The end for Mandriva [LWN.net]

The end for Mandriva

Posted May 26, 2015 13:31 UTC (Tue) by amimjf (guest, #506) [Link] (1 responses)

Its a shame, for a long time the Mandriva installer was the best in the business, i must have installed Mandriva 10/11 on loads of computers and laptops 3/4 years ago. I have a powerpack edition with real printed manuals somewhere.

The end for Mandriva

Posted May 26, 2015 13:51 UTC (Tue) by moltonel (guest, #45207) [Link]

Mandriva is the first and only distribution I payed good money for, back when it was still Mandrake. Good old days. I've moved on long ago.

Conectiva return?!

Posted May 26, 2015 13:49 UTC (Tue) by higuita (guest, #32245) [Link] (1 responses)

When Mandrake acquired Conectiva (and merged it) i saw the end of a great distro. Maybe now it can reborn.

Conectiva was big in South America and while less known in the USA and Europe, it was as good or even better than redhat,mandrake or suse.

Conectiva return?!

Posted May 26, 2015 15:39 UTC (Tue) by paulj (subscriber, #341) [Link]

Conectiva had apt-rpm, which made life a lot easier for RPM distro users (people had apt-rpm repos for RedHat / Fedora). This spurred RedHat into adopting yum and working on it, to provide a similar ability to be able to easily search for and install a package + all its dependencies (though, incredibly slower in its early days).

The end for Mandriva

Posted May 26, 2015 15:07 UTC (Tue) by tonyblackwell (guest, #43641) [Link]

Fortunately its successor Mageia is only days away (all things coming together) from its 5th release, so at least the heritage is well passed-on.

The end for Mandriva

Posted May 26, 2015 17:22 UTC (Tue) by josh (subscriber, #17465) [Link] (9 responses)

It seems like many of the most fundamental issues with desktop Linux have been addressed. Desktop environments and other software are mostly nice and polished before distros even have to touch them. Much of the integration and cohesiveness that was once the domain of a Linux distribution now occurs upstream.

There's still room for non-traditional Linux distributions to innovate and experiment, and there's still a baseline level of polish expected of a distribution, but I wonder to what degree traditional Linux distributions are suffering because the baseline has surpassed what they grew up having to provide? How much room is there for more traditional Linux distributions?

The end for Mandriva

Posted May 26, 2015 18:08 UTC (Tue) by rahulsundaram (subscriber, #21946) [Link] (6 responses)

Can you clarify what you mean by traditional vs non-traditional? Distribution vendors like Red Hat and SUSE continue to do a lot of integration work upstream rather than just at the distribution level (installers, configuration tools, package management etc) so I am not sure how you are looking at it.

The end for Mandriva

Posted May 27, 2015 2:44 UTC (Wed) by josh (subscriber, #17465) [Link] (5 responses)

By "non-traditional", I meant distributions that are pushing some novel approach to packaging Linux, such as Nix, the former rPath, or similar.

The end for Mandriva

Posted May 27, 2015 6:28 UTC (Wed) by rahulsundaram (subscriber, #21946) [Link] (4 responses)

Would you count Fedora Atomic, GNOME Sandbox, Ubuntu Snappy etc in that list? If so, it looks like traditional vendors are willing to try out novel approaches as well.

The end for Mandriva

Posted May 27, 2015 6:54 UTC (Wed) by josh (subscriber, #17465) [Link] (3 responses)

GNOME Sandbox is more what I meant about upstream taking care of what distributions used to do.

Fedora Atomic is an interesting experiment, as is CoreOS; that's the kind of thing distributions need to try to remain relevant and compete.

My opinions on Ubuntu's "experiments" these days are not appropriate for civil discourse.

The end for Mandriva

Posted May 27, 2015 18:12 UTC (Wed) by rahulsundaram (subscriber, #21946) [Link] (2 responses)

> GNOME Sandbox is more what I meant about upstream taking care of what distributions used to do.

Ah ok. Now that I understand your perspective better, I can highlight the direction this particular effort is moving towards to the extend I understand it.

* GNOME Software has support for abstracting out the differences between distro native packages and sandboxes apps
* Fedora workstation is looking into enabling more of this support out of the box. See
https://lists.fedoraproject.org/pipermail/desktop/2015-Ma...
* The exact line on what would be supplied by the distro vs shipped as sandboxes apps is a ongoing discussion. System runtime vs individual apps seems to be a natural fit. Support from third party vendors might move that needle further.

The end for Mandriva

Posted May 27, 2015 18:36 UTC (Wed) by josh (subscriber, #17465) [Link] (1 responses)

I appreciate the extra info about GNOME Sandbox.

However, that was just one example of upstreams working on what would previously have been the work of a distribution. More generally, much of what a distribution would traditionally have done as "integration", such as making programs work together or use consistent appearances, has now been done in many upstream projects. Not perfectly, but to a degree that drastically reduces just how much distros need to do.

Similarly, Linux is much better about working with hardware out of the box, and many other pieces of software Just Work without needing changes.

There's still work for distributions to do, but it's much less than it used to be, and doing that work alone doesn't create something uniquely valuable.

The end for Mandriva

Posted May 27, 2015 19:02 UTC (Wed) by rahulsundaram (subscriber, #21946) [Link]

"There's still work for distributions to do, but it's much less than it used to be, and doing that work alone doesn't create something uniquely valuable."

Possibly. Depending on how successful the sandbox stuff is adopted across distros, users might not care about where exactly there are getting the base runtime from. That is already true for containers and VM's to a good extend.

The role of GNU/Linux distros

Posted May 26, 2015 18:36 UTC (Tue) by coriordan (guest, #7544) [Link] (1 responses)

Well, what do distros do? Besides the technical work of integration, they verify freeness and act as a filter to anti-features.

Cooperation between upstream projects is great, but sometimes upstream is a company with internally conflicting interests. Some companies wish they could gather data about users, or mix some proprietary software in with their free software. Distros are the filter, and because the companies know that bad stuff would be filtered out, they generally don't put bad stuff in in the first place.

I'm suspicious of projects that try to bypass the distros. I do install a handful of plugins from outside my distro (Debian), but I get everything else from my distro.

The role of GNU/Linux distros

Posted May 27, 2015 2:46 UTC (Wed) by josh (subscriber, #17465) [Link]

Distributions absolutely do serve a critical function. But for the most part, all distributions do reasonably well at that. (There are some variations; for instance, I trust Debian's vetting of software freedom more than others, but for basic integration most distributions get things right.) What room is left for another distribution to differentiate themselves from those that already exist?

The end for Mandriva

Posted May 26, 2015 17:27 UTC (Tue) by samuelgraeff (guest, #102782) [Link]

I'm very sad with this bad news.

Conectiva did many contributions to the free software community, one of them that I remember was synaptic (a graphical package management), that is in use, even nowadays by many distros.

It's a huge lost.

The end for Mandriva

Posted May 26, 2015 20:24 UTC (Tue) by xorbe (guest, #3165) [Link]

Luckily openSUSE is a solid KDE distribution, and filled the gap nicely for me. Mandrake had horrible policies resulting in nearly broken ISOs every single release.

Mandriva & Debian

Posted May 27, 2015 4:09 UTC (Wed) by pabs (subscriber, #43278) [Link]

Ah, fond memories of migrating a LTSP system from Mandrake to Debian sarge via debtakeover are coming back to me.

The end for Mandriva

Posted May 27, 2015 15:10 UTC (Wed) by kloczek (guest, #6391) [Link] (14 responses)

As long as Linux lost his combat about desktops by broken by design ALSA end of Mandriva was obvious consequence.

The end for Mandriva

Posted May 27, 2015 20:46 UTC (Wed) by branden (guest, #7029) [Link] (13 responses)

return -EPARSE;

The end for Mandriva

Posted May 27, 2015 22:38 UTC (Wed) by cesarb (subscriber, #6266) [Link] (12 responses)

I think he's blaming ALSA for the end of Mandriva.

"(As long as ((Linux lost his combat about desktops) by (broken by design ALSA))) ((end of Mandriva) was obvious consequence)."

That is, ALSA -> no Year Of The Linux Desktop -> end of Mandriva.

The end for Mandriva

Posted May 27, 2015 22:48 UTC (Wed) by nix (subscriber, #2304) [Link] (2 responses)

Yeah. Everyone knows that only Mandriva used ALSA: everyone else is using the widely-supported thoroughly free and well-designed OSS4. See? Obvious!

(when all you have is a hammer, everything looks like a conspiracy, and kloczek clearly *really likes* the long-obsoleted OSS4...)

The end for Mandriva

Posted May 27, 2015 23:05 UTC (Wed) by kloczek (guest, #6391) [Link] (1 responses)

Of course only Mandriva failed on desktops. Isn't it?

The end for Mandriva

Posted May 27, 2015 23:29 UTC (Wed) by rahulsundaram (subscriber, #21946) [Link]

Plenty of desktop products have failed including "Java Desktop System".

The end for Mandriva

Posted May 27, 2015 23:05 UTC (Wed) by sfeam (subscriber, #2841) [Link] (8 responses)

Which would be weird, because ALSA worked fine on Mandriva. It was pulseaudio that broke Mandriva's sound support when it arrived :-)

The end for Mandriva

Posted May 27, 2015 23:08 UTC (Wed) by kloczek (guest, #6391) [Link] (7 responses)

Lack of in-kernel software mixer few years caused complete fail Linux on desktops.

The end for Mandriva

Posted May 27, 2015 23:23 UTC (Wed) by sfeam (subscriber, #2841) [Link]

Mandriva had a user-space sound demon that mediated multiple-source sound input. I do not know if other distro desktops had an equivalent, but it worked fine for me. Better than the early version of pulseaudio that eventually replaced it.

Even now there is no in-kernel audio mixer that I know of. Why would you need one?

The end for Mandriva

Posted May 27, 2015 23:25 UTC (Wed) by flussence (guest, #85566) [Link]

dmix worked fine for me in 2004. PEBKAC.

The end for Mandriva

Posted May 28, 2015 5:54 UTC (Thu) by bandrami (guest, #94229) [Link] (4 responses)

Why on earth would I (or anyone) want to do audio mixing in Ring 0?

The end for Mandriva

Posted May 28, 2015 10:08 UTC (Thu) by SimonKagstrom (guest, #49801) [Link]

Because it gives meaning to lyrics such as

"I fell into a burning ring of fire,
I went down, down, down as the flames went higher
And it burns, burns, burns,
The ring of fire, the ring of fire."

;-)

The end for Mandriva

Posted May 28, 2015 13:51 UTC (Thu) by ms_43 (subscriber, #99293) [Link] (2 responses)

Ask the FreeBSD developers why they are doing exactly that.

Just think about the possibilities, GNU/Linux could be as successful as FreeBSD on the desktop if only audio hadn't been made gratuitously complex with user-space mixing.

The end for Mandriva

Posted May 28, 2015 17:39 UTC (Thu) by Cyberax (✭ supporter ✭, #52523) [Link] (1 responses)

> Just think about the possibilities, GNU/Linux could be as successful as FreeBSD on the desktop if only audio hadn't been made gratuitously complex with user-space mixing.
/me looks at Internet traffic statistics.

Linux Desktop - 1%
FreeBSD+NetBSD+OpenBSD - 0.05%

You know, I think you're totally correct!

The end for Mandriva

Posted May 29, 2015 8:31 UTC (Fri) by nhippi (subscriber, #34640) [Link]

I think he's totally sarcastic

The end for Mandriva

Posted Jun 10, 2015 15:08 UTC (Wed) by tbm (subscriber, #7049) [Link]

This article has more details about why the company shut down:

http://www.businessinsider.com/ceo-employee-lawsuits-kill...