on a work-based machine, just in case anyone thinks I am being wilfully partisan). I had moved away from KDE as a disgruntled 3.5 user who wasn’t able to get over himself. I understood that KDE has been developing so thought I would give it a go. So far, I am happily surprised – someone above called it a ‘Windows clone’ but interestingly I have installed a KDE-based distro on account of two days of hair-pulling realisation that a lot of the foibles of Vista are still buried in W7 – and in that sense, what’s wrong with a Windows clone with some genetic augmentation !
Equally however this new experience of KDE means a lesson for me when it comes to evaluating GNOME 3 – count to twenty (or more), listen to the arguments (not your own cherished opinions), and try to work with what there is, instead of going off in a huff!
This article seems a bit one-sided to me. Gnome has accepted and incorporated various projects into their core product, and its community is open for everyone to join.
There is a number of reasons why libappindicator is not included in the Shell, but perhaps the most essential one is that app indicators simply don’t fit into the Gnome Shell interface, where all the system-related components are presented in the top panel and all the application-related components are in the message tray.
For a well-documented, yet wonderfully readable development of the libappindicator issue, read this: http://bethesignal.org/blog/2011/03/12/the-libappindicator-story/
My understanding of the spec is that it is up to the desktop shell to provide the visualization of the data sent over dbus. So it the desktop shell wants to show system related components in in a top panel and application related things in the message tray that is fine. The spec doesn’t mandate that everything should be shown in a single panel.