Canonical To Launch Some Sort Of Ubuntu Developer Site
Phoronix: Canonical To Launch Some Sort Of Ubuntu Developer Site
The development discussion surrounding whether Ubuntu 11.04 should default to the classic GNOME desktop rather than the Unity Desktop being developed in-house at Canonical is ongoing. One of the latest discussions is over the lack of system tray support in Ubuntu 11.04 Unity and the new style of application notifications. An interesting comment was just posted by one of Canonical's employees...
I hate when they make it either/or. No, you won't have standards anymore, if you code to those, your $FEATURE will not work on Ubuntu $FUTUREVERSION. And vice versa.
Death to the notifier!
The notification applet is a pain in the ass. I just hate it so bad.
I have to install on application basis the .deb's from Debian repositories just to get rid of them so i would not be affected. That seems to be what a lot of people are doing and what is recommended by some users in the Ubuntu channels.... sad state of an distribution.
I also would like the splash to be gone forever and only installable as an option for those who want it. I prefer to see the text if there is any problems, "no-splash/nosplash and verbose" does only remove the splash for the first seconds not the rest of the boot process, ridiculous.
Just had to vent.
Yay, finally someone doing something to ease the gripes of corporate developers. I hope they'll continue working with LSB, as unawesome their tools might be, on this.
Aren't you getting a bit too worked up over these tiny things? Are they really that big of a pain?
Originally Posted by judgen
Besides, if you don't like them, why don't you start distro hopping? It's not like there aren't a zillion ubuntu/Debian derived distro's out there. Surely, one of them has to have the things you like?
Ubuntu is aimed to be user friendly. If you want some h4x0r 3lit3 feeling install something else.
Originally Posted by judgen
Erm, the site already exists. It's just waiting to be properly filled with content when the time comes:
Is there a chance a site like this (a real one, not developper.ubuntu.com) will be available for non ubuntu developpers (reading only I mean) ?
Exactly this. Unless the Ubuntu market share relative to the rest of the GNU/Linux distros becomes as dramatic as Microsoft's market share relative to other OSes, you can't properly capture support for "Linux" without considering, at the very least: Ubuntu, Debian, Fedora, RHEL, and OpenSUSE. Each of them has a significant enough part of the market share that leaving them in the dark is a costly mistake.
Originally Posted by phoronix
And since companies who (at this early stage) author software that runs on Linux already have enough foresight to break away from Microsoft, chances are that they possess the foresight to realize that leaving people in the dark is very, very bad for PR, and excludes whole segments of the market from using or purchasing your product.
So if in the future, Ubuntu has 95% of the Linux market share, then okay -- I guess the Ubuntu APIs would become de facto by their essentially universal market share. As long as they are free software APIs with a free software implementation, I'll go along with it (grudgingly).
But if the Fedora/RHEL/OpenSUSE type distros out there continue to command measurable market share (especially in enterprise workstations and servers) in the 25 - 50% range, you can't ignore them. So you have to write to the lowest common denominator.
That probably means that the best API targets on Linux for the next couple years will be the classic GTK2/Gnome2 platform (Gnome3 is too new to be "universally available" enough), Qt4-Core/Gui/Net, Gstreamer 0.10 for media, OpenAL-soft or PulseAudio 0.9.2x for low-level audio, OpenGL 2.1 for 3d rendering. And some companies will still link directly against libX11 if they want to do very custom stuff without worrying about whether the platform supports their library stack.
Open source works generally won't have a problem targeting newer APIs such as Gnome3, or less universal APIs such as the "Ubuntu APIs", because anyone finding the program valuable can add support for their favorite desktop environment. But trying to influence the commercial-proprietary folks to target Ubuntu APIs to the exclusion of all others? That's sinister.
Then again, so is proprietary software in general, so I'm not sure that it's too much of a loss. But still, justifications exist for making and using proprietary software in some situations; it'd be nice if the software would work on all distros instead of just Ubuntu (otherwise it becomes a classic case of vendor lock-in).
What always amuses me is that when Ubuntu tries to propose their in-house APIs as "standard" for other distros to pick up, by and large they are not adopted at all, and forcefully rejected on technical grounds. The one counterexample I know of is upstart, which made its way into many distros and was widely recognized as an improvement to the boot system. But I think systemd is another leap ahead of upstart and will (should) eventually replace upstart, even on Ubuntu.
But if anything, upstart is empirical proof that most third parties judge in-house Ubuntu/Canonical works by their technical merits only, not by the political aspects of it. If it were political, then even the technically superior upstart (for its time) would have been rejected. But enough distros adopted it, while not adopting other Ubuntu/Canonical technologies, to suggest that upstart was an exception to the usual: the usual being that Ubuntu just doesn't create good, maintainable, generally useful APIs that are superior to existing solutions. If they did, then you'd see a wider adoption of things like libappindicator, Unity, Ubuntu One, their restricted drivers applet, and so forth.