Slashdot Mirror

← Back to Stories (view on slashdot.org)

Mandriva Juggles Multiple Codebases

Posted by samzenpus on from the give-me-more dept.

jfruh writes "In the wake of its decision to cede control of its Linux distro to its community, Mandriva is trying a tricky balancing act: offering Linux products based on two different code bases. Desktop and OEM offerings will be based on the Mandriva distro, while server products will be based on the traditional Mageia codebase." Update: As babai101 points out the codebases were reversed in the original post.

44 comments

  1. OP got it wrong! by babai101 · · Score: 5, Informative

    From TFA "According to CEO Jean-Manual Croset and Director of Community Charles Schulz, the Mandriva server products will be based on the Mageia distribution of Linux, while desktop and OEM products will be based on the historical Mandriva Linux distro." Desktop and OEM offerings to be based upon Mandriva not on Mageia and server to be based on Mageia not Mandriva.

    1. Re:OP got it wrong! by Anonymous Coward · · Score: 0

      And now the sloppy summary has been changed to refer to the "traditional Mageia codebase".
      Nice work confusing everyone, ./ "editors" !

    2. Re:OP got it wrong! by unixisc · · Score: 1

      But shouldn't it have been the other way around? Mandriva use their own in-house code for their servers (like Red Hat does), and use the community fork (like Fedora) for the desktop/OEM versions? After all, servers are what they can actually sell, whereas desktops are usually given away. So it would make more sense that their employees work on the servers, wouldn't it?

    3. Re:OP got it wrong! by eric_herm · · Score: 1

      In fact, the whole point is that in house is a simplification. mandriva SA is controlled by 2 heads, one from JM Crozet, a switzerland business man, and the other by leonid Reiman, a russian "businessman", whose record on the web should explain why no one would say no to him. So basically, what is the reason of a complex strategy is the result of a compromise :
      - desktop is mainly targetted at russian school and the rosa distribution. Rosa pay people who pushed the incompatibility with Mageia for innovation ( ie, rpm5, new installer, trying to drop drakx tool for python based one ), so they do not want to go back to what they have seen as non working ( of course, only in their mind, just see the drop of mandriva on distrowatch ).
      - server is where JM Crozet aim to get money, and for these, he need a solid ground, and that's not with the current desktop offering ( where almost everything is ligthly maintained, without a clear roadmap or server feature such as pxe/auto installation, proper ldap/kerberos integration, non broken software like puppet, cfengine, apache, etc ). So he took the safest way of using Mageia, if only because Mageia is dog fooding servers ( Mandriva either switched to Debian or kept unupgraded servers ).

      So yeah, that's what they try to explain, without saying the real reason. And the whole "let's give back to the community" is a mess too. If you take a look at those who are in the idea, that's mainly people who are on Rosa side, because they have seen no one competent besides them want to work with them. But the whole stuff is broken, just look at the current controversy around mandala Linux. Despites being the first result ( http://poll.pollcode.com/g9vn_result?v ), russian investor push people to vote for something else, because manda mean pussy in russian. They have some teenagers on a random forum claiming to have gamed the poll ( without any proof, as if people where not boasting on forum ), and so decreted that the current poll is flawed ( but of course, only after seeing the result were not what they wanted ), and they will not use it. Not a good start for community.

  2. No problems by colinrichardday · · Score: 2

    Yeah, no problems keeping those straight.

    1. Re:No problems by Anonymous Coward · · Score: 0

      With a distro name like Mangina, I don't think they care about keeping anything straight.

    2. Re:No problems by Anonymous Coward · · Score: 0

      From the highly respected piece "The Gay Linux Conspiracy" come these informative facts:

      The Mandrake product is run by a group of French faggot satanists, and is named after the faggot nickname for the vibrator. It was also chosen because it is an anagram for dark amen and ram naked, which is what they do.

  3. Redhat/Fedora by gmuslera · · Score: 3, Insightful

    They hadnt a lot of problems juggling between those codebases neither. Not sure if Fedora is to Redhat Enterprise like Mageia to Mandriva, or is a totally different beast, but it could work as precedent.

    1. Re:Redhat/Fedora by Anonymous Coward · · Score: 0

      They hadnt a lot of problems juggling between those codebases neither. Not sure if Fedora is to Redhat Enterprise like Mageia to Mandriva, or is a totally different beast, but it could work as precedent.

      Hollar to ya fellow Huckleberry Finn Language Institute grad!

      We got it real tough nowadays.

    2. Re:Redhat/Fedora by nxtw · · Score: 1

      Each major RHEL release starts with a fork from a Fedora release.
      RHEL6 is apparently based on Fedora 12/13

    3. Re:Redhat/Fedora by Yaa+101 · · Score: 1

      Mageia is a fork of the rock solid Mandriva 2010 codebase, Mandriva crashed with their experimental 2011 codebase and threw out the main builders of the product, these builders united in september 2010 to start Mageia.

      They used the Mandriva 2010 to build the servers that make the distribution (puppet based) and made the Mageia 1 distro with the servers in june 2011.
      At this moment Mageia 2 is out there but it is not as stable, the product is in the middle of changing from startup scripts to systemd and has dbus and policy kit problems, half of my system does not startup in systemd mode so I am using startup scripts as I always have done.

      The Mandriva they want to use for the desktop is less stable than the Mageia 2 they want to use for servers.
      I use Mandrake since ancient times and stayed with Mandriva and now with Mageia, I am used to it and it mostly works out of the box.

      Only that French company, it has always been and probably stay a disaster.

    4. Re:Redhat/Fedora by unixisc · · Score: 1

      Wasn't Fedora actually a fork from RHEL? At what point did the 2 get interchanged in terms of roles?

    5. Re:Redhat/Fedora by mellyra · · Score: 1

      Fedora is basically the experimental version of RedHat - Red Hat uses Fedora to test the integration of bleeding edge (Desktop) technology which then will eventually end up in RHEL.

      In my (3-4 years old) experience this results in every Fedora version upgrade breaking something new.

      I think the original Fedora codebase (when they migrated from being a set of extra repositories to being a full-blown distro) was taken from the (discontinued) Red Hat Linux, with some caveats (see above rgd stability) Fedora has taken the role of desktop Red Hat Linux.

  4. Diety du Jour by Anonymous Coward · · Score: 0

    Mandriva - which Hindu god/goddess is that?

    1. Re:Diety du Jour by unixisc · · Score: 1

      None. Mandriva = Mandrake + Connectiva. That's how it got that name - from their merger.

  5. who cares? by THE_WELL_HUNG_OYSTER · · Score: 0

    Titles says it all.

    1. Re:who cares? by MickyTheIdiot · · Score: 2

      Who cares that you don't care??

    2. Re:who cares? by osu-neko · · Score: 3, Insightful

      I love it when people do this: click through to read articles they claim they're not interested it, apparently unaware of how websites track reader interests. Every click on an article, going in to read it in full, is literally a vote for more articles like it. It's your way of saying, "Hey, I love these kinds of articles -- they interest me -- please post more like this one!" Of course, if you're afraid one click isn't enough, there's a way to totally trump that and magnify your vote for more articles of the sort: actually post a comment! That indicates a level of interest above and beyond, and adds more content to the site, which sites crave. The more discussion an article generates, the more sites love it.

      If your idea here was to indicate how much you really want to see more articles on this topic, you've done good. OTOH, if you would rather not see more articles on this topic, you've just done the most stupid thing you could do to try to indicate that. The bean-counters don't take time to actually read every comment, they just count the votes, so what exactly you post, what you say in the post, is irrelevant. All that matters is that you posted, and people responded, generating even more content for them.

      --
      "Convictions are more dangerous enemies of truth than lies."
    3. Re:who cares? by THE_WELL_HUNG_OYSTER · · Score: 1

      Your mom?

  6. Step one: Call it Mandrake again. by Picass0 · · Score: 4, Insightful

    This is exciting to see. Giving the community greater influence over the future development of the distro has put this on my list to watch. I've used Ubuntu and Fedora (laptop and desktop) for years, but I used Mandrake years back and would be open minded to doing so again.

    1. Re:Step one: Call it Mandrake again. by dmomo · · Score: 1

      I also enjoyed Mandrake for quite a stretch. At some point, they started littering the install process, and then the desktop with ads. I respect the need for a revenue model, but I feel the experience was too intrusive. There were plenty of alternatives, so I jumped ship.

    2. Re:Step one: Call it Mandrake again. by Anonymous Coward · · Score: 0

      There should be a meme of "What linux looks like when left to the community." I'm guessing it looks like monochrome background, CLIs galore, no real mouse drivers to speak of, and support for 1,000 monitors simultaneously.

    3. Re:Step one: Call it Mandrake again. by Picass0 · · Score: 1

      There are many example of open source projects where the feedback of the community is solicited by developers and over time results in improvements. Blender and Gimp are two that jump to mind, but there are many others. Mandriva becoming a distro that is tailored by it's fans has great potential to serve as an example.

      Cynicism is an ugly trait. You should work on that.

    4. Re:Step one: Call it Mandrake again. by unixisc · · Score: 1

      But wouldn't using Mandrake re-invoke copyright battles w/ Hearst corporation, who had already won a case against them in 2004 over the use of that name? As it is, Mandrake Corporation was already in trouble due to the use of the magicians's bowler hat and wand as its logo, and to top that, the use of a Linux tool named Lothar, who in the strip was Mandrake's friend and principle body-building aide. The merger w/ Connectiva and the resulting name change resolved that issue. The last thing the company needs to do is start up new court battles.

      That said, I liked the look of that distro when it was out - a pity I couldn't use it b'cos at the time, it couldn't recognize my network card (in fact, none of the Linuxes I tried out - TurboLinux, Caldera, Corel Linux, Storm Linux, et al could). Which is why Linux didn't work for me @ the time.

  7. Whats the difference? by vlm · · Score: 1

    Whats the difference between desktop and server other than what marketing has tried to create?

    Imagine having to support two versions of mysql, one on the KDE desktops and one on the backend server. Lovely.

    --
    "Science flies us to the moon. Religion flies us into buildings." - Victor Stenger
    1. Re:Whats the difference? by MickyTheIdiot · · Score: 1

      About $5000... ba da bum bum...

    2. Re:Whats the difference? by Anonymous Coward · · Score: 0

      > Whats the difference between desktop and server other than what marketing has tried to create?

      Look at other things besides your navel and you'll there are other sciences beyond the engineering / marketing dichotomy.

      What's the difference between a desktop and a kiosk at a shopping? And a point-of-sale thin client? And a node at a beowulf configuration? Do you think we could use the same thing all the time?

      Just a hint for you: servers must have maximum performance, sometimes not from the processor but from the I/O subsystem; desktops must have only enough performance -- in fact, sacrificing performance for less latency is a worthy goal to pursue.

      That alone implies different kernels.

    3. Re:Whats the difference? by vlm · · Score: 1

      What's the difference between a desktop and a kiosk at a shopping? And a point-of-sale thin client? And a node at a beowulf configuration? Do you think we could use the same thing all the time?

      Uh, yeah, thats kind of the point of a universal operating system. Dramatically lower support costs.

      sacrificing performance for less latency is a worthy goal to pursue

      You've got to be kidding. Why would I want a high latency server? Besides you're arguing very small percentage gains.

      --
      "Science flies us to the moon. Religion flies us into buildings." - Victor Stenger
    4. Re:Whats the difference? by Anonymous Coward · · Score: 0

      > Uh, yeah, thats kind of the point of a universal operating system. Dramatically lower support costs.

      Why don't you invent a special tool which is a spoon and also a fork. Think of the scale savings! Oh, wait, someone probably already did it.

      I wonder why people still use spoons. Why? (irony)

      I have a phrase for you to think when you're close to your pillow: "the right tool for the right job".

      > Why would I want a high latency server?

      For the same reason people build trucks which have greater latency than a bike: to get other features of interest (like increased cargo capacity on a truck or providing centralized services on a server). Why would anyone pay to have a nimble truck?

      > Besides you're arguing very small percentage gains.

      Which can mean the difference between life and death! (when you're playing a FPS, that is). On the serious side, latency is important when several small ones accumulate and give a sluggish feeling or when you're doing multitasking and the other tasks make your life miserable. To feel its importance, choose a perfectly working system and add small delays (of few microseconds) in all tasks the system do (not only at the interface) and then you feel like your brand new machine suddenly looks like from 7 years ago (I'm certain a Windows joke would fit nicely here, but this is getting too long).

  8. How the f** is it this company still exists? by Anonymous Coward · · Score: 0

    I really have to wonder who on earth uses Mandrivia. It died in the 1990s. The only thing interesting that has come out of it in the last 15 years has been its recent fork and that was mainly due to it having a grass root organisational structure. This however isn't exactly new. Debian is also run using a similar non-corporate model.

    I would love to see what a list of distributions and what differentiates each. What are the aims of Debian? It would seem to me it's aim is a universal distribution. Or in practice a distribution in which to base other distributions. In practice its model also leads itself very well to web servers. Ubuntu seems to be a distribution in transition from targeting desktops to tablets/phones/home theatre entertainment systems. Redhat Enterprise Linux is clearly aimed at servers used in a corporate environment that need really really long term support. Fedora seems to be a base for Redhat Enterprise Linux. As a end user distribution it doesn't work terribly well (too many bugs- you would think things like AR9170 USB chipset would be well supported given its a free software friendly chipset- but it's not). Linux Mint seems to be a hobby distro which fixes some of the mistakes of Canonical related to UI (although adds some new mistakes in the process). Trisquel is clearly a desktop distribution aimed at the freedom loving crowd. It doesn't just claim to love freedom it actually implements it (not something that can be said for any other major distribution- although Debian gets pretty close and would be a major distribution). Then you have OpenSuSE. Does anybody use this? It reminds me of Mandrivia. It doesn't have a big enough base and bugs don't get fixed that should. I guess it's the base for Novell's practically defunct OS. Novell is now owned by Attachmate. A company known for ending development and barely keeping products on life support. Cent OS is the distribution for individuals at companies too cheap to pay for the real thing (Redhat Enterprise Linux). Puppy Linux is a distribution for users whose systems are so old all they are really doing is playing around for fun. PCLinuxOS seems like a dead OS although had a lot to like for desktop users. Unclear and lack of info just makes me wonder whats going on? Some clarity and updates on some sort of blog would be nice. I know the lead developer got sick and is taking time off and there is no clear leadership with the lead developer absence. The project is or was "falling apart" due to all the fighting. Ultimate? What's that? Lubuntu is a version of Ubuntu that seems designed for the desktop that is light and good for older systems or those looking for a non-KDE/GNOME/Unity desktop. Sabayon last I heard was a distribution designed to show off the cool effect of 3D desktops. Anybody who gets a desktop for this reason is just messing around. Freebsd is... wait that isn't a GNU/Linux distribution! Chakra? Whats that? Pear? hmm sounds familiar but not even going to try. Zorin looks like a really cool up and coming distribution designed for desktop users. It's targeted at Microsoft Windows users. It seems to be working on a potentially successful business model although is still young and with growing pains. Mistakes here and there (stop selling 'premium' flavours of GNU/Linux- it won't work- but your hardware angle has merit- and you could safely charge a few dollars for downloads or maybe access to your update server.. just keep it freedom friendly...). Overall a nice distribution. Then you have some like Slackware and Gentoo. If you actually utilise them you are probably a developer. Knoppix is a dead distribution that fostered the first great regularly updated live distribution (a prior distribution predate it that never got updated). Backtrack is a distribution for "hackers" (ok, lets me honest, you aren't a hacker if you use this distro, you are a script kiddie wannabe who doesn't know a damm thing about GNU/Linux, freedom, or what a hacker really is, because if one did you wouldn't be using backtrack). Tails is a unique distributio

    1. Re:How the f** is it this company still exists? by BanHammor · · Score: 1

      Parted Magic is useful for recovery partitioning, the kind that you couldn't really do on a hot install. I would like to know why are you so antipathetic to Backtrack, as it ships with some tools that other distros wouldn't. Wireshark, aircrack, all that stuff. You have mentioned so many distros, including forks of Ubuntu, some absolutely unknown ones like Pear and Zorin, but only mentioned Mageia by its' grassroots structure? That is not as important as the aim of the distro. The aim is to have a clean, minimally tampered with user-friendly distribution while still preserving some tools like MCC of Mandriva's.

    2. Re:How the f** is it this company still exists? by Anonymous Coward · · Score: 0

      :) that was my post. Or more of a rant. Surprisingly it really wasn't terribly biased. I support GNU/Linux for a living so I actually get to work with various distributions on a regular basis. I'll give you some clarity as to some of frankness and perspective of the comments of each.

      As far as I've been able to tell the majority of useful utilities in Backtrack are available in nearly all other distributions. Why bother downloading it? If you are a technical user which is what hacker implies this is sort of a dumb distribution. The other reasons I put it down is because it's a "hacker" distribution when hacker is a term to describe endearment to someone who likes to learn. While it can be applied to someone who finds holes/bugs, breaks into systems, and so forth that in and of itself is not hacking. Hacking is more along the lines of what the Tor developers do and others. They research, learn, implement, and enjoy that process. Running a point and click tool that somebody bundled (what backtrack does) and somebody else wrote is not hacking. Trisquel developers are hackers. They are breaking down barriers to learning (a truly free distribution that excludes non-free software/drivers... something very rare amongst distributions).

      I don't recall what Pear does. It might be a mac emulation distribution or something. Although I think that was PearPC- a project that had nothing to do with a distribution. Anyway. The mentioning of a distribution merely means I've heard of it. The comments are what really matter and you can't put too much value in them. There is no way I could fairly evaluate all these distributions. I merely summarised (quite badly) what I saw as the gist of the distribution. Zorin is a fairly obscure distribution although there is something behind it. It's basically a guy (I'm imagining) who has written some free software to pretty the desktop and implemented various 3d pieces. He has made it very usable and user friendly. Besides from that it's basically Ubuntu. The usefulness in this distribution is its ease of use and potentially the hardware/support it ships with. His business model seems good although I'm not entirely sure he alone is going to be able to handle it all. He should be working with a company like ThinkPenguin. ThinkPenguin has operations in the US and UK. They focus on free software support from a hardware angle. They work with chipset vendors and manufacturers as well as distributions like his. They don't ship hardware dependent on proprietary drivers, have a huge catalog (the largest in the world), multiple distributions points (US & UK), and end-user support.

      Mageia might have some potential because of its grass roots effect. People tend to favour distributions with a community component (even where they are owned by corporations like Canonical). I probably should have mentioned I feel that the bureaucracy is holding Mageia back. Despite that it could be a distribution that is around for many many years to come.

      I agree that the aim of the distribution is more important. I just fail to see what differentiates Magia from other distributions. They have some pretty interfaces and it might work well. However you need to have a large community with lots of developers to fix bugs. That is something Magia does not have. Most of the Ubuntu distributions work so well because of the larger developer base Canonical has and the community around it. Even with a shrinking Ubuntu people are moving to other Ubuntu derived distributions. That means those users are still going to be contributing back to Ubuntu (big fixes, etc).

      Unless there is something revolutionary about the MCC I don't think it is enough. In 1990's it was something. But back then we didn't have solid debian based derivative distributions aimed at the desktop. Now we do so MCC is only part of the equation. What is probably missing from many of the Ubuntu derivatives is probably MCC. Ubuntu 10.04 was the prime example of a solid usable distribution. It wasn't perfect, but overall it worked well. I'm hav

    3. Re:How the f** is it this company still exists? by unixisc · · Score: 1

      Pear, at one point, seemed like it was trying to create a MacOS desktop out of GNOME3. It took GNOME3, which everyone at the time was complaining about, made some changes to it to make it look completely like OS-X, even w/ its own app store and everything. Only way one would notice the difference - Opera instead of Safari (which doesn't exist for either Linux nor BSD, AFAIK), and a few other apps in the docking table. But more recently, after they became Comice, they seem to have shifted to covering their bases by having separate distros for KDE 4.x - the desktop, netbook and even Plasma Active.

      The things that I start w/ in most distros is whether they support networking - preferably wireless, but even wired? If they do, I'm willing to try them out further. Since the KDE team has put together a whole host of apps, I'm more willing to give that a try. I certainly hope Calligra picks up & does well.

      Personally, the things that intrigue me more are the BSDs. With Linux, I've seen cases where unless I have the right combination of an OS version plus the relevant driver, it will not necessarily work - I saw that clearly w/ ALSA. Apparently, the lack of a device driver ABI in Linux - the way it's there in BSD - is an issue. W/ the BSDs, they tend to be less distro-happy, even though there are a few moribund distros there as well, such as Desktop BSD, PicoBSD and a few others. Since they tend to provide as many desktops as they can, you don't see too many distros in the BSD side, particularly if the only distinction is the choice of UX. It would also be nice to see Minix making some inroads, riding as it is on the coattails of NetBSD.

    4. Re:How the f** is it this company still exists? by eric_herm · · Score: 1

      Burocracy, the name that you use because "discussing with other and try to be democratic" or "planning before doing" is not a good term when you want to criticize people.

      And you do realize that Mandrakesoft start in 1998 ? So when you say in the 1990, you are just saying it was good the 2 first year and that's all ?

  9. Mageia = name fail by Anonymous Coward · · Score: 0

    Can't the combined candle power of the world's smartest people come up with something besides "Mageia"? Even the Color + Noun meme of the 90s (Red Hat, Yellow Dog) would be better. Or the Adjective + Animal meme of the 2000s. Instead we get a name that's unpronounceable in any language?

    How about calling it New Drake Linux or something!?

    1. Re:Mageia = name fail by sumdumass · · Score: 1

      I'm sure they could. However, I'm not sure they are connected with mandriva in any substantive way

    2. Re:Mageia = name fail by Anonymous Coward · · Score: 0

      Can't the combined candle power of the world's smartest people come up with something besides "Mageia"? Even the Color + Noun meme of the 90s (Red Hat, Yellow Dog) would be better. Or the Adjective + Animal meme of the 2000s. Instead we get a name that's unpronounceable in any language?

      How about calling it New Drake Linux or something!?

      To be honest I always thought Mandrake (and the wizard logo) was the dorkiest thing ever. Followed up by Mandriva which is one of the clunkiest names ever. Mageia is infinitely sophisticated in comparison.

    3. Re:Mageia = name fail by unixisc · · Score: 1

      Why not combine forces w/ PCLinuxOS? Egos? B'cos that name at least seems more generic, and people might think of it as less strange, than say, something named Zorin, or Comice, or Mint, or Ubuntu, or so on.

  10. Must... resist..... by Anonymous Coward · · Score: 0

    All your codebase are belong to us!

  11. Mandriva 2011 was a disaster. They should ditch it by MROD · · Score: 1

    Having used Mandriva (and Mandrake before it) ever since Redhat split its distributions I tried the 2011 version... It was a complete pig's ear of a release, especially if you want to integrate it into a shared network or use it for real work. The worst part (other than systemd and its intrinsic brokenness) is the default "Start" menu replacement. (Oh, and the WiFi is completely broken, the wired networking half so.)

    Mandriva 2010.x was stable and worked very well and this is the basis for Mageia.

    If there were anything they should kill it would be the "desktop" version, start with the old code and move forward.

    As to the anonymous coward who wrote the essay on how bad Mandrake/Mandriva is, I'd just show him URPM, the distro installer (for 2010.x) and compare them with the other distros' solutions. They pale beside them.

    --

    Agrajag: "Oh no, not again!"
  12. MAGEIA code base is not MANDRIVA by Anonymous Coward · · Score: 0

    Mandriva is trying to use the community to stay alive. MAGEIA forked from Mandriva a couple of years ago now and is NOT Mandriva.

  13. Mandriva/mageia:there's nothing left to talk about by Iron_Fist · · Score: 2

    The Mandrake Team created a dependency tool (urpmi) at a time when only debian did, and poor redhat users had to download dependencies by hand.
    The Mandriva Team improved on the -drake family of tools, and came up with a centralized configuration panel : the MCC ; SuSE was doing the same ahead of 6 months, and poor debian users had to dpkg-reconfigure each packages by hand.

    In all that time, it was still the same people doing the good job (Pixel, warly, fpons, and so on).

    Now that they have all left (fired or underpaid), and are not contributing to mageia either, you should realize that you are talking about a completely different product which only retains the name and the history of its ancestor, with a freshly hired off-shore development team.

    Even the Mageia Team is no more than a shadow of the original team, with former interns, and support engineers made software developers.

    There is no magic is this world.

  14. Mageia/PCLinux by unixisc · · Score: 1

    Why don't they use Mageia for servers and PCLinux, which is another Mandriva fork, for desktops?

    1. Re:Mageia/PCLinux by Yaa+101 · · Score: 1

      PCLinux is an old fork, probably it is too far from the codebase of Mandriva.

      Texstar (PCLinux' maintainer) used to be a packager for Mandriva years ago, he rpm'd a lot of applications in those days, and the packages were of exceptional quality.