I’d just post a link, but I’m pretty sure the olverhelming majority of you won’t bother of clicking it.
This post comes from Planet Compiz, and it’s written from a Compiz Developer.
Maybe it’s finally time we put some serious consideration into a Compiz Desktop Environment. With GNOME pretty much throwing us aside on support and with KWin replicating our features left and right, it’s hard to stay on board. This all comes on top of the numerous issues we’ve always had with the two big ones.
So, I figure it’s time we look a bit bigger. As 0.9.0 starts to build back to the status of 0.8.* in plugins, we should look at things we need to build a solid, effect-filled DE. Sam mentioned the possibility of forking GNOME, but do we really want to go that route? I can see the backend systems in GNOME being reused in the “CompDE”, but I don’t see us forking their panel and desktop: we don’t need to. Compiz has always been an add-on to existing DEs, and we need to consider that: We can build the basics, we already have. We have numerous panels, and we’re working on a deskop. The rest is integration systems.Hm. I’m limited on time here, so I need to wrap it up. I’ll try and make a more suitable argument later when I have time.
And:
I only mentioned forking gnome-shell, not all of GNOME. GNOME isn’t interdependent on shell, so this could be done.
However, even forking shell, let along maintaining our own DE will be problematic. Remember that we have like what, 5 developers? And we all have full time jobs/school/university? We barely have enough time to maintain the compiz stack, let alone our own DE. (I barely have enough time to get all my homework done these days!)
Unless one of us wins the lottery, gets a job with google/redhat/canonical/novell or we put a ‘donate’ button on our site and we end up getting over $5m, somehow I doubt whether this would be physically possible.
I’d hate to see compiz locked out of GNOME due to some silly little design desicion, but the GNOME guy’s are basically telling me ‘Compiz sucks, nobody cares about you, users never cared about you’. Which is completely wrong. Oh well, if people want to use compiz, they will have to use KDE/QT. GNOME’s loss.
I won’t add much to this, but I really prefer Compiz to Metacity right now, mainly because of many little usability enhancements, and for the incredible configurability (and yes, it has a full GUI devoted to its configuration). If they switch to KDE or whatever, I’ll probably follow them.
You may want to read the original article to check for any update or to discuss with the author.
Update: Kevin answered some questions made by some of you. Don’t miss his comment.
Now, I don’t use Gnome much. I’ve always preferred KDE (nothing against Gnome, KDE is just more to my liking) but I can honestly understand why the kwin-devs decided to go their own route.
Why? Because in my experience, compiz has never worked as well in KDE (v3 as well as v4) as it has in Gnome. Maybe it’s just me, but there was always an issue here and and an issue there that made Compiz + KDE a royal pain in the nether regions. In fact, it wasn’t worth the effort. Granted, I haven’t tried in a while so it may be different now. (and I hope it is, for those that prefer compiz to kwin). And no, I don’t know where the problem is/was. All I know is that I never had the same problem(s) in Gnome or with KDE + Kwin instead.
Still, I don’t really miss compiz anymore (maybe some beryl decos…) since the features I used are now a part of Kwin and much better integrated. Actually, I think the biggest drawback of compiz is that its emphasis on effects, while nice, tend to overshadow other aspects of a window manager. Making a window into a paper-plane when closed? Honestly? Yes, I know you don’t have to use all the features but it still feels like they should have diverted some more effort into other aspects of the window manager as well.
Compiz is needed at the moment because it the only decent, working, hardware accelerated compositor for gnome.
With a half modern PC you’re crippled running Metacity because it’s dire and there’s practically no noticeable hardware acceleration present.
XFCE et al are faster but the compositors are equally as bad as Metacity’s.
Gnome needs a compositor that is fast and makes good use of today’s cards and not last century’s relics.
Compiz almost fits the bill. Where’s the competition?
@ste: the only competion I know of is Kwin. Which KDE-only of course and probably born out of necessity (see Jonas comment).
Even if Gnome were to improve their compositor I highly doubt that they would ever bring all the goodness Compiz brings, in terms of plugins and configurability (they’d just put a configuration window with 3 options and tell you how easily you can hack GConf if you need other things).
Even further, I’m a heavy user of Screen Edges and Expose and I can’t see gnome implement those features any time soon. I like gnome, but I really think they hide themselves behind HIG too much.
Even the author of the article still hasn’t finished it, I guess, he was referring to the gnome shell. So I watched the demo videos ( http://www.gnome.org/~otaylor/gnome-shell-talk/ ) and would like to tell why imho it hinders more than helps:
1) There are already various great tools for searching apps in linux (gnome-do, meta tracker, beagle and so on);
2) Compiz expo is superior to that workspace switcher used in gnome shell;
3) There is no need for “recently viewed docoments” list because many applications can ever save sessions or show a list of files recently opened with that app;
4) That slow slide effect will be really annoying to most users who just want to quickly open some app;
5) Compiz scale plugin also makes more sense than that scale effect in gnome shell which only works when previewing workspaces.
Anyway, as we talk about gnome in general too and I have used it and watched its development for quite awhile, would like to share what annoys me most about its development. So here are my points:
1) It seems that gnome devs care even less about stability than KDE. Examples? Gnomes session NON-manager is a perfect example – they released “STABLE” 2.24 version with completely broken session manager ( http://np237.livejournal.com/22014.html ) and if that’s not enough for you, 2.26 session manager is still quite broken – they plan to have correctly working session manager in 2.26.1 ( http://bugzilla.gnome.org/show_bug.cgi?id=552387 )
2) Even there were some serious stability issues in KDE, at least unlike Gnome they move forward. Now in gnome, imho, the only recent innovation worth mentioning was introduction of tabs in nautilus, but, you know what, they failed this time too – after the release the feature wasn’t really usable because of major bug ( http://bugzilla.gnome.org/show_bug.cgi?id=542396 ) and it took half a year to fix it.
3) Another reason why gnome seems to be stagnating is that gnome bugzilla is like some black hole where user submitted patches die. Here are a couple of cases, I think worth mentioning:
a) Emblems and Icon size in Sidebar Place (the patch used to work with gnome 2.22 but now needs to be rewritten because none of the devs noticed it):
http://bugzilla.gnome.org/show_bug.cgi?id=477470
b) Support for compiz wallpaper plugin (same fate):
http://bugzilla.gnome.org/show_bug.cgi?id=444320
c) Composite support to main-menu (same):
http://bugzilla.gnome.org/show_bug.cgi?id=419840
I have also filed bug reports for meta tracker and after about a year they just closed it because it was already invalid and nobody bothered to even comment my gthumb bug report even more that a year has already passed.
4) It seems some main developers think that new features might confuse users more than help them even the feature was already requested by thousands of users (at least judging by ubuntu brainstorm clicks). Examples:
a) Split view file browsing (patch proposed)( http://bugzilla.gnome.org/show_bug.cgi?id=309646 )
Mailing list: http://www.nabble.com/Split-dual-pane-view-again-(but-this-time-with-code)-tt22099182.html
Video: http://www.youtube.com/watch?v=yqBDEi17l2A
b) open desktop folders in new tabs:
http://bugzilla.gnome.org/show_bug.cgi?id=550909
5) Some projects seem to just die silently (even sometimes they are Google SoC projects):
Different wallpapers on different workspaces:
http://gsocblog.jsharpe.net/archives/15
I was going to blog about this if I’d find any time but you beat me.
I hope canonical is going to act on this because they use compiz in their gnome edition if it can run (it’s nice eye candy). Also I expect a Novel reaction because they use compiz in all their openSUSE commercials (I hope they’ll use their connections and power for the common good one time).
My opinion: I love gnome, but I don’t love it if it doesn’t have my compiz.
Wouldn’t the ideal be to integrate Compiz into Gnome instead? Merge Compiz and Metacity? better desktop effects/enhancements for Gnome, the whole of Gnome backing Compiz. KDE have Kwin so they really don’t need Compiz anymore, and you don’t run XFCE or ‘box to have fancy 3D effects.
And Compiz already have a plugin-model, so the standard Gnome desktop wouldn’t need all the stupid stuff like paperplanes, waterdrops and other proofs of concept.
But Compiz still doesn’t work with my Ubuntu and Nvidia-card
@audunmb
On the Compiz(-fusion) planet you can read that the gnome people responded very rudely to the compiz guys when they said that the changes to gnome-shell broke things. They claim that the average user doesn’t care about them etc. (I think that’s wrong, compiz eyecandy is the trick to get new users to TRY linux, and then they fall in love with everything about it (including compiz).
As Bert_2 said in a previous comment, I am pretty sure Novell and Canonical (and even Fedora) won’t ever let down Compiz, as it’s something they rely so much to get new users and publicize Linux (see for example the Novell commercial cited in this post).
Still, is saddening to see Compiz developers feeling set aside.
It’s not every day I get quoted by someone syndicated on the Ubuntu Planet, and far less often am I quoted in agreement. I’m on a TCP-over-DNS tunnel to my server at the moment, so much like my blog post, this is going to be short and probably seem like rambling:
Maybe nouveau will support it…
- Yes, the article is referring to the direction GNOME is taking with gnome-shell. They’re doing whatever they can to integrate key parts of the desktop into the shell, leaving other window managers (like Compiz) to fend for themselves. While there are an abundance of replacements for the things gnome-shell is trying to bundle, but that doesn’t go very far: once you start replacing bits of gnome-shell so that you can run Compiz, you’re getting into the realm of “not even running GNOME”.
- Honestly, yes, a merge /into/ GNOME would be great – the KDE guys are doing a fine job replicate parts of Compiz, and since they’re not just pushing in parts of the desktop we still respect them (though they /really/ need to fix the reflections in their cube, have they done that yet?)
- I’ve applied for GSoC with a project that aims to add a feature to the GNOME desktop that people have been asking for for quite some time (multiple wallpapers in Nautilus, sound familiar?); while my primary goal would be to do what I describe, I really want it to be a connection between GNOME and Compiz and try to settle some of these disputes.
- audunmb: What card?
- Has anyone read through all of the GNOME HIG? There’s some really… odd stuff in there.
As I’ve said, I’m on a bit of a thin pipe (not to make anyone jealous, but I’m in Hawaii, but my hotel does /not/ have free wifi…), so I’ll have to adjourn. I’ll edit this (and a few other comments) into the blog post some time this week.
Gnome was my first Linux desktop, but now I’m seriously considering switching, especially if the compiz team direct their efforts to KDE or their own fork of Gnome.
Metacity is old. Dead. Decrepit. Much of the fun I’ve had with Linux is thanks to Compiz.
I take offense to the Gnome developers comments that no one cares about Compiz. From what I’ve been reading, no one cares about Gnome. Take that!
@Kevin: yes I did read Gnome HIG. Some stuff is weird, some other stuff is just often took as an excuse for lazyiness. Thanks for coming here and clarifying things a little bit.
Here is another gem for you – Linus Torvalds writes utility for Gnome, but it’s rejected because it’s “too complicated for users”:
http://distrowatch.com/weekly.php?issue=20090202#feature
LT: Since I’m on Fedora, I got hit by the (bad) transition to KDE4, and as a result I’ve been using GNOME for the last year or so. It’s still somewhat painful, more so when I’m on my laptop, mainly for the same old reason: you cannot fix the mouse buttons in GNOME. (The reason this hits me more on the laptop than anywhere else is that most laptops only have two buttons, making the middle-button press much harder. And middle button is what you need for the ‘send to back’ window action.)
I wrote the patch (including even the graphical configuration management), I sent it in, and it got rejected as “too complicated for users”. Frickin’ idiots (and I’m not talking about those alleged users).
But right now, KDE is worse. I’d like to explore alternatives, but if you’ve followed my answers this far and are perceptive, you’ll probably already have figured out that the programs involved aren’t on my list of things I care about that much.
I’m well known for disliking GNOME, but it’s not the “using it” part that I dislike as much as the apparent mentality of the GNOME people who think that all users are idiots and then limit what I can do with it for that reason.
only can says is kwin is not stealing code from compiz:
kwin is “c++” and compiz is “c”, is not so easy like copy-paste
and most of the “plugins” of compiz really share code with the core of compiz(plugins??), this is why kde dont use compiz in kde4, just read the kde-tech-base docs.
bye
Well, the compiz guys are transitioning to C++ as we speak, compiz++ will be the default for 0.9.0 I believe (some plugins have been rewritten but the most difficult, like cube and wobbly aren’t done yet).
But of course you’re still right about the code sharing
@bypasser
While some of your statements do have a point, I don’t agree to the
big picture that you’re drawing.
I am the author of the Nautilus split view branch that you linked
to. The situation which you depict is not quite correct, though. The
announcement mail of some draft work yielded a lively and useful
discussion. I then cleaned up the patch and modified it according to
the suggestiong that came out of this discussion.
The patchset is non-trivial, pretty big, and affects core parts of
Nautilus’ codebase. So, before formally requesting a review (knowing
that there is manpower missing in the Nautilus project in the first
place!), I’d like some testing from other users except my own desktop
usage before requesting a review. As people (like you) try to make
believe, there’s thousands waiting anxiously for that very feature.
I’ve announced it in various places, and blogged about it a few times,
always asking for testing support. And now, guess how much feedback I
got. One comment that it works (though not stating how long it has
been used — my guess is that it was freshly compiled). Well, of
course, I got a few “will try it now”, or “can’t wait to see it in my
distro” (both of which is nice feedback, but neither of which is
technically helpful), but not a single “have been using it for a
month, it works flawlessly” or “found the following bug”.
Of course, that’s fine, nobody has to help out. But please, don’t use
my work to try to accuse the Gnome developers of ignoring users’
wishes or polished patches. It’s (in my case) simply not true, because
neither did upstream say split view was too complicated and will be
rejected, nor was the patch ready for inclusion.
Surely, there may be polished, much-wanted stuff rotting in Gnome’s
bugzilla. But in my experience, it’s either a lack of manpower for
review or outside the project’s scope (not every application wants to
become emacs) if it doesn’t get applied – and not arrogance or
indifference of Gnome developers.
By the way, Torvalds’ patches for mouse-button modifications are in
Metacity since a long time now, so I don’t really get that part
either. While we’re at it, Torvald’s very own views on accepting or
rejecting patches are an interesting read, too:
http://lkml.indiana.edu/hypermail/linux/kernel/0210.3/2128.html
http://lkml.indiana.edu/hypermail/linux/kernel/0210.3/2141.html
I think Compiz was one of the most innovative opensource projects just a couple years ago, I still like it very much and use it, of course. Anyway, my question is “to create a DE is the next level into the Compiz development”? I don’t think so, I think they must be more creative and follow the path that they have been following til now. Maybe when this sudden desperation is over, they will think clearly about their future, and ours like costumers.
(Sorry if my english is bad).
I like the news. A Compiz DE will only add competition and maybe a wake-up call to the other DEs.
As for the debates about the problems with Gnome, I honestly believe it’s just a problem to do with lack of manpower. A lot of people working on Linux aren’t do it as their first job, just like the people doing Compiz. So it’s expected for development to be slow, and most radical non-essential suggestions and features to be rejected because they simply don’t have the time or manpower.
The thing is, down to the user level, if one was smart enough or brave enough to use Linux (switching from Windows), let alone Gentoo or Arch Linux (horray!), one should have the ability to customize or change anything they didn’t like. I started off with Ubuntu for 2 years, switched to Arch Linux and tried KDE for one year but went back to Gnome. It’s all about personal preference. You tend to not rely so much on ‘developers’ but yourself. You ‘develop’ your own system.
It’s not about Gnome thinking all users are idiots. Yes, bugzilla and the HIG need some tidying up, but for a small team (small compared to the gigantic Microsoft Coorporation) to manage these things is very tasking. But if the issue is a low priority or if implementing them requires many code-changes, then they’ll have to re-think it.
@ hb
I’ve been using your dual-pane nautilus patch since I found it last summer. For me, it worked flawlessly in Ubuntu 9.04 and continues to work flawlessly in 9.10. I use it extensively in my day-to-day operations and I’ve never noticed any bugs or hiccups of any kind.
Im using Gnome Shell right now only be using it for like 2/3hr’s i like it but it has a long way to go before i would put this on my laptop they need to integrate gnome do more in to it so its easier to add programs to it and add options so we can change things maybe add more effects if they where smart they would build compiz in to it and save them having to do the work on any effects
I think Compiz will probably be kicked out of the Linux environment entirely. The problems with such an event are too numerous to count. Firstly I converted because of COMPIZ and the fact it was actually usable and fun (compared to Windows, etc.) But without that as an option I’ll be weary of staying with Linux since independent developers for platforms like Windows are already starting to clone their own versions like CubeDesktop or DeskSpace. As they advance their versions even though very different, Windows can only benefit. It is Linux that suffers as it will then be even further behind in desktop environments and when the next person considers changing they’ll say “hey there’s no real flare on Linux it’s boring.”
Integrating Compiz into Gnome is lame, it’s basically already done once you turn on Compiz. What they need to do is completely trash the Gnome interface in it’s current form and take lessons from the Compiz team in designing interfaces. What user doesn’t want a “colorful experience” when using their PC?
As for this nonsense about Compiz being unstable, lol It’s very stable as long as your drivers are stable, it’s got nothing to do with Compiz itself. That’s like not having drivers for your videocard installed then claiming games don’t run so their shit. Lamers make this argument which is a LIE.
Frankly if I were the Compiz team I’d say screw Linux and just stop development since nobody seems to care about them except a handful of people.
Gnome really need to incorporate compiz into gnomeshell, if they don’t, they are complete morons.
Wouldn’t that be something? gnome and compiz to become one. It only makes sense. KDE has kwin – and good for them. I would hazard a guess that most people use compiz over metacity. But it’s not going to happen, however open minded the compiz guys are the gnome folks are not!. So I put it to gnome like this: get compiz on board or risk loosing a large user base to compizDE when it surfaces. I mean for the love of jeebus, there is no point reinventing the wheel! why redo your own effects when you can get into bed with the experts over @ compiz
.
if GNOME were run by socially cognizant people, they’d find a way to funnel support toward Compiz and make them part of the package. It’s silly that people want to reinvent the wheel so often.. the excessive forking that makes up the canvas of the gnu/linux scene is really convoluted. It took a lot of self-determination to put up with the confusion when I made my switch from Windows.
Anyone complaining about Compiz having workarounds doesn’t know how software development works.. especially when you have such a ragtag squad that aren’t being given the respect they deserve.
I find the Shell idea boring and annoying. Talk about fragmenting more and more.. they should be trying to figure out a way to unite Linux under a more common flag. Sick of all the G-this K-that bullshit.. at this point, it’s worse than WinSuchandsuch.
Maybe Compiz should focus on working with the smaller DEs – XFCE, LXDE, Enlightenment and others.
The developers are likely to welcome the interest, as they have limited resources themselves, and are unlikely to want to maintain their own equivalent of functionality there is a perfectly good third party component for.
Its sad how Gnome is not teaming up with Compiz… I found Gnome Shell a great idea but lacks a lot of Compiz “fun” and customization… No wonder why Linux does not grow faster,,, as Dave says “they should be trying to figure out a way to unite Linux under a more common flag”
I switched to Linux because of Compiz by the way!!
I switched to linux because of compix also. I’m finding the lack of support for it to do what it used to be able to is simply going to drive me back to where I came. What has happened to linux?
Dan
Ubuntu is keeping Compiz by default with Unity. Can’t say I like Unity, but atleast they chose Compiz over the sluggish featureless Clutter (Mutter) compositer
Gnome3 in it’s current state on the Fedora 14 alpha release is the most sanitary, academic piece of crap I’ve seen in a while.
I’ve got Compiz with a 6 sided ‘cube’, expose, scale (for all windows) and a simple alt-tab window switcher (for current) on the traditional Gnome desktop with two auto-hide panels of application shortcuts.
It’s all fine an lovely having a brilliant menu to start new applications… although it’s not the starting of applications that I need help with, it’s how best to organise and view the ones I’ve got running. I need the applications grouped into task domains that I can switch between in different contexts and I want to decide how to do that!
Gnome3 so far, aside from not providing the pretty visuals of Compiz and not providing us with the configuration options of Compiz, seem to have completely missed my persona from their use case diagrams. As far as I’m concerned I’ll be finding a work around to Gnome3 if it doesn’t improve significantly.
Much love to the Compiz team for increasing my productivity with an utterly beautiful solution. Bonus points for excellent by-products! +10000000000000000000000….
Compiz+Panel+Desktop icons/folders + SomeOtherBits=”Compiz Desktop Environment”
GET CODING, LAZY COMPIZ DEVS!!!
I would be interested in helping with a Made For 3D Compiz Desktop Engine … You could accomplish this with a Kinect and a 3D TV =D really go crazy with a completely interactive touch and move environment and just blow Gnome out of the water
I am a little bit out of date to this topic since I don’t really care about the window manager until fedora released 15 and chose gnome 3 as the default window manager. Then all of a sudden, my Compiz tricks are no longer working in gnome-shell. The google search engine links me to this thread here, and it is sad to know that gnome3 and Compiz developers do not get along very well.
By the way, the Compiz tricks I use are really helpful to me to fulfil my daily tasks. They are not available in gnome3.
IMO, Linux and Compiz rocks because there are many ways that users can configure their Linux, unlike the way how Windows does. However, gnome3 took away many configuration possibilities, even including Compiz. Up to now, all the Linux users I know who uses gnome have the similar feeling for gnome-shell. Gnome developers think they know better than everybody else, we will remove Compiz since nobody cares about them. I mean, oh my god, this attitude could eventually bring gnome to dead, seriously. All the people who used gnome so far thinks they will give KDE or other window manager a try. Wake up, gnome developers…
I definitely will use gnome3 for a while, because I do a lot of my work in console, I don’t really care about the window manager.
@compiz rocks
ubuntu unity interface that comes with the latest version of ubuntu gets along quite well with compiz (Canonical even hired one of the main compiz developers). Also latest version Linux Mint (11) comes with gnome 2.32.
Unfortunately, switching to another Linux distribution is the last thing I want to do, or for most people I know want to do. The Linux distribution I have been using since the first day I knew Linux is Fedora, and I loved it. You can say I am loyal to Fedora, and at some extent, I am.
Somebody please suggests Fedora to hire one Compiz developer.
I must be MAD! I posted here last 4th of May last year!
Well, since I’m an avid gentoo user I gave XFCE 4.8 a whirl, yes totally left gnome, recompiled everything -gnome (minus gnome), and I must say I really like it, been using it for 2 weeks now and I’m happy! good job gnome!
Compiz DOES suck, fact. Now that Unity is using it, that also sucks a fat one.
@sdfgsdfsdfsdf Troll much?
what if I like both compiz and gnome?
I hope people in gnome development change their feeling about compiz. At least they use the 3d cube desktop in new gnome, it easy to move around workspace in compiz 3d cube, usefull for facing a lot of windows, just scroll the mousewheel, right?
but gnome shell is more clear and not make my eyes frequently blinking,
yeah, that’s it, I hope compiz still can used in gnome in the future.
I am a partially blind user, and I will say (though this post is a little late compared to others) that I could not use my system efficiently without Compiz.
I completely rely on the eZoom plugin for screen magnification, and it is THE reason I finally switched to Linux in 2007. The GNOME magnifier (with no offense to its developers) is inferior in quality to the Compiz magnifier.
I tried to use both GNOME Shell and Unity with Ubuntu 11.10 recently, and BOTH of them were not accessible to me using eZoom. This is not by any means the fault of Compiz.
To those who are working so hard on Compiz, I care for compiz and I depend on it for my computer use. It is a great shame that users and developers are being tossed aside so that something “new and flashy” can be developed to appeal to new users. I am not against new developments, but when these new developments throw aside loyal users and developers then a problem arises.
I really don’t know what to do anymore. Ubuntu is no longer accessible to me, so I do not know where to go. I won’t go back to Windows, so I will find someplace where I belong as a user.
You guys keep up the good work. It is not in vain.
R. Cole. I recommend then looking more into source distros so you can build up your computer to your needs… eg. Arch linux is a very nice bridge between source/binary.
But needs studying the archwiki when you are shot. Although if you wish for solutions out of the box, try the newest mint version perhaps?