Misinformation and miscommunication

A while back I said a few (looking back on it, rather harsh) words about gnome-shell. I’ve probably miscommunicated what I was trying to say with regards to that situation.

First of all, such comments are (roughly) my own opinion and my stance has changed to be a bit more neutral since then – of course with the status quo and the way GNOME-Shell is written, it would be difficult to make it WM-agnostic as it is now. Looking at the code the job may be a little harder than I thought. Of course, it’s important to note that though I am a member of the compiz council (an ill fitted one at times), usually my comments don’t represent the entire view of the compiz project (even though it might look like it).

Secondly, I get the impression that people are starting to think GNOME is some kind of evil project. They’re really not – and I think GNOME-Shell has a lot of merit to it and is certainly a step in the right direction for bringing a composited desktop to the masses.

The only concerns that I had outlined with it was that it integrated the panel with the window manager. This has it’s merits and can make things a lot easier, but also means that if you were to start another window manager (i.e compiz) you would lose your panel. There are replacements for the panel (like AWN, Kiba) but for people who like the normal panel (like me), this isn’t the best scenario.

I guess the best thing to do now is to let matters cool down and cross the bridge of the issue of the panel when we get to it.

So please – don’t turn this into some kind of holy war, because it really isn’t. I always try to live by the rule ‘Assume best intentions’ i.e people aren’t evil for the sake of being evil – there is always some good intent in actions. The shell issue will sort it self out, it’s just one of the many little things that should probably be addressed in the future.


32 thoughts on “Misinformation and miscommunication

  1. I don’t get it. Why can’t gnome-shell handle both, but when it detects another window manager starting up, it’ll stop managing windows and switch to managing only the panel? Or is it too different for that?

  2. I think it’s difficult for Compiz to find a role in the future. Initially, it was the only way to get cool desktop effects, and could be dropped into almost any environment. But with both Gnome and KDE looking to integrate compositing features natively into the desktops, it’s hard to see quite where Compiz fits in. Getting them to make things modular enough to support Compiz too doesn’t really work, because it *does* make their job more complicated for (from their perspective) no real benefit.

    1. For this reason, Compiz probably has little choice but to become a desktop environment in itself. Whether that means forking Gnome, working together with some smaller window manager project or building one from scratch, it doesn’t matter. I know the council doesn’t have the time or resources to support a full desktop environment, but they may have little choice in the future if they want to stay relevant.

      I personally think the best choice would be to fork Gnome, every major release, and document every change you make to make it Compiz friendly. Some of those changes have been long coming, such as multiple desktop backgrounds and suppressing startup to allow apps on launch to launch with Compiz instead of before it. With a little bit of effort, the Gnome guys could see everything from your perspective, and with the patches you provide, might end up cutting the fork work in half by implementing the small things into the core. If done right, distributions may end up choosing to launch with your version instead of Gnomes, and that would really light a fire under them.

      1. Forking GNOME as a whole is ridiculously drastic – not only would that make us look completely inconsiderate and uncoperative but is also impossible considering the 10mins people like myself get a week to work on compiz.

        I think a more elegant solution is required. GNOME isn’t some kind of evil project that wants to take over the world – and we shouldn’t treat them like that. I was only fired up because of a previous comment made but I believe it has been retracted.

        If it comes down to it, we might be able to get some kind of panel/shell spec that could be implemented in a similar way across both projects. The idea would be make applets not know the difference between the two panels and that is mostly already possible (to an extent)

  3. @Micheal W.

    Unfortunately, code wise – it was going to be a little harder than I first thought. shell is based on clutter, runs with the window manager process and because it’s based on clutter we would have to change compiz to be based on clutter.

  4. Everything that gnome-shell is trying to achieve within the next 5 years, will be doable in KDE4 with better backends, design and freedom within one year.

    Please stop wasting time and money while pissing on linux’s identity with crappy freedom of choice and deprecate gnome already. Contribute to kde4, it’s way ahead of what gnome-shell will ever be.

    (Gnome-shell is rather stiff, hitting aside a lot of projects like compiz, AWN, gnome-do and so on, duplicating effort once again)

  5. Actually, IMHO, compiz devs should just cut the crap and contribute to KDE/KWIN directly… or even GNOME, as long as they do things properly.
    I don’t see any good reason to have compiz as a ‘3rd party’ implementation, no wonder why GNOME is not taking you guys seriously.

    I don’t want to be the bad guy, but we need to wake up and start making some *real progress* with linux… http://dreadknight666.com/?p=19

    1. Oh hi KDE/Gnome, We’ve decided to work on your project. Our compisiting manager is far superior to yours, so we’re just going to turf all your hard effort and jam ours in there instead.

      Oh yeah, I can see that going down well. Code doesn’t get duplicated by evil demons, it happens because someone seriously believes they can do something better than someone else. If the Gnome or KDE people thought Compiz was better than theirs, then THEY would approach the Compiz people to work out intergration, not the other way around.

      Unfortunately for us, neither KDE nor Gnome like Compiz, so their efforts will be focused on their versions. This is how code duplication is born.

      1. Quit working on compiz and just help out KWin then 😛
        Don’t see any point in compiz if it’s always going to go against the wave, seriously.
        It was nice for it’s time, but devs need to know when to end certain things (like adept was discontinued).

    1. That’s just acting stupid. It’s enough freedom of choice that you can pick something else besides windows, mac, freebsd.

      I don’t care about them. They should just spend the and get along with KDE. It’s not that hard. All are basically desktop environment with applications doing pretty much the same thing. WIth more support, contributions etc, KDE can satisfy more people, because it’s very powerful and allows for a lot of customization, while gnome-shell is very stiff.

      Those who don’t want kde can go download windows from some website or get themselfs a mac. We need a clear identity on linux, a desktop environment that’s the best thing and less crappy choices and applications that don’t get the love they deserve nor meet their average potential.
      It should be about running linux or not running linux. Users barely contribute to anything, so if they want a DE for lame reasons, they don’t really help it anyway. More people tend to make more distros and from time to time a new DE appears. We need to believe and care for the same thing, not let people piss on linux’s identity because they just don’t want to get along with a distro or some packages. Let’s be rational for a moment.
      But damn geeks… always unsatisfied with stuff, too much free time and no real hobbies… we need a webbrowser with webkit to be able to surf any website, a music player that lets you easily explore all your music, a video player you can see your movies even with subtitles, a nice image viewer to explore your pictures, an instant messenger to talk with people and not only by IM, but audio and video as well, because the 70’s are long gone and they’re not coming back.. and some other applications doing a specific purpose. It shouldn’t be about the application, but what it does and what you can actually accomplish with itt.
      We don’t need 2 toolkits to make 2 average applications doing the same thing more or less.

      The main problem with distro is that they rebrand linux again and again, creating confusion with no real arguments, i say those people are confused.
      Linux should allow you to easily configure anything, from looks, packages, workflow etc. without you having to always consider another distro and spend ntime investigating and comparing stuff like that. Distro watch much become deprecated.

      1. Yes, let’s not feed him. People are better off using windows and mac, if paying for them or not.
        Linux really needs to be threated like shit by people misunderstanding freedom of choice just because they like to pick between 10 average file managers instead of just having one that’s fast, stable, full of features, secure, nicely designed, customizable and integrated.

        Anyway, I’m sick of the whole ‘windows sucks, freedom is so awesome’ crap. Go play with your terminal and compile some kernels.

      2. Dread Knight – talk about heart and soul. look, there’s much reason there. and yes that is the road to having what pretty much constitutes a winning product.

        the following isn’t a rant directly into your face DK. the thing is, the crowds which are drawn to Linux for the idea of NOT being forced into a “system” … that one player, one browser, one editor mindset. even if they/we end up being dedicated/used to just one app/kernel/distro/whatever, the idea that we could switch to another one which is just as good is what gives that feel of freedom. however useless it is, however “perfect” the one we’re using now is, there’s more out there.

        an analogy: you are given the choice to live on in this life, or you could be plugged into a machine which will feed you emotion, vision, taste … all the perfect ones, all amazing, all suited perfectly for you. choice one is organized chaos, but provides endless possibilities. choice two gives you one taste, one emotion, etc. now you may already know the answer. i know i do. and if you or any reader don’t/doesn’t, no worries. it’s a difficult choice. my point? i have 2.

        1. Not everyone will make the same choice.

        2. This analogy, when applied to choosing an app/kernel/etc changes to not being able to cater to each individual to their perfect idea of perfectness. and what’s good to you is not necessarily good to me.

        when you present me with an OS, i may dislike one simple feature which will bug me to no end, and a few others will dislike it as well. we will band together and come out with a better product which makes us happy. a few other people will like it as well, tell their friends, and we have a challenger to whatever the mainstream system(s) are at that point in time. sweet deal. but of course some others will be even still unhappy with the current choices, complain, and they (or someone else) will create a system which will do better for their purposes. here i am again saying a few things, which pertain to the development of products which perform differently than the norm, and to the need for developers and complainers alike. otherwise we get one product, the non-stagnancy of which still confines it to one system which is upgraded, but remains one system nonetheless. the open source situation with which computer users are presented provides the necessary platform for open human creativity, which inevitably results in change and inevitable division of people into teams with/due to differing opinions. this is only human, and no amount of that species’ bantering will prevent/compel the development of a SMALLER number of kernels and apps.

        not that it’s not necessary to pop out a sweet product, but after that’s done, move on and make something else. things will only get better. and they’re moving fast enough even if it seems like they could be moving faster.

  6. i don’t know the ideas about gnome-shell, and how they coded. but i think (if possible) gnome team should use an abstraction blabla, “if” the user loading another wm, to save panel.. whcih i like too.. 🙂

  7. Stop telling to people to JUST choose KDE and live with it and start thinking about some REASONABLE suggestion. Linux is all about freedom of choice and if you don’t like that, go to windows or mac – their devs will make all the decisions for you ;]

  8. (Flame wars = bad)

    I should probably clarify something. Unless we rearchitecture the majority of compiz, integrating with shell won’t be possible at the moment because of the way clutter works.

    Anyways, the point of this post is that a number of people were offended by my previous comments and I wanted to rephrase them in a nice way. Obviously I haven’t done that either.

    I’m starting to reconsider my role as a councellor.

  9. i think compiz as a windows manager was a hell of a tech example for all of us about what can be accomplished using composite rendering techniques but i agree is outdated today, why? is hard as mentioned above to have a fully integrated enviroment to be trully useful but is true too that other composite groups like kde4 kwin arent even close to the power of compiz in performance or effects, gnome in another scenario is hell hard to integrate whatever you want. so the problem here is not kde4 nor compiz (gnome well they gain some momentum cuz the trully fully geeky gpl stuff but qt 4.5 being gpled and multi plataform + comercial support in case someone need it i think kill that momentum + they are pretty behind KDE in all terms). So how we can fix the kwin issues and give compiz they rigthful place? easy convert compiz from a windows manager to a modular C++ api independant library, in that way anyone can integrate compiz in their enviroment and you can provide an stable API so anyone can write plugins that will work in any enviroment wich use this library so no more code duplication (i think api independant so other projects can create binding to this library so it gets easier for specific API developer to develop even faster stuff) + you can get a lot more debugging information and coder help that way, so we all are happy again + a freaking amazing desktop experience.

  10. I don’t think the GNOME Shell devs have anything against compiz in particular, but I do see a resistance to putting in the major effort to make it window manager agnostic. As a contributor to the GNOME Shell project, I can say that I have always been a big supporter of Compiz(use it on my netbook to boost productivity and provide eyecandy for onlookers) and would like, at the very least for GNOME Shell to incorporate the best aspects of compiz even if it doesn’t end up being compatible.

  11. hello. that is a cool idea… compiz devs working to make kwin better.. think about it… “Kompiz” but hey then i guess we would have one less choice which is where the road to Lindows begins…

  12. Hey, hey, hey, HEY!

    Cut that crap of “stop compiz and start contributing to Gnome/KDE” you guys. There are out there more desktop environment options than just that Gnome/KDE shit, in fact, I don’t use any, I just fire up compiz directly.

    Seriously, who needs a desktop environment when you can just use compiz and a good set of applications?

  13. SMSPillaz, what about the idea that was mentioned earlier about transforming Compiz into a library? This would certainly require a WM to support a plugin arch (or perhaps something specific for compiz, though that seems at cross purpose) but it would provide a nice set of (potentially) OGL accelerated functions for a WM. Assuming the other the WM devels agree it would be a much more efficient use of resources for all involved, and Gnome could control how much/little of Compiz they choose to expose. However, I wonder if it wouldn’t be easier to rewrite compiz using Clutter. That should make development faster.
    Thanks for all your work. Compiz is a huge time saver and it also just makes the desktop experience so much more pleasant, BUT I’m also very interested in SpringDesk. I’ve struggled with coming up with a way to deal with Desktop(the directory) clutter(not the library) and SD may provide an interesting solution(as I see it, the problem is one of sorting, and one could start with magic numbers then continue from there). Hope you move forward with that.


  14. Сорри за оффтоп, кто-нить смотрел ролики на ютьюбе про конец света? Ну, про андронный колайдер Ваще страшно!

  15. Занятно пишете, жизненно. Все-таки, для того, чтобы делать по-настоящему интересный блог, нужно не только сообщать о чем-то, но и делать это в интересной форме:)

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s