The three CDs for the third beta of Mandrake Linux 9.1 are still getting uploaded to most mirrors around the world, but you can already download it from here or here for now, until all the mirrors receive the files.
Mandrake Linux 9.1 Beta 3 Available
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
Is there a change log? Do they usually include one?
Does this mean we are getting close to a final release?
Not really. The previous release of Mandrake, included 4 Betas and 3 RCs, which was spanned across ~4 months.
I would install it but I’m runnin’ kinda low on CD-Rs. If anyone here would be kind enough to provide some sort of synopsis of the changes it would be greatly appreciated. Mandrake betas are usually stable as necessary for most users willing to use experimental software (Including myself), but I’m more concerned with how well everything “Just works” as well as good looks.
It has been barely 2 weeks since the last beta and they come out with another one. I don’t believe Mandrake has enough time to get feeback, get time to fix the problem reported and release. It is like constantly running after a moving target.
you are not subscribed to the Cooker mailing list, are you?
They are receiving reports of (and working on) about ~70 bugs a day.
That’s plenty of work, if you ask me.
I certainly have to agree with your assertion. My gut feeling tells me here that they are trying to release as soon as possible, before another release of Red Hat or SuSE takes place, because otherwise they won’t sell as much, and MandrakeSoft is with one foot in the grave, financially. atm.
I am really not sure that if they are so quick on their betas that their final release won’t have many bugs. I mean, 9.0 was buggy enough for me already…
Good luck to them though. I believe that they are waiting for the Gnome 2.2 and XFree 4.3 releases in February and then they are going directly for a release… I just hope that this time the boxes will be ready for purchase at the same day the free version comes out. The 9.0 release had almost 1 1/5 months of delay between the free and commercial version, and that really killed them financially. I hope they won’t do the same mistake twice.
Damian, you don’t think you understand what Jean_J is saying. Of course and they have many bug reports. Of course and that’s a lot of work. The problem is that creating betas of a whole OS every 2 weeks and have thousands of downloads and subsequent reports is just overkill. For open releases like these betas, they should at least wait a month or more, before rolling out a second beta. That would give them more time to iron out most of the problems, instead of getting the same reports over and over again in the mailing list from new users that they haven’t seen the previous reports. It is just too early to release a beta every 2 weeks for a whole OS. If that was a single application, I wouldn’t have a problem with frequent releases. But these releases for an OS every 1 or 2 weeks are overkill.
two weeks is better than the 8 day time frame Mandrake had between beta1 and beta2’s release. I’ve finally got beta2 config-ed to a point where it’a semi usable. I think I’ll hold out until that first RC cnadidate before overhaulign my linux partition once again.
All Cooker bugs are reported thru Mandrake’s bugzilla. Duplicate bugs are easily
and quickly marked as such… But then, who knows, maybe you guys are right
and two weeks is too little time… and maybe they had a reason to feel it was
time to give people a third beta.
The only way me or anyone could really tell is if we were working with them…
There were only 13 days between this beta and beta2. I really believe that MandrakeSoft just tries to release ASAP. It makes sense actually, if you think of their financial situation. However, I don’t keep my hopes high that their final release will be either stable or somewhat bug-free, if they keep this crazy pace.
Eugenia, When I was beta testing Windows 98 I was getting new builds about every two to three weeks. Admittedly that was a closed beta program but it shows that a two week timeframe is not in and of itself abnormal.
mandrake really want to get the 9.1 out of the door quickly. Maybe, it’s true, this might be the LAST ONE.
When you only have 1000 or 2000 testers, there is not a big problem. But when you have 100,000 ones and in every beta you get _new_ testers who don’t know what it has already being discussed/submitted before to the list, it really becomes a problem (even cleaning up the db for duplicates is a lot of work – an engineer’s time is always valuable). Also, Win98 is less than a CD. About 300 MBs I think. Mandrake is 3 full CDs with a zillion apps on them. When Apache or your Konqueror crashes, you still submit the bug in the MandrakeSoft’s bugzilla, so that really adds up a lot.
If Mandrake had a closed beta or a known number of testers, things would have been more manageable, but this is not their case. For these public betas, they should take their time.
I remain firm on my opinion that these public betas are rolling out too early and the end result won’t be what it is ought to be. I did say that there is a good reason for doing so (financial reason), but as an end user and consumer, that bothers me.
If they are speeding up to 9.1 (which IMHO doesn’t seem like that to me since 9.0 and 8.2, IIRC, started firing betas and RCs this fast), they are dead wrong if they think that 9.1 would bring them to safety.
Their product is almost okay, the big problem is with the services. Geeks hardly buy services and there is not services from Mandrake for the consumer and corporate market (the latter being more profitable). In other words, even though 9.1 comes out, beats every other distro in town, gets a nice review from you being better than Windows and Mac OS X, etc., they would still go through financial ruins.
Why? Their main revenue stream (boxed software, MandrakeClub) isn’t making enough especially since the free download edition fits almost everyone.
How many really serious testers are there? If you look at the number of downloads of the stable version at cnet, it would surprise me hughey if there were not morre than a few hundred. Most of whom will not work with the releases, they evolve with the cooker. If they keep the same cd’s they just keep getting bug reports on solved issues.
Eugenia, I was not saying that there is not something wrong with this release cycle, just that I my experience it is not unusual.
Getting reports of the same issues is okay. Getting reports of issues reported in previous versions is downright perculiar. Would a shorter time span for each beta and RC means more bugs being fixed? I doubt it. By increasing the time span between betas, they can fix more problems and when they release a new beta, most of the problems would be new or unreported problems.
I am a yellowTAB developer ( Future BeOS Solutions ).
We have been busily developing a distrobution of BeOS named ‘Zeta’ based on licensed material created since R5 and BeIA.
We have reached from Beta 1 to Beta 3a in relatively short fashion for the task at hand, and yet we are constantly being insulted that we have not yet V 1.0.
Of course, the biggest reason is very much the lack of public relations, but one nearly as big is the lack of patience. Not from the BeOS community as much as it is from the outside community wanting to try something new.
yellowTAB’s team is small. All of us who work on Zeta are doing so with whatever free time we have. And with this, we have managed Four betas (counting 3a vs 3) in relatively short order. Each version light years from the previous and actually essentially ready in merely a week or two, though bugs are waited for and resolved prior to the next release to the internal developers.
yellowTAB has a slow bug turn over, though we have less bugs to begin with ( and far less, but uniform, code ). Mandrake has a high rate of bug turnover, and are very familiar with their code. From what I have heard, each beta has been an improvement over the last. This is all that matters.
You are not being forced to download the latest. If you do not wish to do so, then simply wait for the final.
Even within yellowTAB, we have a simple, but effective, bug tracking mechanism. And with that, cvs is also a big help in organizing such a trully monstrous task of developing an Operating System. Mandrake undoubtedly can outdo our little yellowTAB startup in pure volume and organization, so why would you not EXPECT them to be able to outdo our (or most others’) abilities?
What I wouldn’t do to be able to release betas as quickly as Mandrake. Zeta would be at version 5.0 (yes, starting at 1.0)!!
Some people just don’t realize what they have when they have it. I’d kill to be complaining about the same things with Zeta’s development!!
–The loon
Mandrake would have been the best User (as in opposite to geek) distro if they would have behaved a little better.
After 8.0 they started swapping out betas and RCs just too fast, made a stupidly high jump from an instable 8.2 release to an instable and buggy 9.0 release. Already 8.1 did not work for me, the networking subsystem was not working, never actually, and I had two professional sysadministrators (Linux and BSD) sitting at my machine, had help from #Mandrake, to no avail. And no, my PC is pretty standard, no exotic hardware. 7.2 and 8.0 worked (except for sound, but that was not a Mandrake thing) flawlessly.
Mandrake moves for about 1 1/2 – 2 years between NiceThing and HypeThing. They lost me as a customer after 8.1 (I purchased the PowerPacks). Now, tell me all, but I am not surprised they go down the line…
.john
Looncraz, you keep posting stuff without bothering to understand what you are reading first. YellowTAB’s procedures have nothing to do with these of Mandrake. YTab only releases to a HANDFUL of BeOS testers and your betas were spawned across many months of releases.
The problem here is that Mandrake does _public_ releases, which draw thousands of downloads, resulting a chaos in the bugzilla and the efficiency of the engineers. YTab’s situation has nothing to do with Mandrake’s. Don’t compare apples and oranges.
A comment From a PCLO registered user:
————-
Re: Mandrake 9.1 Beta 3 Released (Score: 1)
by illogic-al on Friday, January 31 @ 21:33:21 PST
If you’re on the cooker mailing list you’ll see that it’s pretty active. As a matter of fact a couple of us had a pretty heated discussion about the login manager being mandrakized. It looked really crappy some words were said and a couple minutes later it got filed as a bug. Of course there are more serious bugs which are being taken care of. These guys are pretty on top of things. We test, they fix. It’s a beautiful thing.
———-
Correct me if I’m wrong, but shouldn’t they fix all the bugs that have been found (best they can) before they release a new beta? I’m speaking from a logical point of view, not a marketing one.
Mandrake wants to stay in the spotlight and maintain interest in their product, as Eugenia has stated, as well as provide beta /users/ more software on disk with each new release. Despite being beta, it’s useable – and the more software included in the release the better. And herein lies the real hook.
Mandrake 9.0 was a buggy release… people know this. People shot it down for this reason, and it lost popularity. But these betas are SUPPOSED to be buggy, so people using them tolerate the bugs just fine and see through to the real juices of the product. In other words, in a beta, users tend to look more towards the ‘neato’ features that are coming, and being on the cutting edge, despite the bugs… whereas someone using the release is smacked and left with a bad taste in their mouths when several overbearing bugs are present.
This sure is the way for Mandrake to get that sparkle back, as long as they continue releasing betas and eventually RC’s until the final release really is stable.
For now, I’m just thinking of these betas not as betas that are meant to find bugs (though that is a major purpose), but generally speaking, as a release with bugs, intended for the savvy user who wants to be on the cutting edge while able to deal with the quips that come along with it. And while they’re at it, they’re keeping the next release in the public eye, encouraging excitement with every beta release, thus reviving the popularity and interest necessary to revive their company from the staggering blow they have taken.
The only questions are… how many betas will we see? Is it worth installing now, with another release likely just over the horizon? What about then? How buggy will the final release be?
Let’s hope that Mandrake doesn’t lose sight of what’s really important – a full fledged release that is as stable and bug free as possible…
Faster betas, more interest, more reports, more bugs, more work.
Sure hope they’re up for it.
I think they should release often mostly to prevent duplicate bug reports with the early betas. They should gradually take more time as they get into later betas and RC’s.
Good point… snag is that you can go too fast… people get sick of installing (and perhaps reviewing) a new release only days after they got themselves set up with the last. They need to spread it out a little more I think, but so far as generating interest goes, so far, so good.
Most of the stuff they will be including is labeled as stable or will eventually be stable once MDK 9.1 is finish( for example Gnome 2.2, E17 etc..). Right know I think the bulk of the work that they are doing is in either re-making, re-designing, updating or fixing Mandrake only applications.
“Their product is almost okay, the big problem is with the services. Geeks hardly buy services and there is not services from Mandrake for the consumer and corporate market (the latter being more profitable). In other words, even though 9.1 comes out, beats every other distro in town, gets a nice review from you being better than Windows and Mac OS X, etc., they would still go through financial ruins.”
There Mandrake Club subscriptions have been on the rise lately and is really moving forward. They have services for every level of clientel from your average user to those who need high end support. Also as far as your “Geeks don’t buy services” crap. Well thats a bunch of bull$hit. Mandrake provides it’s user via the MDK Club with RPMS that a lot of us Geeks like to use. In fact their service subscriptions for the Mandrake Club services are very vauleable to me.
“Why? Their main revenue stream (boxed software, MandrakeClub) isn’t making enough especially since the free download edition fits almost everyone.”
Again you seem to be uttering a lot of non-sense. Mandrake sales have actually gone up in terms of software sales as of late along with their Mandrake Club Subscritpions. Their only problem is the contracts that their previous CEO locked them into when it came to providing for Linux Learning Centers. This is the root of their problem and not the so called lack of sales that you keep rambling on about.
There Mandrake Club subscriptions have been on the rise lately and is really moving forward.
Let me tell you why. Mandrake’s financial problems. Heck, Mandrake Club was created precisely for that. Hardcore Mandrake fans feel that it is right for them to join the club.
Also as far as your “Geeks don’t buy services” crap. Well thats a bunch of bull$hit.
Before you call any argument “bullshit”, why not you compare the amount of downloads and the amount of subscriptions. Most of the subscribers are sympathy members, not members for the software (if you want that, there was always the boxes..).
Their only problem is the contracts that their previous CEO locked them into when it came to providing for Linux Learning Centers.
Actually, if Mandrake actually develops it, these learning centers can actually be profitable. Now, just say Company A wants to move to Linux. They choose Mandrake Linux. So now Mandrakesoft has an opportunity to provide reeducation, which they get a rather inviting royalty.
But why it isn’t making any money? Simply because there isn’t much companies even thinking of moving to Linux, much less to Mandrake Linux. Mandrake’s sales and marketing forces are pathetic in comparison with the likes of Red Hat and Sun.
Mandrake has been releasing betas like gangbusters. This comes in the light of heavy criticism of thier 9.0 release and thier impending future after filing for bankruptcy. Perhaps beta3 fixes issues that 9.0 raised but were as yet unresolved in betas 1 and 2. Perhaps beta 3 fixes bugs from beta1 that went unresolved in beta2. I agree it’s a tight timeframe to work in to fix bugs judgine solely on the last release alone. I applaud Mandrake’s efforts of continually putting out these betas and staying on the bleeding edge of what linux has to offer. It’s a far cry, and a much better one that what other distributions are doing.
moo
yay more toys to play with lol
Mandrake is CR@P
are there any md5sums for the beta3-isos yet?
I’m looking for the md5sums as well. I have three ~650mb files sitting here waiting to be burned as soon as I can verify they downloaded properly.
vetinari: here are my md5sums:
CD1: 120610ea02bb6a3ad0f0de594723b4a2
CD2: 8114ad9397a08fa7d4aea94e34f183f0
CD3: 42e97b16574350ad8c15a109657b34fa
Let me know if anyone gets the same. Thanks.
So, they have cranked up the beta release schedule, that means nothing. All betas are just snapshots of cooker. You can buy cooker snapshots any day of the week. Therefore you can’t draw any conclusions as to how “fast” they are working.
Reading reviews from distrowatch.com do seem pretty encouraging, especially part 3. HPT372 onboard raid controller issue works perfectly?!! That would be a first. Also Mandrake may FINALLY be working on the font issue, at least going in the right direction. And although I never had ANY WHATSOEVER stability problems, only bug problems, stability looks to be even more enhanced. Anyway, I’ll download beta 3 and get it a whirl. Maybe getting the “bankruptcy” issue rolling and having the government apporval to “operate as normal” is allowing the programmers to focus more itently on the OS. It can make a difference when you know that you probibly won’t be unemployed next month.
My downloads are finally finished, I got the same md5sums.
“Let me tell you why. Mandrake’s financial problems. Heck, Mandrake Club was created precisely for that. Hardcore Mandrake fans feel that it is right for them to join the club.”
That may be part of the reason why it was created but they also knew that they need a way to support the cummunity at large and this was one way of doing it. As I say their RPMS and add-ons are very much appreciated in my opinion, as these provide me with what I need and want so that I won’t have to scour the net to look for stuff or waste time tinkering around with spec files to build a rpm for MDK.
“Before you call any argument “bullshit”, why not you compare the amount of downloads and the amount of subscriptions. Most of the subscribers are sympathy members, not members for the software (if you want that, there was always the boxes..).”
That doesn’t matter because they still showed a profit over last year and lowered thier expensive. You could easliy say the same of RedHat.
“Actually, if Mandrake actually develops it, these learning centers can actually be profitable. Now, just say Company A wants to move to Linux. They choose Mandrake Linux. So now Mandrakesoft has an opportunity to provide reeducation, which they get a rather inviting royalty.”
That’s stupidiest thing I have ever heard. That is the precise reason why they got in trouble. They screwed around and went in a differrent direction and by doing so they took away much needed resources from their main goal. Which is bring Linux to the desktop. They need to first focus on getting their distro ready enough for the masses and then once they captured enough customers they can dick around with other ideas if they have cash to burn. Right now their only goal should be to develop their distro and Mandrake customization tools along with support and services systems and gear them towards their orginal goal of a bringing in a Linux desktop.
“But why it isn’t making any money? Simply because there isn’t much companies even thinking of moving to Linux, much less to Mandrake Linux. Mandrake’s sales and marketing forces are pathetic in comparison with the likes of Red Hat and Sun.”
Hello !! Mandrake is making money ! Again it’s those stupid contracts that the previous CEO got them locked into which has put them in the red instead of in the black. They just need and will get back on track of focusing on the desktop.
I just installed Beta2 to discover the Beta3 is out. About 30 seconds to go now and I have Beta 3.
Anyway I like the speed and polish of this release. I also like that they don’t mess too muich with default KDE and Gnome.
If 9.1 is as good as it looks like it is going to be I’ll by the box in the mean time I think I’ll join the club. Well maybe.
looncroz’s statements parallel my own opinion. If you think they release too fast, then don’t download it–wait for the final.
“YTab’s situation has nothing to do with Mandrake’s. Don’t compare apples and oranges.”
His analogy has more sound reasoning than your posts Eugenia. And since when have you been on the Mandrake developer team Eugenia? I’m pretty sure they know what they are doing in releasing their betas and the reaction it will get on Bugzilla from experience (as opposed to your wild assumptions).
Ever since your last ‘review’ of Mandrake 9.0, I have realized how biased you are and seem to wish to only bash Mandrake every chance you get.
And to those saying the jump from 8.2 to 9.0 was poor and buggy–I think 9.0 to be an excellent release. I say this because this was the distribution and version (9.0 that is) that impressed me so much that it made me finally dump MS and move to the Linux world.
If the timespan of two weeks is to quick you can always skip a revision 🙂 It’s not that the new version depends
on the older one.
BeOS beta’s took months to appear.
When will wi-fi be included?
As long as I’ve been following them (about 3 1/2 years), they’ve kept a really short lag time between “betas”. This gives them the ability to get fixed software into the hands of testers earlier, so that the fixes can be tested.