Slashdot Mirror
Mandrake Linux Development Process Changes
joestar writes "Just found at MandrakeLinux.com: 'MandrakeSoft today announced a major evolution in the way that future Mandrake Linux distributions will be engineered and released. The purpose of this new development process is to provide the highest level of new features, as well as maximizing the quality of new products.' In short: for each release, there will be a 'Community' release, equivalent to a common Mandrake release, with all latest features. Several months later an 'Official' release - based on the 'Community' - will be available. Both of them will be released publicly and supported. The new process will start with the upcoming Mandrake 10.0."
Here is the problem I see with this. They are trying to have their "Official" release be less buggy than recent releases. They claim that the problems with the recent releases are because not enough gets ironed out in the betas.
So, they are breaking the final release into "Community" and "Official" branches. Won't the "Community" release eventually become synonomous with "beta." In the end, fewer people will run this community release, and fewer bugs will be found in it. If this happens, problems will undoubtedly creep into the "Official" release and only be found then because more people are running it.
Anyway, it seems to me they are just trying to rename the word "beta," which is not a solution to the problem they are trying to fix.
-- Fighting mediocrity one bad post at a time.
Nobody gets paid to use the 2.5 series kernels to prove 2.6 is good. There will always be people out for the latest & greatest... and plenty of whom would not care about being paid.
SIG: HUP
I honestly don't think it sounds like that bad an idea. Most home users don't need the testing and would like the features. With easy updating most home users can afford to use a less tested package. And for those who do not like the idea, they can wait for the official release. It gives them a situation akin to Debian's unstable/stable development where the stable branch is solid but aged, and the unstable branch is usable but current.
I do security
I think it's a way of getting people to join the Mandrake Community which is not very expensive anyway. I see it as a gentle nudge for those who actively use Mandrake and want immediate access to new releases. It really is for a good cause. I support their decision as I use their Linux distro on many customer sites.
Suncoast Linux - Sarasota, FL
...as far as I can tell, is that they're adding a "gamma testing" phase between the open-beta-test phase of the Cooker process, and the official put-it-in-boxes-and-call-it-done release. Seems like a reasonable move, because it lets users be a little more granular in deciding just how bleeding-edge or risk-averse they want to be with new versions.
http://alternatives.rzero.com/
And Mandrake Cooker is 5 years old. Both Fedora & Cooker are experimental and quite buggy Linux distributions. The new Mandrake development scheme is an extention/evolution of this process, but you cannot compare it to Fedora. If you really need to compare it to something else, compare it to Debian stable/unstable branches...
Well, the community version of Mandrake is just an early build of the release. Unlike Red Hat, Mandrake is actually releasing the "official" version for public download, as it's stated "after a short delay, Mandrake Linux 10.0 Official Download Edition will be made available on public FTP mirrors". This is a signficiant difference from Red Hat's business model.
It's a bit more like Debian than RedHat's model I think. The Community release will be functionally complete, but has bugs.. i.e Beta. The Official release will be the Community Release put through a QA process which seems to depend heavily on feedback from Community users. This is pretty much how I've seen Debian handle it's stable/unstable branches, although I'll admit I pay less attention to the Debian dev process than RedHat's.
Personally, I think it's not a bad model for getting higher quality on a shoestring. I don't think Mandrake is out of the deep water yet, so I definately commend their ability to find innovative solutions to providing higher quality in their products.
Fedora seems to be a sort of less public version of this policy. Fedora (Community) users add features and test the Beta quality software. The cream is incorporated into RH products and put through traditional QA testing, which is probably a much larger operation than what Mandrake can muster.
Just my 0.0160900 EUR on the announcement.
Arrogance is Confidence which lacks integrity. -- me
And the great thing about all this is that those who want the bleeding-edge stuff get it, and the others (e.g. newbie, corporate user, etc.) get something that's stable right out of the box. I can see this as being a great way to sort out the bugs before the corporate guys install it and toss it to the side after playing with it for 10 minutes. This way, they get something very stable and usable that could gain widespread use throughout various companies.
This is a very smart move by Mandrake. The Cooker idea worked well, but it was never officially promoted in a big way i.e. you never hit the Mandrake home page and saw "Cooker release 9.2 available for download"
,you must know what you are doing - and you'll just end up using Cooker anyway.
,but not experienced enough to compile their own kernels. And that's a good thing - more exposure to a wider range of platforms and better bug feedback.
Dare I say, but it sounds very like the Debian way of doing things (unstable - testing - stable).
But there's a double-edged sword with doing things this way , in that you'll never have the bleeding edge stuff in a "Community" Mandrake release.
But then, if you want that
imho,the Community thing is more aimed at the general casual Linux user - a bit experienced
As an example on why they had to introduce this (possibly), the much advertised MandrakeMove Live CD doesnt even recognise some PCMCIA wireless cards in laptops. A bad oversight.
A MandrakeMove community edition would have helped in identifying this glaring omission.
Overall, it's a big big thumbs up from myself - well done Mandrake for introducing the Community Edition idea.
We'll end up with a Mandrake stable, which we've never had (and probably keeps them off some corporate desktops) and I'd imagine the testing Mandrake will be pretty much the same as regular Mandrake has always been (bleeding edge, sometimes buggy and still the best of both worlds).
For those posters complaining about the new 'Official' release being out of date, bleeding edge will *still* be in the community version, nothings changed. I'd guess the 'official' version will focus more on thier new Corporate desktop push and configuration/usability technologies. Makes perfect sense to me and maybe we can see some more serious usability enhancements (DrakConf is great, but not much has changed lately) now that some of their costs will be more focused (if the community comes together, which seems pretty active already in the club).
Quack, quack.
What they ultimately GET is self-evident. If you're downloading betas and release candidates and providing feedback (via mailing lists, bugtraq, whatever), what you end up getting is a more stable distro with features yourself and others have requested. Input is valuable in projects of this scale *particularly* with the unbelievable variety of x86 hardware. What works for you may not work for someone else with identical hardware, differing only by something as miniscule as a single piece of usb hardware.
Good Troll, but *Red Hat* is the one innovating here.
I'll be sure to tell the Debian project that the way they've been doing things for the past 10 years is now an "innovation" from Red Hat.
Jay (=