Catch up on stories from the past week (and beyond) at the Slashdot story archive

 


Forgot your password?
Close
typodupeerror
×
Programming IT Technology

Interview With The KDE And GNOME Release Managers 162

Posted by timothy from the dedication-defined dept.
An anonymous reader writes "It has to be tough, keeping projects as big as GNOME and KDE organized, but that is the job given to those projects' 'release managers.' In an interview on Linux and Main, KDE's Dirk Mueller and GNOME's Jeff Waugh discuss their wacky, devil-may-care, hell-bent-for-leather, zany, fun-filled world -- the shadow, as T.S. Eliot put it, between the idea of a release and its reality."
This discussion has been archived. No new comments can be posted.

Interview With The KDE And GNOME Release Managers More

Interview With The KDE And GNOME Release Managers

Comments Filter:

  • I have a question (Score:1, Interesting)

    by Anonymous Coward
    Gnome and KDE seem to be geared towards windows migrating ppl. They all emulate the same basic look and feel. My question is, is there any project of the same calibre (of would be soon), that does a native look and feel (modern and cool, like in movies) for Linux/BSD's ?

    Thank you.

    Amma Fui

    • Re:I have a question (Score:2, Interesting)

      by dotgod ( 567913 )
      3dwm [3dwm.org] is an interesting 3D window manager project (not an X11 window manager though)...definitely a bit different from the "same basic look and feel"
      • Interesting. But for the same exact reasons 1D commandline hasn't died out to the 2D desktop and can be more productive, the 2D desktop could be more productive than the 3D window manager.

        On the other hand (no puns intended), the command line needs the keyboard. The desktop needs the mouse, even though the keyboard alone can be used. Which brings us to the 3D desktop. Without a faster interface (e.g. data glove), the mouse might not be the ideal interface for it.

        Food for thought: how many people would prefer the trackpad over the rubber joystick on the notebook for extended desktop work?

    • Re:I have a question (Score:4, Informative)

      by paladin_tom ( 533027 ) on Wednesday September 04, 2002 @12:05AM (#4193385) Homepage

      My question is, is there any project of the same calibre (of would be soon), that does a native look and feel (modern and cool, like in movies) for Linux/BSD's ?

      Amma, the destops you see in movies look cool on the screen, but that's their only good point. Their user interfact is horrible. Movie producers purposefully give computers on movies a "computerish" look; for example, movie computers often have green text (a la the text in The Matrix, when Neo sees it), which has been obsolete for years -- it's hard on the eyes.

      Movie destops also have an excessive amount of animation. Trust me, you'd hate any desktop that worked that way -- it'd run like absolute molasses.

      The user experience I've had that most closely resembles a movie desktop is application built using Flash, like you find on web sites and enhanced CDs. These apps try to emulate the look and feel of movie desktops.

      • Movie destops also have an excessive amount of animation. Trust me, you'd hate any desktop that worked that way -- it'd run like absolute molasses.

        Kinda like Mac OS X, eh? Yeah, yeah, it's getting better, but it's still pretty slow compared to other operating systems with less flash.


        The user experience I've had that most closely resembles a movie desktop is application built using Flash, like you find on web sites and enhanced CDs. These apps try to emulate the look and feel of movie desktops.

        Interestingly, most of those movie GUIs are built using Macromedia Director (or similar), so building the same thing out of Flash makes sense. I wouldn't be surprised if some of those movie web sites actually took most of the code directly from the mock-up made for the movie itself. It shouldn't be difficult to do.


        Personally, I like usability with a bit of flash. Something like Window Maker [windowmaker.org], while very useable (it should be, since it's based on NeXT), is rather boring. Enlightenment [enlightenment.org], while flashy, isn't that useable in my opinion. In that vein, I'm pretty happy with Windows XP [microsoft.com]. It's themeable [themexp.org] (link is down at the moment, but according to the notice it should be back up in a half an hour -- I doubt that, but check back in a day or two), so I can get my eye candy, but it's also very useable. Say what you will, Microsoft has spent a small fortune on useability testing, and most of what they've done works well. Brush it off as familiarity if you will, but there are concrete examples of Windows useability getting better (small example: the Start button now has infinite borders, just like the Apple menu in Mac OS -- throw the mouse down to the lower left and click, you'll get the start menu).

        • i read this online, that XP is whoohoo great. everyone i've talked to who's actually used it either says "whah? its different from anythinge else?" or "yeah...XP kinda sucks." Overall, i've gotten >10 nays and one yay. It's fine that you like it...but my personal data-polling (my most trusted source) tells me that XP isn't that hot. Usability testing's great. just give people something that makes reasonable sense, and let them learn it.
          • Overall, everyone I know (myself included) that's given XP a chance has liked it. The things that make it great are the details, and therefore tend to be small items that people don't think about until they go to another OS (even previous versions of Windows) and realise they're not there. Basically I look at it as Win2k with more compatibility and a lot of little feature improvements (like ID3v2 tags and other metadata as sortable columns in detail view, images placed in a folder used on the folder image in thumbnail view (irreplaceable for browsing ripped MP3s), the recently-used-apps column on the new start menu (plus pinned items), etc.).
        • Not just the start button, but the entire task bar has infinate bounds. It's about time though.

          • Not quite the entire taskbar. The systray area still requires precision mousing, but then you're generally not going down there very often, and when you are you're probably trying get a tooltip (say, of the time and date), and so will be moving the mouse around carefully. The Quick Launch bar (and therefore, probably other toolbars) also seems to not have infinite borders. And of course you can change the size of the taskbar so that you end up with multiple rows of buttons, at which point only the "bottom" (assuming taskbar at the bottom, of course) row has an infinite border. But hey, at least some of it works! Now if only I used the mouse that much, rather than hitting my Window key, alt-tab, Window-R, Window-M, etc . . .

      • "This is a Unix system... I know this!"

        "I think I can get Jurassic Park back online."

      • Um no.. (Score:3, Funny)

        by majestynine ( 605494 )
        ..movie operating systems have bad useability?!..

        Ive never had problems with my movie operating system. Say... you want to hack someone? Simple!

        Step 1: Open 'Elite Hacking Utility' in the 'Extras' submenu of the startbar.

        Step 2: Enter the IP Address of your target, or click 'Hack Previous Download Stream' to hack the last server you sent data to. Then click OK

        Step 3: 'Elite Hacking Utility' will automatically determine which OS the hackee is running, and then will prompt you if you wish to proceed. Click OK to continue, or Quit to exit the program.

        Step 4: The hackee computer is running "Debian Linux." If you press okay, Elite Hacking Utility will hack through any firewalls and give you full access to that computer, using simple translators to convert anything to the file types supported by the OS. You will then gain complete access to their computer and all their files. You have not owned them! Although Elite Hacking Utility is a very safe program, there is a rare chance that the hackee could make a counter-attack, that Elite Hacking Utility may be unable to stop. To proceed, click OK, to choose a different IP address, click Back, to exit the program, click Quit, if you need assistance, press the Help key on your keyboard or choose Elite Hacking Help from the Help menu.

        Simple!

        • Re:Um no.. (Score:3, Funny)

          by Roblimo ( 357 )
          5. Run "Guess Password" utility. It tries the sytem user's address, phone number, variations on wife or girlfriend's name, and several significant dates in the user's life, plus the ever-popular "rosebud." The current version of this utility only works correctly when the bomb countdown is almost at zero.

          I believe an improved version of this utility, one that isn't bomb countdown-reliant, is supposed to be included in KDE 4, but of course that's up to the release manager.

          - Robin
          • and don't forget that it even works on alien navigation and communication systems, even if your host computer has problems communicating with earth-born operating systems.
      • The flash app of which you speak --- may we view it? Grazi.
    • Have you tried windowmaker (part of the gnustep desktop)? Its not as full featured as KDE or Gnome but it runs their apps fine, is much more light weight and doesn't feel like windows at all.

    • Re:I have a question (Score:5, Informative)

      by Arandir ( 19206 ) on Wednesday September 04, 2002 @01:04AM (#4193548) Homepage Journal
      Let me take your unwritten assumptions in order.

      Assumption: The goal of GNOME and KDE is to convert Windoze users.

      Truth: There are as many goals for these projects as there are developers. The goal of converting Windows users is definitely on the list somewhere for some developer, but overall it is very low on the totem pole. Much higher are the goals of "scratching my itch", "improving the desktop I use", and "writing my application in this awesome development framework I've found."

      Assumption: Windoze users will be more comfortable in a cloned Windoze environment.

      Truth: There are two kinds of Windows users. Those that can't stand change and those that want to get away from that crappy desktop. Nothing is going to please the former but the genuine article available only from Microsoft. The latter don't want that article, which is why they're leaving.

      Assumption: GNOME and KDE emulate the Windoze look and feel.

      There is some resemblance between GNOME/KDE and Windows. But it's superficial only. Take a second glance and there's no comparison. The Windows desktop is ugly and very low in usability. On the other hand, both GNOME and KDE have great usability and look good. There is no comparison between Kicker/Panel and the Windows taskbar. There is no comparison between KWin/Sawfish and the Windows "window manager". There is no comparison between Konqueror/Nautilus and the Windows Explorer. Etc, etc, etc.

      Assumption: There is a native look and feel for Linux/BSD somewhere if I can find it.

      Truth: There is not a native look to Linux/BSD. Trust me, I've checked. That's because there isn't *a* native look, but multiple native looks. Among them are GNOME and KDE.

      There is no centralized authority over the *Nix look and feel, so there will never be just one, no matter how hard Redhat tries.

      Assumption: The desktops in movies are modern and cool.

      Truth: They sure look modern and cool, but they also look like usability nightmares.
      • Assumption: There are two kinds of Windows users. Those that can't stand change and those that want to get away from that crappy desktop.

        Please don't assume that just because you think your enviroment is surperior, that everyone is either on your side, or a stubborn fool.

        Believe it or not, some of us who have used KDE & GNOME & Windows, would still choose the Windows enviroment, not because we resist change, but because it's what works for us.
        • Okay, question then. Without referring to the availablity of applications (which is a separate issue) or the difficulty in installing and configuring the underlying operating system (another separate issue), why do you prefer the Windows *desktop* over the alternatives?
    • You want to try Enlightenment 0.17.

    • KDE with different settings.
    • They all emulate the same basic look and feel. My question is, is there any project of the same calibre (of would be soon), that does a native look and feel (modern and cool, like in movies) for Linux/BSD's ?
      Sure, I'll get right on it. Could you just take a minute to define what the new look and feel should be?

      The fundamental problem in UI design is that since the "ah ha!" insight that created the WIMP interface and the desktop metaphor, there hasn't been another key insight or breakthrough. Having once seen an actual Star workstation in action I can attest that a Star user wouldn't have any difficulty using the current Mac or Windows interface - they are fundamentally the same. Everyone agrees that we need something better than the W95 desktop - its just that no one can figure out what that is.

      sPh

    • > (modern and cool, like in movies)

      You apparently have seen different movies than I have. The movie
      desktops I've seen don't provide any facility for launching apps,
      switching between apps, or anything else useful. They usually
      consist of one big screen-filling dialog box containing options
      for performing impossible feats, such as trying all the
      possibilities for a sixteen-character password in thirty seconds,
      or uploading a fatal virus to an alien computer system about
      which nothing is known. Also, they have about 160x50 resolution,
      which is anything but modern and cool.
    • Look at Berlin Project [berlin-consortium.org].

      And if you like it then read here:

      Berlin is a windowing system derived from Fresco, a powerful structured graphics toolkit originally based on InterViews. Berlin extends Fresco to the status of a full windowing system, in command of the video hardware (via GGI, SDL, DirectFB or GLUT) and processing user input directly rather than peering with a host windowing system. Additionally, Berlin's extensions include a rich drawing interface with multiple backends, an upgrade to modern CORBA standards, a new Unicode-capable text system, dynamic module loading, and many communication abstractions for connecting other processes to the server. It is developed entirely by volunteers on the internet, using free software, and released under the GNU Library General Public License.

  • I applaude the work of leaders, teams ,and freelancers, who have given their free time(and a lot more) to a project that will mean so much to poor communites around the world. These poor communities now have access to technology that can uplift and enrich their lives without breaking their wallet. perhaps .. technologies such as these will do much to ammend the huge discrepancy between rich and poor. Thank you, in the purest sense possible.

  • Free and commercial release processes (Score:3, Interesting)

    by ColdChrist ( 543345 ) on Tuesday September 03, 2002 @11:38PM (#4193315)
    One of the things that's interesting about this article is how closely related the release process for free software is to that for commercial software. Commercial products have many things in their genesis that free products do not, but when it gets to be time to produce a clean release "the aim of the release process is to finish software, not to develop it." Waugh cites that quote from Havoc Pennington as applying to Free Software, but in fact it applies to both. It's a kind of convergent evolution into a niche; the ecological imperatives of product survival force projects to adopt these mechanisms in order to be successful.

    The really interesting question is where does this convergence start? Are the reward systems, involving kudos and problem-solving pleasure for free software, and money for commercial software, fundamentally different? I suspect they're not, and that there is much less difference between an open source project and commercial product development than is sometimes thought. I'd guess that the more successful examples of each strongly resemble each other.

    • Re:Free and commercial release processes (Score:1, Interesting)

      by Anonymous Coward
      Think for a second about how works on free vs. commercial projects.

      For the most part, the people working on free projects are a self-selected group of motivated talented people. For commercial projects - well - you probably remember doing 'group projects' in school.

      The rewards for free software are small, and not at all required for existance (unlike say, a pay check). As a result, anyone who is not interested or talented enough to contribute leaves. For commercial projects, this just isn't the case.

  • There is help! (Score:2, Funny)

    by stox ( 131684 )
    Dear Mr. Dirk Mueller and Mr. Jeff Waugh:

    The ACME Sales Corporation is proud to announce a solution to your problems. The XJ19 Laser tatoo removal kit. It will even remove the "Sucker" tatoos clearly emblazended upon your foreheads. How did you get conned into taking these jobs? Cat hearding is easy in comparison to the tasks you each wield. For only $99.95, and three sacrificial rubber chickens, we will gladly send you our kit.

    But seriously, thanks for all the fish!

  • I can associate with this ! (Score:4, Informative)

    by StArSkY ( 128453 ) on Tuesday September 03, 2002 @11:54PM (#4193359) Homepage
    As a project Manager I can really associate with these guys..

    I became a release manager at the company where I used to work by volunteering... it nearly gave me a heart attack after 18 months.

    Release dates are set at standard intervals becasue theat's the way it has always been done ;)

    Scope changes to meet time available ;) If it looks to tough to finish in time, delay it for the next release.. I have seen releases with 50 updates and fixes scaled back to 10-15.

    Time to code ! bah no release manager has time for sleep, family or counter-strike..

    and all of the above explains why I now do Business Continuity rather than release management ;) Mind you I would recommend it to anyone who wants to see how project teams, communities and stakeholders react to what happens with software releases. As the release manager you are one of the few who ever gets to see all sides of the argument.
  • Being responsible as a release agent for anything
    like KDE or Gnome has got to eat a BIG hole in your daily time line, I have a few moments in my day to help friends with small problems out of my normal daily work cycle, I can't imagine the time it takes to get involved in a project of this magnitude. My Hat is off to you. I have a small job as a SMT operator/prgmr,eats up 8-5 real fast, my brain even faster!

  • Mozilla did it better (Score:3, Interesting)

    by sohp ( 22984 ) <.moc.oi. .ta. .notwens.> on Wednesday September 04, 2002 @02:03AM (#4193690) Homepage
    KDE and Gnome releases are fine, but compared to the Mozilla build/release process, managed by the enigmatic Leaf, they are 2nd class. Mozilla developers created their own tools to do it, too. Mozilla is cross platform, continuous builds [mozilla.org], bug tracking [mozilla.org] integrated with version control [mozilla.org], and they released regularly on a five week cycle [mozilla.org] (now quarterly), and daily build and smoketests [mozilla.org]. And once again, Mozilla is cross-platform -- Linux, Windows, and Mac OS 9/X.

    Sorry to crash the party, but I have yet to see KDE or Gnome approach the bar that Leaf and Brendan Eich set high.
    • "bug tracking"

      What can I say: the world's buggiest code required the world's best bug tracking software.

      (this is not an insult - Mozilla has become a respectable, stable browser)

      Who here remembers M7? It's so cool to watch a browser progress from it's early primitive roots to what it is today. Amazing.

      Bravo, Mozilla developers - for developing the world's best browser.

    • Mozilla is a single application suite; it is small compared to either KDE or Gnome.

      • Mozilla is a single application suite; it is small compared to either KDE or Gnome.

        True, but this is only more reason to use advanced release management and QA tools. I'm not sure if KDE or Gnome has continuous builds or something, but I doubt it (at least I've never seen them) and in this respect it's true that they can learn a lot from Mozilla (or even use some of their code ;-).

        • Developers like me do continious builds all the time, many are running the latest CVS code as primary desktop or alongside a stable release.

          With a developer base in the hundreds of active people, there is little need for automated builds. Mozilla's developers base is way smaller, so they need a more automated QA system.

          Furthermore, KDE does not provide binaries itself at this moment, although there are packagers who will probably do the necessary QA themselves.
          • Other old hands will tell you war stories, I'm sure, but a developer being able to build and run the latest code is not the same as continuous integration on a designated integration station with an automated build and smoketest.
            With a developer base in the hundreds of active people, there is little need for automated builds.

            That's about as scary an assertion as I've ever heard from a developer on a large software project. Quite the opposite, the bigger your team, the more you need that automated build and smoketest. But I thinkJoel [joelonsoftware.com] can say it better than I.
            • Quite the opposite, the bigger your team, the more you need that automated build and smoketest.But I thinkJoel [joelonsoftware.com] can say it better than I.

              Joel's main argument for an automated system is:

              "A tester finds a bug in the code, and reports the bug. The programmer fixes the bug. How long does it take before the tester gets the fixed version of the code? In some development organizations, this Report-Fix-Retest loop can take a couple of weeks, which means the whole organization is running unproductively. To keep the whole development process running smoothly, you need to focus on getting the Report-Fix-Retest loop tightened."

              I still believe (based on experience) that this is not a problem for KDE. There is a very large group of people who build KDE from CVS or CVSup sources and have instant availability from the patch, which was quite possibly also sent to the respective bug report system (and mailed to the originator) and project mailinglist.

              My experience tells me KDE is not one of the development organizations where this is a problem. Patches spread quickly within the KDE community itself, every major release (6-9 months) is followed by several minor releases with bugfixes only (every approx. 6 weeks, unless an urgent fix requires a faster release).

              Maybe Mozilla did not have such an active community, but fortunately KDE has and it seems to work.

              Of course there are some automation tools, such as build scripts in kdesdk, but those are still there to be used by end-users themselves. A community also scales the variety of compilers, platform and dependency versions over a much broader area than any automated system could. It would be a mere impossibility to maintain that variety.
      • Mozilla is a single application suite; it is small compared to either KDE or Gnome.

        This is a joke, right?

        I mean, there's even a company [oeone.com] that is using Mozilla to create a nice interface for Linux machines. How exactly is this different from KDE/Gnome?

        It's every bit a platform as KDE or Gnome: It provides a user interface via xml, scripting language support (javascript), the ability to write add-ons for it, and it includes an html rendering engine, a complete email program, an WYSIWYG html editor, an address book, and (soon) a calendar/scheduling program. And anything it gives up in size to KDE/Gnome, it makes back due to the cross-platform complexities.

    • Re:Mozilla did it better (Score:3, Informative)

      by luge ( 4808 )
      Um, GNOME does all of those things (except smoketesting, which we hope will be coming soon). We release on a regular 3 week schedule (with 6 month major releases), we use bugzilla religiously, and we use tinderbox and release daily snapshots on multiple platforms (via both ftp and Red Carpet.) So... frankly, you don't know what the hell you're talking about :)

      P.S. Leaf and Brendan would probably consider you a troll. I sure do.
  • They keep mentioning it in the article and I must admit I have no idea what that is. My best guess would be it's some crazy abbreviation for installation.

    ...which usually includes docs, i18n, etc. as well as maintainers...
    • Internationalization. It's abbreviated "i18n" because there's an "i", 18 more letters, then an "n".
      • Yes, and basically because Internationalization (did I get it right?) is an incredibly long word that is extremely easy to misspell. Especially for the people involved in the process, because they mostly do not have english as their main language, which in fact is the reason for i18n in the first place.

        Gaute
    • "il8n" is shorthand for the word internationalization. In this context, making KDE and Gnome work for the parts of the world that don't speak English.

      • Re:Can someone explain what "i18n" is? (Score:5, Funny)

        by wgmari ( 215656 ) on Wednesday September 04, 2002 @03:00AM (#4193796)
        In this context, making KDE and Gnome work for the parts of the world that don't speak English.
        You mean like America? ;P
        • Let me think... Quebec, Mexico, Guatemala, El Salvador, Honduras, Nicaragua, Costa Rica, half of Panama, Venezuela, Colombia, Bolivia, Ecuador, Peru, Brazil, Chile, Uruguay and Argentina

          v.s.

          Rest of Canada, USA, Belize and half of Panama...

          Yup, it seems that most of America doesnt speak English (at least as a native language)

        • Amusing, but in many ways modern American English (especially the English spoken by backward folks in rural areas, like Appalachia) is closer to the English Shakespeare spoke than modern British English is. For many words that Americans and Brits pronounce differently (e.g. "schedule"), the American version is the older one, the one formerly used in both countries.

          The story is similar in France vs Quebec: it is the French who have changed the language more.

    • Re:Can someone explain what "i18n" is? (Score:4, Informative)

      by JabberWokky ( 19442 ) <slashdot.com@timewarp.org> on Wednesday September 04, 2002 @02:58AM (#4193792) Homepage Journal
      It's all the translation strings for every language. If you write an app for yourself, you can just write it in the language you use. If you want it to be used around the world, you have to provide a seperate translation for every single string: the menus, buttons, documentation, mouseovers... everything. In an open source project you then have to run it past seperate language maintainers to proof read all the translations.

      KDE has very impressive i18n tools (Fire up kbabel and take a look if you're running KDE right now), and I'd assume that any major project (Gnome, Mozilla, etc) all have nice tools as well. Unfortunatly, due to an oversight in Unicode, KDE's support for tlhIngan Hol uses the english alphabet.

      --
      Evan

    • Internationalization.
      There are 18 letters between the I and the N.
      There's also l10n, which stands for localization.
  • that the story has MS visual basic ADvert right in the middle?

    Sorry for OT trolling but kinda thought it to be real real ironic considering that Kdevelop is direct competitor of MS programming environment.
    Moving back online to the topic, I felt the interview a bit more general with very general questions with even more general answers.
    I guess more hard hitting interview is the need of the hour with the interviewer baying for blood ;-) just like larry wall!
    Better still get both of them together and lets have a flame war about wether KDe or Gnome is better. Too radical... I guess not i would really like to know what the KDE developers and leaders really feel for Gnome and vice a versa.. some interesting interview will be that!
    • I would really like to know what the KDE developers and leaders really feel for Gnome and vice a versa.. some interesting interview will be that!

      Most of them don't have any opinions either way - the main KDE/Gnome developers normally run an exclusively KDE/Gnome desktop, and so don't think about the other desktop at all. They're a very good example of convergent evolution. There *are* some times when something passes from one desktop to another, but they're relatively infrequent (for example, the Gnome 2 icons are *very* influenced by the KDE 2 ones, and in the other direction the KDE file manager got an improved look-n-feel after the release of Nautilus).

    • Re:aint it ironic....... (Score:4, Interesting)

      by LMCBoy ( 185365 ) on Wednesday September 04, 2002 @10:08AM (#4194861) Homepage Journal
      Kdevelop is direct competitor of [the] MS programming environment.

      No, it isn't.

      Most importantly, KDevelop is not a commercial product, so it has nothing to compete for. Sure, it's nice if lots of people use it, but ultimately it matters not at all how many "customers" KDevelop has. As long as there are interested developers, the project will thrive.

      More obviously, KDevelop targets only unix apps; MSVB targets only windows apps. They're in completely different "markets".
  • I found this quote interesting:
    Any organization or person is welcome to contribute to KDE, and anybody is permitted to distribute modified versions of KDE under the permissions granted by the individual licenses of the KDE components.
    This should be food for thought for all the KDE zealots who have their panties up in a bunch over the latest Red Hat beta. ;)
  • I think what this article is really pointing out is that you _don't need to be a coder to contribute to open source_. I know that people who code on open source projects constantly say that, but other, non-involved people don't seem to understand that improving documentation, testing, PR, scheduling, and the other "support" tasks are damned important to a good open source project. A room full of coders will not get you a good open source project!

    -Erwos

Slashdot Top Deals

Our business in life is not to succeed but to continue to fail in high spirits. -- Robert Louis Stevenson

Close