Are we being used by Linux based Companies?

[BaVinic]

I don't know, but I get the feeling we are being used as pawns in this "Linux Chess Game" that is going on, First Red hat pulls out of the desktop market, and now Suse has been bought out and is also pulling out of teh desktop market in favor of the business ( and money) side of computers.

Are we being used as "Beta testers" and are we starting to see a trend here?

What is your thought??


BaVinic

[bwkaz]

There will always be Debian, LFS, and friends. Perhaps Gentoo as well, though I think I remember rumors of them doing same type of thing a few months back.

(There is no cabal....)

[dek0nstructi0n]

true. Linux can have a place with technophiles, can fit in education, small business, probably other niche markets. but to compete in "Enterprise" it needs a solid support foundation which I don't think is there.

Redhat has basically become another Unix vendor. They will be competing in an arena with companies who are the staples of techonology who have tried and proven and mature products.

don't worry evolution will sort this out soon enough. companies will come and go but linux will always survive.

[je_fro]

Originally posted by dek0nstructi0n
don't worry evolution will sort this out soon enough.....

Heh...I don't think the words "evolution" and "soon" belong in the same sentence....
But that's just me....

I don't think it can be helped. People are going to try to extract money any time they see an opportunity. Of course GNU/Linux is going to grow a couple of new faces as big companies (like IBM or Novell or TheUnspeakableOne) try to take "control" or "lead" the project. Eventually Linus will retire. All things change. Thankfully, there will always be Debian.

[El_Cu_Guy]

First Red hat pulls out of the desktop market,

Hmmm, didn't bother to do your research I see. Red Hat does stuff like this all the time. It's then followed by a press release about how people should stick to Windows for now because Linux isn't quite mature enough yet....

SuSE too. I think Novell will see this in time. If not then SuSE will die a horrible death like so many of Novell's acquired companies.

Are we being used as "Beta testers" and are we starting to see a trend here?

It's OSS. Essentially all users are "beta testers".

Redhat has basically become another Unix vendor. They will be competing in an arena with companies who are the staples of techonology who have tried and proven and mature products.

This is where Red Hat and SuSe shine and where they'll find the highest ROI. The desktop really isn't a viable solution right now for them. Especially, when some have the mentality that these companies should be giving everything away.

I myself have no problem paying for quality OSS (let's face it there's a lot of crap out there) and funding development. Sure Debian.org (or some other quality dev team), I'll be more than happy to click there to make a small donation.

don't worry evolution will sort this out soon enough. companies will come and go but linux will always survive.

Some people might think it's funny to watch some retard dev team die a slow painful death, but some "me-too" desktop distro "providers" should have been buried long ago.

[gehidore]

does it matter if they all pull out and start charging?/?????


there will ALLLLLL ways be " US " if "they" abondon us we can make our own distro. its really that simple there are enough of us most are just to lazy

[pwhitdog]

My only question is how do we support them and make the demand for the desktop user more of a reality? Several companies have bowed to the pressure as have been noted here, and though I have never tried them I would have to say that Lindows is at least fighting the fight along with Mandrake about making it a viable desktop solution. I think in time as the tides come in things will change, until then the best we can do is either use it, report bugs and be vocal on improvement ideas to the development teams, write documentation on what you get to work, and also we can try and learn languages to help create the programs to make this more of a reality.
Basically if we want this to work we all need to put in the time and make the effort for tommorrow.
my $.02 though, your results may vary,
Paul

[hard candy]

If a small number of users (compared to windows) and a lack of corporate support determined whether a OS died- why is BSD, gnu/HURD, etc still around?
Companies come and go, but the source is still there. And as long as there are ego's there will be someone building a new distro (the "I can do it better" motivation).
Sure, people get used to Suse and Redhat, but switching to another distro is no more difficult than switching from Win98 to Win XP (IMHO).
Just take a look at www.distrowatch.com , there are 100 distros listed.
When you find your favorite distro, send money, bug reports, and emails- that will help. And stop by their forums and answer a few questions.
Slackware, Debian, Fedora, OpenOffice, KDE, Gnome, Xfree86- these all have thousands of people involved and new ones are coming along each day.

[Satanic Atheist]

Am I correct in saying that Debian is the only non-commercially available Linux flavour out there? Something like that anyway... so that'll always be around.

I've noticed that RedHat have gotten their act together over the last couple of years or so and DO provide a comprehensive technical support system that cannot expand without higher revenue and income. On this note, I think they've done the right thing by going commercial. They want to provide a quality product and service to a bigger sector of the market.

The UNIX vendors don't seem to acknowledge Linux as a threat to them and I think it's pretty obvious why - If you need the power of a Sun SPARC Station would you get rid of SunOS (or Solaris) in favour of Linux? I'd probably stick with the native OS that was designed for that particular machine. At least with OSS you can compile the source code to run on a SPARC.

Good luck to RedHat, though. They provided us with a good, comprehensive distro that has now out-grown us. I hope they do well in the commercial market. We'll have Mandrake, Slackware, Debian and Gentoo for now.

James

[hard candy]

Am I correct in saying that Debian is the only non-commercially available Linux flavour out there? Something like that anyway... so that'll always be around

I tried to hold back, but I juuuussst haaaad tooo poooossssst!

No, you are incorrect.

Wheeeeeew! That feels better.

[Satanic Atheist]

Smart-arse.

Not sure where I read that, but I'm fairly sure it said it was non-commercial. Maybe you can elaborate on why I am incorrect?

James

[amgeex]

I really think that hard candy is wrong on this, Debian IS the only linux distro which is not commercial in any way (that I know, anyway), BUT I may be wrong, so correct if necessary.

[ricstr]

Am I correct in saying that Debian is the only non-commercially available Linux flavour out there?
What about slackware, gentoo, jamd are they availble comercialy ?

[je_fro]

I'll bet 'ol Dan Robbins wants to be a millionaire. Who else thinks that gentoo is ripe for getting bought by a big company who wants to play? I can't see IBM doing it, but someone else? HP? Nah...a company who lives a little closer to the edge....

[ricstr]

Originally posted by je_fro
a company who lives a little closer to the edge....
Microsoft ?

:D

[kam]

Oh my god, this is just flamebait.

[Satanic Atheist]

Originally posted by je_fro
a company who lives a little closer to the edge....
Pizza Hut?

James

[Rommani]

The majority of distros are non commersial.
btw, RH wasn't that good to begin with :D

[rbrimhall]

Hey Now... I like RH a lot... it's the only distro that has fit my needs (ie less technical than most but not too simplistic like Mandrake)... I'll use and support Fedora... I have not outgrown this distro... I'm just starting to feel at home...

[hard candy]

knoppix, fedora, slackware are considered non-commercial in the sense there is not a commercial version for servers/desktop such as Redhat, Suse, Lycoris have available.
Any distro can be used for a business/commercial network set-up but the difference is some distro's sell support services (what else are you going to sell when you have free software?). Take a look at Distrowatch- out of a 100 on the list, probably only 25 (estimate), if that many, are pushed commercially in some way (not counting selling coffee mugs, cd's , t-shirts, etc.) .

[je_fro]

Originally posted by kam
Oh my god, this is just flamebait. This should be deleted.

Saying that is kinda like me saying that you should keep your opinions to yourself.

[blimpieboy]

ok, just so you know where i stand, i am very new to linux. i have just installed lycoris and red hat 9 on my comp a mere 6 days ago. (took two tries to set up my dual boot on my own). ask yourself, why did you switch? is it because linux is "free"? or is it because windoze is buggy, unstable, and unsecure? were you tired of ms messenger taking over your computer? how about blue screens?

am i wrong to make the assumption that linux is only free in the sense that it is "open source"? if someone puts together a company to build a distro, isn't it for OUR convienience? and why would they make a company to not make money? that doesn't make any sense.

there are many companies out there still making and developing distros aimed at the desktop "market". i feel better giving my 30-80 bucks to them to keep them in the development business than to give one more dime to micor$oft.

if you feel abandoned by your favorite, all you have to do is switch. that's what happened when we left windows behind.

[Satanic Atheist]

What a profound first post!

Welcome to JL forums, blimpieboy!

James

[hard candy]

ask yourself, why did you switch? is it because linux is "free"? or is it because windoze is buggy, unstable, and unsecure? were you tired of ms messenger taking over your computer? how about blue screens?

Naw, by the time I had bought all the hardware, I only had $44.63 left to spend and since Win98 and Win XP were > $90, I bought Redhat 8 for $33.20 (with tax). :D

[amgeex]

ask yourself, why did you switch? is it because linux is "free"? or is it because windoze is buggy, unstable, and unsecure? were you tired of ms messenger taking over your computer? how about blue screens?

15 BSOD in a period of 10 minutes (win98Se), no bs.
And a month later widowze froze, rebooted, and didn't get past BIOS. The poor b$tch died, along with 234 of my precious mp3s. Damn you M$!!!

[El_Cu_Guy]

Who else thinks that gentoo is ripe for getting bought by a big company who wants to play? I can't see IBM doing it, but someone else? HP? Nah...a company who lives a little closer to the edge....

IBM won't buy up any distro provider. They're not distro providers which is one of the reasons they won't indemnify customers (stupid SCO). It is possible that SCO might opt for a desktop provider then choose to incorporate that into there commercial offering.

I personally would llke to see some big names buy up some of these guys.

1. The smart dev teams know that the desktop is about a user's experience and ease of use. It's not about simply gobbling together cool icons and a long list of included software that's utterly useless to Joe Blow end luser.

2. If mutliple companies were bought up it might end some of this "me-too, nothing more than a new name" distros. Dev teams would be forced to put aside some of their difference (some of which merely involve what desktop background to use by default, no I'm serious) and create a quality distro.

Certainly I don't think that every one of them should be bought up. Some of the drive and the skills but don't have the business sense to save their lives.

[The Linux Kid]

While it is sad seeing Suse and RedHat leaving the desktop arena, things must change. Obviously, RedHat and Suse have decided that they could put their weight behind servers and wait a bit to come out with desktop (I think the next desktop release to these two is goig to be BIG!

Until then, we still have Slackware, Debian, LFS etc etc etc etc. Even if those all went down, we would still have one of their last versions that we could continue dev with.

Just my $0.02

</rant>

[emus]

The follownig is a quote from the SuSE website



Novell/SUSE LINUX to become the world's largest supplier of desktop-to-server Linux solutions and technical support


http://www.suse.com/us/company/press/press_releases/archive03/novell_suse.html


I'm somewhat confused. Earlier it was stated that SuSE will pull out of the desktop market. Is SuSE just going to provide desktop solutions for a limited time still?

[ColeSlaw]

Personally, I don't think anybody has to worry about a company buying up Gentoo. If for no other reason, Gentoo is a source-based distro. I can't foresee sizeable company wanting to compile all their programs on their servers. Besides, when it comes down to it, Gentoo is pretty much just LFS + Portage. (I know there are some other differences, but Portage is the main one).

(Feel free to correct the following if I'm wrong, because my understanding of the GPL may not be totally correct)

What I understand is that even if Suse and Redhat don't want to make binary images of their distro available without paying for it, they still have to let anybody who wants look at their source code. This means that any money they make will only help them improve their source code (They have their own R&D department, correct?). This source code then is viewed by other Linux Developers and used in other distros. BLAMMO! We still get the benefit, even if we aren't running their particular distro.

I'm not worried about being "beta-testers", etc. There will always be stable versions to run, even if they are a little dated.

Once again, feel free to correct the above if I have misinterpreted the GPL.

[blimpieboy]

oh, one other thing, didn't i read somewhere that red hat is simply changing the name of their next release to fedora, which is essentially red hat 10? and stamping the red hat name to their enterprise products?

[BaVinic]

this is what I get for staying away from the computer too long :D.....

My point was , are we being used as guinee pigs? did they offer to give us something ( not as in for free) and then once we helped make it (perfect?) they pulled the rug out..

I have purchased every version of Red hat I have used, starting with version 6.0 and every point there after. I have supported red hat with bug reports, documentation errors . and in beta testing. and I would have continued to do so, but I feel like I have been used, and it is not just me, there are at least 30 other people who did much more than I could ever have done, who feel the same way.

I am happy for Red hat, I wish them no ill will. and once I am able to do so, i will get Fedora and check it out, but as far as I know, it is not RH10, there is no RH10. Fedora from what I have heard is not even a product of RH, but of another team who shares the same ideals that RH did, RH owns the copyright of the name (Fedora) and there will be some interaction with the RH devl. team.

I don't know. or even care what RH does, I like RH, I have enjoyed using it, but like has been said many times in this thread, I will just move on to something else.

:D

BaVinic

[Citadel]

Maybe the Linux vendors should work with a software layer which would provide their customers with productivity advantages and that way the vendor can decouple through abstraction Linux the product from Linux the platform shared by the open source community.

[Lostman]

Originally posted by BaVinic
I don't know, but I get the feeling we are being used as pawns in this "Linux Chess Game" that is going on, First Red hat pulls out of the desktop market, and now Suse has been bought out and is also pulling out of teh desktop market in favor of the business ( and money) side of computers.

Are we being used as "Beta testers" and are we starting to see a trend here?

What is your thought??


BaVinic

Linux is free. It was around before a lot of these Distro companies. That's the best part about it, not one person owns the code. So SuSu can fall off the face of the planet and Red Hat can decide they want to make cars insted, Linux will still move on.

M$ users are beta testers. But they pay for that right. They pay for they're bugs and have to wait for one company to fix it. In Linux, you don't. Everybody collectivly works to fix the bugs.

[dek0nstructi0n]

Originally posted by Lostman
M$ users are beta testers. But they pay for that right. They pay for they're bugs and have to wait for one company to fix it. In Linux, you don't. Everybody collectivly works to fix the bugs.

In theory yes... In practice, Linux is no better. :) This is what has created the market for "enterprise" linux. Eg. say I'm a commercial software developer with a product that runs on both linux and windows. Say a big bank or something uses my application in a production environment. Then my application goes down, stops production, bank customers are pissed etc and I have to investigate. I find the bug is really not with my program but with the OS...

With Microsoft or Unix vendors I can get an escalation manager with 24x7 pager, daily status calls and assurance the problem is being worked on.

With Linux I can beg some volunteer developers on a mailing list and pray someone does something before I lose my customer.

Red Hat and Suse want to play in the same arena as MS, and Unix companies where they have a stable support life for their products (like 5 years or so) issue officially tested patches etc. They also don't need to wait for Linus's blessings to get critical patches out to their customers. Some of these patches will trickle into the official kernel branch.

[bwkaz]

Originally posted by dek0nstructi0n
In theory yes... In practice, Linux is no better. Wrong, sorry.

I have fixed somewhere in the neighborhood of ten bugs in my operating system. The latest one was in sysvinit -- if you started a sulogin shell in single-user mode, but you had the /etc/inittab line for single-user mode looking like:

su:S1:wait:/sbin/sulogin instead of like:

su:S:wait:/sbin/sulogin which is how the developer assumed everyone did their inittab, then you would end up getting both your sulogin program and an extra /bin/sh process. Both reading the terminal at the same time.

This plays hell with trying to type to only one of those processes...

Try fixing a problem like that with Windows. And good luck.

Most of the rest of the bugs were compile bugs (me using gcc 3.3 when people hadn't updated their code to remove multiline string literals like they should have been doing), but they were bugs just the same.

I fixed all of these less than ten minutes after opening up the first source file. I'd like to see you get a patch from Microsoft within ten minutes...

------

The problem with your scenario is that with Linux, you don't wait for someone else to fix the bug. You fix it yourself, and provide your customers with the patch to their sources.

With Microsoft or Unix vendors I can get an escalation manager with 24x7 pager, daily status calls and assurance the problem is being worked on. [/b] Umm... that doesn't necessarily mean much. Congratulations, somebody's working on the problem. So? You have no more of a guarantee that they will fix it "in time" than you had that your Linux vendor would fix it "in time". If you want something done right, do it yourself.

[dek0nstructi0n]

We're getting a bit off topic here. I only wanted to point out to the previous poster that in reality you're not always going to get an army of kernel hackers fixing all your bugs.

But I'll play devil's advocate...

Originally posted by bwkaz
Wrong, sorry.

I have fixed somewhere in the neighborhood of ten bugs in my operating system.

This plays hell with trying to type to only one of those processes...

Try fixing a problem like that with Windows. And good luck.


That's totally cool! But not everyone has mad kernel skills to fix Linux problems on their own. You think you could have done the same with say a stack corruption where kernel memory was being corrupt somewhere completey unrelated from the area of crash? Or some hairy intermittent race condition in the kernel causing a semaphore count to become inconsistent? And people looking over your shoulder waiting for a fix even though it is not your code?


Most of the rest of the bugs were compile bugs (me using gcc 3.3 when people hadn't updated their code to remove multiline string literals like they should have been doing), but they were bugs just the same.

I fixed all of these less than ten minutes after opening up the first source file. I'd like to see you get a patch from Microsoft within ten minutes..


Not really relevant to this discussion since those aren't the "bugs" I'm talking about. Even more so since MS users don't typically need to compile their programs.


The problem with your scenario is that with Linux, you don't wait for someone else to fix the bug. You fix it yourself, and provide your customers with the patch to their sources.


Riiight... My company wants to spend their resources having me fix bugs in code we don't develop. Perhaps my company will only support "enterprise" linux distros with support contracts then.


Umm... that doesn't necessarily mean much. Congratulations, somebody's working on the problem. So? You have no more of a guarantee that they will fix it "in time" than you had that your Linux vendor would fix it "in time". If you want something done right, do it yourself.

You're right. Maybe said OS vendor won't fix the problem in time to avoid a disaster in the situation I'm talking about. Luckily me or my company is no longer accountable for that problem as all eyes are now on the OS vendor instead. We may still lose the account but at least we won't get sued (successfully).

I also have some peace of mind knowing that someone somewhere under direct orders of management is working to fix the problem because it is their job to do so and it is in their best interest to do well in their job. Rather than say some week-end volunteer hacker who may look at the problem when they get back from snowboarding and if their favourite TV show isn't on later, etc. Motivation plays a role here.

I'm not talking about the Redhat and Suse engineers who would count as being part of an OS vendor. If I tried to "fix it myself" by modifying their kernel code their support contracts would be cancelled. Doesn't look like having source code access in this case helps much but that's a different story ...

[bwkaz]

This is about as long as one of El_Cu_Guy's posts...

Originally posted by dek0nstructi0n
But not everyone has mad kernel skills It wasn't even a kernel problem, though... sysvinit is userspace.

to fix Linux problems on their own. It's not hard to figure it out. The source code is right there, after all. I'm a firm believer that anyone can discipline themselves enough to learn to write code. Most people don't have the required interest in it, but that's irrelevant.

You think you could have done the same with say a stack corruption where kernel memory was being corrupt somewhere completey unrelated from the area of crash? Not that hard; UML with gdb. Also take a look in the kernel debugging menu sometime, at the "check for stack overflows", "debug memory allocations", "mmio debugging", and "spinlock debugging" options. ;)

Or some hairy intermittent race condition in the kernel causing a semaphore count to become inconsistent? That wouldn't even happen though. There are several extremely well-known ways to implement semaphores (most of them based on mutexes, IIRC, which are usually based on loops and atomic increment-and-test or decrement-and-test machine instructions (spinlocks)), so it's not like there can even be bugs in the semaphore functions. Unless the function itself doesn't implement the well-known algorithm properly, which is easy enough to see.

Plus, I wouldn't be the only one looking for the problem. "Given enough eyes, all bugs are shallow" (ESR) -- I would simply be another pair of eyes. Report a true kernel bug to the lkml list, and you'll have hundreds of people looking into it. Of course, this means you have to do your homework first (make sure that it is actually a bug in the kernel), then figure out how to reproduce it, but it's not like there's only one person doing bug-hunting on lkml.

And people looking over your shoulder waiting for a fix even though it is not your code? That's no big deal. I get it happening all the time at work. But again, even if I can't find it, someone can.

You're confusing the bug-hunting process of large open-source projects with the bug-hunting process of proprietary software. With a large enough piece of OSS, you have hundreds (if not thousands) of people looking at the code. You are simply not depending on one "weekend hacker"; you're depending on ten thousand of them. Which means that on average, a thousand or so of them will be available, not one.

And the support departments of most of the companies I've talked to have been abysmal. I don't doubt that this is due to the perception (at the management level of that company) that support is secondary to making money off the "good" that they sell. Too bad the software industry parallels a service industry much more closely than any goods industry... *shrug*

Riiight... My company wants to spend their resources having me fix bugs in code we don't develop. So what are you doing in the meantime if your vendor is working on the problem? Constantly bugging them via that 24x7 pager? If you're doing something else, then I'd say this "problem" isn't quite as big a deal as it first sounded.

Why not spend some time trying to help the people that are helping you? In the case of an open-source OS, you can be another pair of eyes looking for the bug. If the source is closed, you are sitting there with your hands tied. That is the basic difference.

I also have some peace of mind knowing that someone somewhere under direct orders of management is working to fix the problem because it is their job to do so and it is in their best interest to do well in their job. And yet, when people write code because "their job depends on it", rather than because "they like to write code", the code is almost inevitably worse... not to mention the benefit of having more than one person (or more than a small group of people) looking at this code.

Rather than say some week-end volunteer hacker who may look at the problem when they get back from snowboarding and if their favourite TV show isn't on later, etc. Motivation plays a role here. Mere job security is a historically poor motivator...

But that's not even the entire point. You're still not waiting for "some weekend volunteer hacker", you're waiting for any one of ten thousand of them.

[kam]

Originally posted by dek0nstructi0n
With Linux I can beg some volunteer developers on a mailing list and pray someone does something before I lose my customer. You could also *gasp* hire sombody to do it.

[dek0nstructi0n]

That wouldn't even happen though. There are several extremely well-known ways to implement semaphores (most of them based on mutexes, IIRC, which are usually based on loops and atomic increment-and-test or decrement-and-test machine instructions (spinlocks)), so it's not like there can even be bugs in the semaphore functions. Unless the function itself doesn't implement the well-known algorithm properly, which is easy enough to see.

Well not exactly the bug I was talking about, but pretty close, i.e. hung sempahore because of a race condition due to lack of an internal lock. see: http://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=105910

What you say is true, but you might be surprised how complex some system calls or libc functions (something as braindead as strcpy() even!) can be implemented on some platforms.

So what are you doing in the meantime if your vendor is working on the problem? Constantly bugging them via that 24x7 pager? If you're doing something else, then I'd say this "problem" isn't quite as big a deal as it first sounded.

My scenario basically puts me in a large software company like an SAP, BEA, IBM, Oracle, etc. The customer problem isn't ciritcal to me (i.e. not my downtime) but it is to my customer because my commercial software product fails to work because of something in the OS. (you can blame poor QA but hell even an OS patch after the fact or something could have exposed the problem).

No, I can be spending my company's time finishing other projects or supporting other customers who found bugs in my code. Critical attention to the original customer can be dealt with the OS vendor company once they acknowledge the problem.

bwkaz, you make good points about motivation and job security. You are a worthy person to debate with and I'm enjoying this! :)

You could also *gasp* hire sombody to do it.

Nice smartass comment, too bad you miss the point. What part of "not wanting to spend my company's resources on a problem that is not my fault" didn't you understand?

My point is there is no good service structure for linux itself. Red Hat and Suse have come along to fill that void for enterprise/mission critical needs. And if you don't believe me take a look at any major commercial enterprise app that runs on Linux and see if that company supports you running it on Slackware or Debian or Gentoo... or even anything else other than Red Hat or Suse, or even Turbo. I'm not talking your basement run ISP running sendmail and apache on Debian here.

go ahead, look up SAP, Peoplesoft, BEA, IBM, Oracle, JD Edwards, Siebel, SAS...