Slashdot Mirror
An Open Letter From Bob Young
Subject: Freedom & personal responsibility good, serfdom & tyrannical control bad.
From: bob@redhat.com
Dear Slashdotter,
The wild and heated debate about Red Hat 7 in recent days has been interesting to follow. It demonstrates the strength of the open source model. By comparison (I'm not sure if anyone noticed this) Computerworld had a front page story a couple of weeks ago about how there were problems with Solaris on Sun's Enterprise systems, but that these bugs were not well known because Sun was making their customers sign NDA's (non-disclosure agreements) before helping them fix the problem.
Consider the contrast between a proprietary vendor's unwillingness to debate the merits of their technology with the open debate that Red Hat Linux enjoys.
This discussion is of such value to the users of Red Hat products that we feel little need to even attempt to comment. Informed readers can read all sides of the debate, download our products, test them, and decide for themselves whether our critics or supporters are correct. Of course the readers who post things like "well I haven't tried RH7 but I've heard..." aren't very helpful, but I trust most Slashdot readers to see through that kind of stuff.
There is one recurring comment that I could not resist addressing. Namely the regular habit of our critics of comparing Red Hat to Microsoft. I just don't get it.
There are many things for which we should be justifiably criticised (I have no idea what these might be, but I'm certain they exist ;-) but trying to act like Microsoft is not one of them. Red Hat's business is built on solving the problem thatMicrosoft's business model has imposed on the software user since Bill Gates disagreed with the members of the Homebrew computing club back in 1980.
The software industry that Microsoft has been the role model for is built on the premise that customers are not to be trusted with the technology that they are building their organizations on. The legacy software industry is built on the proprietary binary-only model where not only does the user not get the source code he needs to make changes, but worse he receives the product under a license that essentially says that if you make any improvements to the technology you are using, if you solve a bug that is causing your systems to crash, or add a feature that your users or customers desperately need the vendor can have you thrown in jail. (If you don't believe me, just read any shrinkwrapped software license). This kind of business model, where the customer is completely beholden to his supplier exists in no other industry in any free market that I know of. It harks back to the old feudal systems of 12th century Europe.
Red Hat's business success is owed to one simple benefit our products and services offer that our larger binary-only OS competitors do not. Namely that our commitment to publish the code that we write and distribute under open source licenses enable us to give our customers control over the technology they are using to build their systems. We cannot promise to deliver perfection. All we can promise is to acknowledge the problems immediately and work with you to fix them publicly and in real time. With control over their systems our users can simply build more stable and reliable systems than the binary-only model allows.
This is why the fear that Red Hat is somehow going to wake up one morning and abandon our commitment to open source is so mis-placed. Open source provides us with -the- competitive advantage that enables us to compete effectively against much larger competitors. To abandon open source is simply not in our customers interest and hence not in Red Hat's financial interest.
So if you want to criticise us for shipping gcc 2.96, you have every right to do so - you'd be wrong, but it is at least a legitimate debate and I'd respect your opinion. But to compare Red Hat to Microsoft indicates an ignorance of what is driving our success.
Remember that this debate was begun by someone going to Red Hat's public site and trying to add up all the registered bugs in Red Hat 7. When was the last time Microsoft (or any other legacy software vendor for that matter) gave you access to their complete bug registration system? Which software model do you really want to see succeed? One where you have to trust your vendor (who can and frequently restrict access to information you does need) or one where you are in control of the technology you are using?
We may be making mistakes - that up to you to decide. Some of them may be important to you and while I have no doubt you will point them out to us, you have control over the technology you are using. We work hard to build products that please most of our users most of the time. But if you don't like something about Red Hat Linux you don't have to use that feature or function. We simply are not pursing a business model that bears any resemblance to Microsoft's, so just quit it.
The next slashdotter who compares anything Red Hat does to Microsoft will be punished. The punishment will be to find the nearest blackboard and write "freedom & personal responsibility good, serfdom & tyrannical control bad" seven hundred times.
Cheers, Bob.
it is exactly this kind of thing (open letter) that makes causes my jaw to drop when i see the MS/RH comparisons. when was the last time such a letter came from redmond? i don't recall ever, EVER seeing an article ANYWHERE pointing me to a letter from bill and friends saying, "we're sorry that some of our customers have trouble with our software. we honestly thought we were improving something, but if you don't like what we've done, you are certainly entitled to disagree with those decisions, that's what makes the next version better." have i just missed that letter, or am i right in assuming that such doesn't exist (obviously, i already know the answer). thank you mr. young for the response, thank you for paying attention to us (all of us, not just the large companies, but also the poor students), and thank you for your belief in the OS model. i've used RH in the past, and i've been trying to decide what to put on my old box. looks like RH wins, if for no other reason than you guys at least take the time to explain decisions to your users. i don't expect perfection, but i like it when the developers can admit that they don't offer it anyway. now, where's that RH7 iso, maybe if i like it, 7.1 will get an order and a new user.
The problem is that what redhat needed was to delay releasing 7.0 and instead come out with a 6.3 releast that fixed all the bugs and numerous security holes of 6.2.
Installing Redhat 6.2 is a very burdensome process because of the number of fixes required. However, the consequences of RedHat 7 are much more severe because of its incompatible version of gcc.
What was such a big deal about coming out with a 6.3 release, which would have been stable and useable, instead of just quickly pushing RedHat 7 out the door?
Well, given that this is on your desktop, I bet you don't see the names that way under your DesTop Icons. In fact, you won't see them that way in your Windows Explorer, unless, of course, you have that infamous check box checked.
The point made, methinks, was that the names you see were not the actual names of the file. WYSINWYG.
You may not believe you should be held to the standards that Microsoft is held. I agree with you: I think you should be held to a higher standard still.
I also agree with you that knowing exactly how bad the bugs are is a blessing. It IS good for us to know if there are problems with a product. And the fact that we have full control of the technology is also a Good Thing; it's the biggest reason to use a distribution like Red Hat.
Whether you like it or not, however, you are in the business of providing many of the same things that Microsoft provides. The software you ship must work. RedHat 7 dies after three weeks. If Microsoft were to ship code like that they would be drawn and quartered in the press. It would literally cost them billions, both directly in stock price, and indirectly in reputation damage, were they to ship a product in that state.
I don't think it's unreasonable to expect Redhat to not only do as well, but to do BETTER than Microsoft does. You have a whole army of people working for you for free that Microsoft does not. Isn't it your job to make sure that the bugs are ironed out? You're starting with some of the most stable code running anywhere, and assembling that code into a system for general consumption. If QA isn't Job 1 for RedHat, what is?
You have frequently compared RedHat to Heinz ketchup. People can make their own ketchup, but Heinz still sells an awful lot of the stuff, because the consumer has come to expect a consistent, quality product. Well, you have just shipped millions of bottles of ketchup that leak. Admittedly, it's good that we know about the problem, but the bottles still leak. What do you think this would do to Heinz' reputation?
Personally, I am firmly in the Eric Raymond corner of the Free Software world -- what I want is software that doesn't suck. I'm willing to pay for it if I need to. I am a professional system administrator. I need to build systems that run and run and never fall over. The control over the technology is important, but it is a secondary goal for me; having access to the source is good, but FIRST it must work. SECOND the source must be open.
The fact that you open source everything is not an excuse to ship bad software. Open Source garbage is still garbage.
That was 98, and 98 isn't exactly known for reliability. You occasionally hear about machines lasting that long, but it's almost unheard of for systems that are actually doing anything. Yes it was a bug, but it was a bug that didn't matter in practical use for 99.9% of the customer base.
A problem is, however, a problem if people notice it. Linux crashing after three weeks *matters*.
RH7 is, for lack of a better description, a stinky pile of horse doo-doo.
heh
The inclusion of heavily patched, alpha, non-standard components in RH7 was, in my humble opinion, a tragic mistake.
OK, please tell me which parts of RH 7.0 are alpha quality , heavily patched. I'd especially like your insight on which parts are non-standard.
Hard-core linux users have been turning away from RH for years because in an effort to make your products easier to use, you have strayed far from the path that is Linux's purpose... stability and accessability.
Yes, we all know that people are swarming to Slackware because it is bare, raw linux. Sure a distribution with no configuration tools which forces the user to dig in via the console is far more likely to seem accessible to users.
Making your products accessable to Joe every day windows user by making it easier to use is great, but serves little use when it is done at the expense of those that brought your company to where it is today, the hard-core hackers.
How are these changes at the expense of hard-core hackers? I have never come upon a task or function that I couldn't make ANY linux distribution do for me. I think you're just being willful, or you are not one of the hackers you mention --in which case, why champion someone's else's gripes?
You claim that you publish the source that your company creates; where is it? I know it's on RH's site, but finding a patch on your site is next to impossible, even for the advanced user.
Yes, even hard-core hackers have trouble finding patches, especially on a site as complex as Red Hat's.
As I write this, I'm downloading the latest revision of Debian to install on the remainder of my redhat servers. Until RH releases another product that conforms more to the spirit of open source, I cannot use its shoddy distributions.
Why don't you wait until the latest revision of ANY distribution has undergone some tire-kicking by experimenters, and refining by the developers before you go placing it on your production servers. Or why don't you just HACK a distribution together yourself, seeing as hard-core hackers like yourself are what basically gave RedHat what they have today, right? As a disclaimer, I don't use redhat. I did for a while back in the 2.x --> 4.x days. I tried all of the distributions available at the time and settled on Slackware. That is my personal preference. Still, I won't let someone spout uninformed garbage when I see it, regardless of who it's aimed at. Moderate down as needed. ;)
The Simpsons are as mass market as it gets. They do feature a large number of "inside" jokes, and obscure references, however. I think you mistake these for esotericism.
...
However, Groening's first major accomplishment, Life in Hell
"Akbar and Jeff, brothers, or lovers, or possibly both, whatever offends you the most."
How's my programming? Call 1-800-DEV-NULL
I wish I could make it to ALS to grab just 5 minutes of your time, but alas I took a teaching job teaching intro to Unix starting this week.
Deus Ex Machina said,
Again, as a Debian user, I feel that there is a hell of a lot WRONG with RedHat's distro - for ME. This more of a personal taste though - and don't anyone dare tell me that Debian hasn't had it's share of crippling bugs....
So I think we should back off and turn down our flamethrowers here - if you want to criticize RedHat, fine, but at least find a good reason.
I too am a Debian and a Redhat user. I like the elegance and commitment to structure that is Debian, but don't be fooled that really doesn't matter to most people. They want it to just go.
Please build or accept somthing like apt-get that doesn't depend on unique ID numbers or "Priority Access." My gut tells me if you don't, one day Linux==Debian instead of Redhat in most peoples minds. What you lose in CD sales you will gain in acceptance of Linux in general.
Please don't lose your dominance. I appreciate your contribution, and I fear the morality of your replacements. Thanks for listening.
Matthew Newhall President of LILUG
Novel theory: Modern Man evolved from psychopath
--
I've been using RedHat since 3.0, and have even spent time rolling my own RedHat-based distro. I also write code and release it as open source. So I think I know what I'm talking about, when I say that I have yet to see RedHat do anything that is against the spirit of the "open source movement". Go pick on SuSE or Caldera...
Unlike you, I'd like to applaud RedHat for some of those "pointless changes" that they made - I love it that there is SSL support in things like Samba and that they finally dropped inetd for the vastly superior xinetd. These are genuine steps forward - they may increase the learning curve for upgrading a bit, but they have good reasons for doing these things. I'm not sure I'm happy with the gcc thing, but I'm not going to let it ruin my opinion of this excellent company. So what if some of the new stuff has bugs? They'll get fixed, just like they have in the past.
Unfortunately, no Linux distro has jails... so I'm installing FreeBSD these days. :-)
--
Host your own websites, anywhere!
The other problem that you mention, exit() being broken, is /not/ a compiler issue. That's an issue with glibc. However, there is an update to glibc available -- 7.0 shipped with 2.1.92-(?), there's a 2.1.94-3 available on your local mirror.
-30-
Although I would argue that Red Hat at least didn't beta test the distro long enough, given the fact that the update daemon can flatline the whole OS in a matter of three weeks, I'd say that the number of bugs in the distro as a whole isn't that bad, in the scheme of things. My bugzilla search turns up 790 bugs as of right now. Of these, 203 are marked as duplicates, leaving no more than 587 non-duplicated bug reports. It's quite likely that a number of the new and/or non-closed reports are duplicates as well, I'd hazard that about 550 are non-dupes. 302 are marked as "Not A Bug", although a number of these are dupes as well, or are something like "gcc 2.96 is unacceptable for compiling" (I don't like the idea of a snapshot compiler being included in a release product, but it looks like Red Hat is committed to supporting what they're calling 2.96 on their own, and stabilizing it as much as possible. Furthermore, I hardly think that the deliberate decision on the part of Red Hat to include this snapshot qualifies as a bug, no matter how bad a decision one might think it was). In any case, I'd say we're looking at about 400-450 verifiable bugs.
Now, consider this: Should Red Hat be responsible for fixing bugs in software they did not develop?
In one sense, I don't think they should be responsible for fixing bugs in software they aren't developing. It would certainly be nice if they submitted patches fixing such bugs, but it's simply not economical on their part to do so. Red Hat is in business, and their aim to make money. Red Hat *IS* kind enough, however, to contribute to the development of certain projects like the kernel, as well as GNOME.
Furthermore, a Linux distribution is in theory a compilation of software the ultimately makes the Linux kernel useful -- and I wouldn't task the publishers of compilation discs with fixing bugs in software on that disc. (Of course, Red Hat *SHOULD* be responsible for bugs in their installation software, etc.
OTOH, Red Hat *should* check thoroughly to make sure that the packages they do include in their distro are stable and relatively bug-free (since it's probably well-nigh impossible to produce COMPLETELY bug-free software).
I would compare this to Linus and co. releasing a new major rev. kernel, it takes a long time, and a lot of eyes.
Actually, it's a whole order of magnitude past the release of a kernel, since we're dealing with hundreds of seperate software packages relying of all sorts of different libraries, written by a number of different authors . . .
What can we do to help this? It's not like people are not reporting bugs in the beta, it's more like there's not enough time to address the issues. Or am I all wet here?
Well, they could do like Debian, and test for years before releasing . . . :-) Of course, you're not likely to be the first on the market with the newest and hotest stuff like that, though.
DISCLAIMER: I'm a Red Hat user. I do, however, know how to compile my own software. (Yup, I'm one of the three!)
----
"A fool does not delight in understanding, but only in revealing his own mind."
Yup, that's much better than "Hey, we're the standard, we're not gonna give you the code or the specs, just buy our stuff. And hey, if anybody else figures out how to do it without us, we'll just change the standard again . . .", ain't it?
----
"A fool does not delight in understanding, but only in revealing his own mind."
As far as waiting for KDE 2.0, why bother? No distribution maker is under any obligation to delay their distro for the benefit of an app vendor. I don't see SuSe, Debian or Corel doing this either. Really the best thing to do - imho - is ship the latest KDE stable, offer the development snapshot as an option and provide the 2.0 stable when available as a download. In fact doing this with gcc might have averted some of the criticism.
Welp, I didn't really mean to imply that they should wait for KDE 2.0 -- more that rather than releasing a 6.3 as a minor update (since 6.2 seems pretty solid in my experiece -- we've got it running on 3 boxen here), release 6.5 as a more major update (with newer release software) while discarding the attempt at making the transition to kernel 2.4 easier (which seems to be the major driving force behind releasing 7.0 w/ "gcc 2.96" and the like).
In any case, I do think you're absolutely right about releasing a stable gcc and offering the next revision as a download . . .
----
"A fool does not delight in understanding, but only in revealing his own mind."
That product will then only work on either Red Hat 7.0 or another platform that chooses to discard the views of the GCC steering committee.
However, the gcc folk only released their statement after the release of RH 7.0, so it's not like Red Hat decided to include what they call GCC 2.96 in spite of objections by the gcc steering committee. On the other hand, I think that the inclusion of a snapshot of the compiler in a distro is at best not wise.
That being said, I do understand Red Hat's motives for doing so. They're caught in a difficult position -- marketing wisdom drives them to try to be the most cutting edge distro (after all, having the newest features first looks good to PHBs), while on the other hand it seems like a much safer move to ship a currently stable compiler (yes, I understand that there are questions about the state of 2.95.5 or whatever it is . . .).
Personally, I think I would have chosen to keep the stable release of gcc in, wait until KDE 2.0 was out, add in XFree4.0 and the latest stable GNOME, upgrade some other useful packages, and call the whole thing Red Hat 6.5. Then, I'd sit tight and wait until gcc 3.0 and kernel 2.4 were on the immediate horizon, and then start developing what would become RH7.0.
----
"A fool does not delight in understanding, but only in revealing his own mind."
We work hard to build products that please most of our users most of the time.
.02
it isn't like MS even gives a fuck about pleasing ANYONE. They kind of just force the software on you and make you like it (b/c there is really no alternative). I want an honest show of hands about who likes that fucking little guy in the corner of Office products that is there to "help" you. When you first started up Win95 did you really like the Start menu or the way that the screen looked? I know I didn't... I hate having the clutter of icons all over my desktop. I don't even like having the taskbar "hidden" at the bottom of the screen.
At least RH is giving us many options to make our systems the way we want them... Until they start mandating a single WM and a single way to setup partitions and programs, I say that they are superior to the MS model.
You can flame me all you like for being pro RH, but for someone who is more interested in having his system work w/o much effort (yes, I know how to setup and use other distributions) RH is good for me. I want to burn through the install w/o having to download this and that to make my system complete. RH is providing a service, they are making Linux easier to use for more people, if you aren't into that, fine... Go off and use Slackware or Debian, or whatever, but don't you dare complain about RH..
Just my worthless
- Bill
Have you ever read The Cathedral and the Bazaar? The whole point of Open Source is to, "Release early. Release often." That's what makes Open Source a colaborative development environment. Linus didn't wait until he had a complete, SMP enabled kernel with full support for every conceivable device. He got something working, pushed it out on the Internet and let other people start hacking at it!
Now, obviously, the situation with an integrated commercial product is different from a single developer putting out pre-alpha code for others to start playing with. Commercial distributions have people depending on them for stability, But to say that Open Source historically waits until something is "done" before releasing it is to be completely ignorant of the Open Source development model. This is a "point zero" release, and as much as I diskile parroting the mantra that x.0 releases are never stable, I think that RH7.0 is another example of that rule. If you don't want to be on the bleeding edge, then wait until the bugs have been ironed out. If you want all the latest new features, then learn to live with the risks.
(BTW, I've been using 2.3/2.4 kernels for months. They don't always work on the first try, but they have the new features I need, and I've gotten used to living with the risk. I haven't had any catastrophic failures yet ((knock on wood)), but I make sure I have good backups...)
--
Your Servant, B. Baggins
The reason that you can't work out who said it, is because it's a misquote. The well-known sentence is actually "You can fool some of the people some of the time, but you can't fool all of the people all of the time." I have no idea where you got the "please" part of it from. I think that the original quote comes from P.T. Barnum, but I couldn't be sure. A slightly drunken Charles Miller
--
The more I learn about the Internet, the more amazed I am that it works at all.
So if you want to criticise us for shipping gcc 2.96, you have every right to do so - you'd be wrong Whaadaya mean?! Half of what I'm used to building, won't - if this was a rawhide release, then fine, but 7.0? And has anyone seen any errata pop up for their system on the Red Hat Network? I'd think we'd see them by now....
Yes, the gcc was included in the beta. This doesn't change my criticism. Red Hat (and other major distributors) should discuss plans to include snapshot releases with upstream maintainers, not to give them a veto, but so that any problems caused thereby can be worked out in advance.
> Um, what if the standards set by the dominant distro happen to be retarded?
Then fix them. Christ, do you people refuse to step up to the plate because you can't guarantee you'll hit a home run? Redhat isn't telling anyone to go away, but they aren't lying down to give everyone a chance at being great. So they push their product like it's all that matters to them -- because it is.
What do you suggest Redhat do? Advertise Debian?
I've finally had it: until slashdot gets article moderation, I am not coming back.
> I can't remember the last time anyone said 'Slackware? That buggy piece of crap? Forget it, I'm going to Redhat.'
I have. You can quote me on it. I don't consider it buggy per se, I just found it even more cumbersome to work with than Redhat. Then I found FreeBSD, which I consider to be the best of both worlds. Haven't looked back.
Guess what, I have different needs than you, and others may have different needs than me. Deal.
I've finally had it: until slashdot gets article moderation, I am not coming back.
It's called innovation.
Yes, I wouldn't be at all surprised if Microsoft considers it an "innovation" that they changed what letter signifies a standard option in a standard utility.
-
Well, it IS true that some people have compared Red Hat to Micro$oft lately. Some because they seem to have a paranoid fear of anything commercial, be it open-source or not. But what has been the most voiced opinion lately, is not that Red Hat's goal, intentions or business methods are Micro$oftish, but rather that Red Hat's .0 releases, and especially the latest, RH-Linux 7.0, has a lot of Redmondish features.
Which, one might wonder?! Well, for one, the .0 releases seem to have been rushed out before going through enough beating. It is indeed impossible to fix every bug, but if the bugs hit the everyday users rather than "just" the geeks, maybe something is wrong with the quality assurance process?
Another thing is the handling of the gcc 2.96beta issue. There is no gcc 2.96, despite what the name of the package in Red Hat Linux 7 might be. There was a developmental branch of gcc named 2.96 which hadn't been blessed as finished yet. This branch is now renamed to 2.97 to avoid misunderstandings. The reason for releasing RH7 with the gcc 2.96 beta branch rather than gcc 2.95.2 is, unless I've misunderstood things completely, because of the flakey C++ support in the latter one. This I can understand, support and agree with. What I don't support, however, is naming this compiler anything else than gcc2.96cvs or gcc2.96beta, something it is.
To quote directly from gcc.gnu.org:
The bottom line is, that the handling of gcc in RH Linux 7 was clumsy. And rather than just saying that "You're wrong", Red Hat should admit that something didn't come out right and at least apologise to the community. Do not give those who believe that there's a secret agenda behind the acquiring of Cygnus get even more reasons to be paranoid. They're far enough from the reality already.
Like it or not, Red Hat has become synonymous with Linux in large parts of the non-initiated computerworld. And therefore it falls on the shoulders of Red Hat to do their utmost to make their product better than any other Linux distribution. Because the verdict of Linux from the "real world" won't be based on the stability and security of Debian Linux or the cutting-edge features of SuSE. It will almost entirely be based on how bugfree, secure, easy to use and easy to install Red Hat is.
Maybe with time, this will change, but until then, Red Hat must never, ever race to a release or include beta-software that isn't clearly labeled so. And indeed, this should be in Red Hat's own interest more than anything else, right?
The compiler not being able to build the kernel is the kernel's fault, not the compiler's.
If I'm not mistaken, gcc 2.95.2 can't even compile the kernel.
Need a Python, C++, Unix, Linux develop
I did attack your claim to womanhood through the nick of Anne Marie.
Still, your points in this forum a quite valid, and I welcome the diversity a man posing as a woman can bring.
Got any pics?
I second that. Except Anne Marie is a man. Read the posting history.
I can totally appreciate the desire he has to explain the situation his company is purportedly contributing to, but realistically, all he acknowledges is the comparison of RedHat to Microsoft.
The real things he needs to be doing is explaining to us WHY they included gcc 2.96 rather than a stable, proven version?
WHY are they encouraging the concepts of propriety in the software that they're distributing?
WHY are they regularly taking software that the open-source community and environment have strived so hard for and creating situations where no one can work with them, and incompatabilities and instability arise?
One could argue that while they are definately users and evangelists of the open source movement, they are still a closed-shop. If they were truly open source, and constantly contributing back to the community, wouldn't someone have known BEFORE it was released that they were doing some wacky things in their software? I mean, consider it...yes, we get to see their sourcecode, but we only get REAL access to it when they release it, not during the actual development of it. So while we've all been hacking away at the RH6.2 source, making improvements and sharing these to the rest of the community, we couldn't plan for and anticipate the changes that were being made and improvments (AND BUGS) that would be introduced in the RH7 source because we only get access to it AFTER it's released.
The point is, it's these kinds of closed-office things that create and instigate the comparison to Microsoft. And in the same respect, it is these kinds of responses and letters (The ones that don't really address the issues, but dance around them) that are typical of Microsoft (and any other software/hardware behemoth) to put out, expecting to somehow use marketing-speak to appease us.
I may be going off too far on one side, being a bit extreme, but I'm simply trying to raise an issue here. I love RedHat... I love the concepts and togetherness of the opensource effort... I just don't want to see them screwed up by any one company just because they become the dominant player in the field (like Microsoft).
I don't know what the problem here is. I ran over a month with RedHat 6.9 (the 7.0 beta) with a dual-CPU and dual-headed setup. How many crashes? 0.
Will my production servers get 7.0? Hell no! Will my production workstations get 7.0? Hell yes! It's stability against features. You have a choice with RedHat ... a free [speech] on at that.
Learn the RedHat . release ruling: .0 = bad, .1 = tolerable, .2 = stable
-- Bryan "TheBS" Smith
-- Bryan "TheBS" Smith
Independent Author, Consultant and Trainer
When I found out that RedHat 7.0 was finally coming out, I basically shrugged and made a note not to go to sunsite for a few days. Now, I'll admit that I did look at some of the 7.0-beta packages, but was dismayed when I found out that I needed to upgrade my version of rpm -- and I did not see a way to upgrade it. (Admittedly, I failed to look in the 6.2 updates directory.)
I'm just going to go on what I have read here and heard otherwise... Apparently, gcc 2.9.6 has issues. Yes, most software has issues, but these seemed to be serious issues.
I'm going to wait on upgrading my version of gcc, it seems. Although my computer was initially an installation of Red Hat Linux (version 5.2, in fact, which had a buggy-as-hell implementation of GNOME on my box), I decided about the time 6.2 went into beta to move away from the rpm system software methodology. Now, if I can't compile the software packages I want to install, then I'm sunk.
I have to agree with those that say that Bob Young's letter was long and pointless. No, Red Hat is not Microsoft. Red Hat will never be Microsoft. It is impossible for Red Hat to become Microsoft without breaking the GPL, which most of the software for Linux is distributed under. It's just not going to happen.
And, for those of you out there who don't realize this, you don't need Red Hat to run Linux. Debian, Slackware, and SuSE are all alternatives if you want major distributions. Or, hey, go hit freshmeat.net for the tarballs and build them yourselves!
I just have a few questions: How long was 7.0 in beta? How many people tried to download the packages? And was gcc 2.9.6 one of the beta packages?
If gcc 2.9.6 was amongst the beta packages and there was a long enough review period, then Red Hat had no reason to use a different version of gcc if they did not receive bug reports or complaints. The purpose of a beta is to iron out the bugs before a gold release. But one cannot fix bugs that one does not know exist.
However, if gcc 2.9.6 was not within the beta packages, the test period was insufficient, or Red Hat ignored the bug reports, then complain, piss, and moan all you want for you are justified.
And, remember, everyone makes mistakes. If you want proof of this, look at some of the changelogs for the 2.4.0 test kernels. (Okay, maybe this is unfair, but it proves a point.)
Linux is an open source, free operating system. Most of its software falls under the GPL, which ensures open source software and whose developers choose to release the programs for free. It is impossible for any one distributor to gain a monopoly over it like Microsoft over Windows.
Linux is a movement; its user base supplies its voice. Yell loud enough and you will be heard.
--CAE
Maybe Bob doesn't get it because he's not really paying attention to what his company is doing.
Redhat, the company, is going about setting up deals with many businesses, such as Dell and IBM, in ways that cause them to prefer the Redhat distribution over others. What Microsoft has done in the past, and what I compare Redhat with, is the practice of trying to make sure that I have no choice in OS for my computers. In the case of Redhat, it's no choice of distribution.
This is different than encouraging some company to offer Linux, and Redhat, with their hardware. These are cases where the hardware vendor will refuse to support their hardware problems when the software being run isn't Redhat (or Windows).
Bob, if you want to be a better member of the Linux community, then work to encourage hardware vendors to not just support Redhat, but to also support their hardware with not just any distribution running the Linux kernel, but other operating systems, such as xxxxBSD, as well.
now we need to go OSS in diesel cars
You know, it's odd you mention that. I've worked on HP boxes, Sun boxes, AIX boxes, etc. And the only place where ping was different as to that -n switch was with windows, which does 4 pings and stops by it's default. I susupect what is happening is that you're seeing new Red Hat users run ping, and are suprised when it just keeps going...
Unless you're talking about a different problem with the number of counts produced with a ping, then I think you'd best ask more than just Red Hat to "fix" their problem. :)
Well put, Bob. Even if there were 2500 bugs in RH7, I'd rather _know_ they exist, and disable the running code that I can. If you run a Windows 2000 Server, you know there are bugs in the code. But where are they? Microsoft knows about many more bugs, because a SP2 is in the works. But Microsoft will not tell you about a bug if you don't find it.
:-)
Would YOU step up to your boss and say "look, you would have never seen this, and you may fire me because I'm telling you, but I did a half-assed job".
I run Win2K on non-mission critical systems. And I like it. But I understand the Microsoft Way of business. I also run Red Hat Linux 7 servers for more important stuff. I know what works and what doesn't. Appreciate both sides of the coin. If you don't like, it, say so and say why, but let's stop crying
Or use a little perl
perl -e 'while(){print "freedom & personal responsibility good, serfdom & tyrannical control bad\n"}'
---
"...silence is a dangerous sound."
Neither will binaries compiled with 2.95.2, so I don't see your point here.
The GCC Steering Committe have declared that gcc 3.0 matters are still subject to change - so both 2.95.2- and 2.96-compiled stuff won't necessarily work with 3.0. So the "gcc 3.0" compability argument is moot.
GNU/Linux. The Freshmaker.
That may be so, their x.0 releases are still buggy as fuck-all, but that is almost inevitable. They are doing something difficult, and they can't possibly test every part of the distribution (including the hundreds of software packages that come with it) on every possible hardware and software configuration. If some things don't work, they'll be patched in later minor releases, or even better, incrementally in the updates directories. The same people who are bitching that x.0 never runs without hiccup out of the box are also the same people who'd piss and moan to no end about how long it took for the next release if they took longer to do more extensive testing. So, in short, buggy x.0 software has been around as long as software has existed, and people can lump it and deal, it is a fact of life, just like hangovers, conservation of energy, and gravity.
---
Play Six Pack Man. I
>model. By comparison (I'm not sure if anyone
>noticed this) Computerworld had a front page
>story a couple of weeks ago about how there were
>problems with Solaris on Sun's Enterprise
>systems, yadda yaddda
I'd have more respect for Bob/RH if he hadn't responded to all this.
Here's a very basic premise of life in general (IMHO) - "Comparing your own stupid, boneheaded move to someone elses MORE stupid boneheaded move doesn't make yours any better."
>Remember that this debate was begun by someone >going to Red Hat's public site and trying to add >up all the registered bugs in Red Hat 7. When
>was the last time Microsoft (or any other legacy
>software vendor for that matter) gave you access
>to their complete bug registration system?
Microsoft giving access to their bug database has NOTHING to do with the # of RedHat bugs.... or does it
Microsoft _may_ be evil (not a silly debate that i let myself get pulled into), but don't try to make yourself look BETTER by comparison. I mean, we could bring Stalin into it and make RH look amazing!
j
All three of the above categories have different reasons to form opinions about Redhat (and other opensource/linux companies, too BTW).
Slashdot readers are a combination of all of these. I think it's important that we consider Redhat from a number of different angles.
1) Opensource developers: well, we'd (I'm one...) probably quit our day jobs and code for peanuts if we had good management, brilliant co-workers, a fun work environment, and good coffee. - Face it, we love to code and we do it in our spare time. We probably don't care too much about a particular distro, since all of our machines are so highly customized as to become almost a custom distribution. BUT... we do appreciate the fact that Redhat and others pay talented developers to write opensource code that everyone can use.
2) LEECH: - don't care about anything except what they can get for free, complains about anything that doesn't work, makes lots of noise, has more free time than money, doesn't contribute back, thinks Redhat sucks because they're trying to make money.
3) shareholder: - wants to know that RedHat is looking out for it's business interests. Giving back to the community has yet to prove it's business value to investors.
it's a difficult balancing act...
As one of the people who's posted some of the stories that Bob may be thinking of here, I'd like to put in a quick word about my attitude toward comparisons in general and Red Hat in particular. a) I like Red Hat -- it's one of the 1st distros I ever installed, and the boxed 7.0 is one of the distros that I ve bought and recommended to other people. b) Comparisons *to* Microsoft and not the same as comparisons *with* Microsoft. Not that either of these might not be appropriate in certain cicrumstances, but there is a difference. Saying that MS has bugs and that Red Hat Linux has bugs may sound alike, but which of these would be you trust to be fixed quicker / better? I'd rather bet on Red Hat for most things. Since I'm not hugely interested in Windows, I can't comment on their bugfixes with great knowledge, but there are certain long-standing bugs which I'm familiar with . Red Hat, on the other hand, seems to fix things pretty quick. That's my take on them. I hope they continue to grow and improve Linux at all levels. timothy
jrnl: http://tinyurl.com/c2l8yr / foes: http://tinyurl.com/ckjno5
Of course Red Hat isn't like Microsoft.
:)
Microsoft is a publicly traded software company which sells operating systems and is currently trading at their 52-week low.
Oh, wait a minute...
(it's funny, laugh)
I don't feel that comment deserved the flamebait moderation, the comment may be wrong, but it wasn't flamebait.
plastickiwi here feels that Bob Young was wrong in defending his company in this matter. Apparently plastickiwi didn't understand how the complaints people were making about RH7 were going to far.
People were attacking the company saying that the entire coroporation was at fault for a buggy release, and that Red Hat wouldn't let the situation get better.
I've never used RH7, but I have used pretty much every release from 4 - 6.2. Some were buggy, some were'nt. But I never saw a major Red Hat-caused bug in two versions in a row.
I also have used pretty much every version of Windows. I don't think I should have to tell you that I have seen many bugs pass through to the next version unfixed.
This is the point Bob Young was trying to make, he was not defending RH7, he was defending Red Hat.
Of course by know this is redundant, doh! slow typing.
Devil Ducky
Devil Ducky
MY peers would get out of jury duty.
I don't know actually. Perhaps I should have said "arbitrarily differential pricing". For instance, I don't think it is wrong for insurance agencies to charge for different age ranges and car models. That's plain sense - it's more risk for them. However, something like Amazon charging *arbitrarily* different prices, for no other reason than to gouge, doesn't seem to have much ethical grounds (yeah, ok, they said it was a "test"...hopefully it doesn't become their modus operandi).
It's 10 PM. Do you know if you're un-American?
I'll repost this:
I don't know actually. Perhaps I should have said "arbitrarily differential pricing". For instance, I don't think it is wrong for insurance agencies to charge for different age ranges and car models. That's plain sense - it's more risk for them. However, something like Amazon charging *arbitrarily* different prices, for no other reason than to gouge, doesn't seem to have much ethical grounds (yeah, ok, they said it was a "test"...hopefully it doesn't become their modus operandi).
It's 10 PM. Do you know if you're un-American?
I think there are some ideas in common between Libertarians and Greens, but they are for different reasons. Libertarians don't believe in the vast wealth and power of some corporations because they hinder the individual's freedom to become vastly wealthy and powerful himself. I think Greens don't believe in the vast wealth and power of corporations explicitly because vast wealth and power gaps are inherently undemocratic and socially unhealthy in general (to paraphrase the quote: "you can have vast concentrations of wealth and power, or you can have democracy; you can't have both").
And again, on drugs, Libertarians think that the government should stop telling citizens what they can and can't use because it restricts their freedom. The Green perspective is that government should stop the war on drugs, not just because people have some right to get high, but because it is fundamentally a war against *people*, which is helping *nobody* (regardless of the freedom issues). Look...these holy "freedoms" and "rights" we have only exist because someone is there to stop somebody else from infringing on them. Freedom comes from restriction. I think we should have as little restriction as possible to maximize the cost/benefit equation...but I'm not so naive to think that totally abolishing any type of government or regulation or arbitrating entity will land us in some wonderful fantasy land. Without *some* sort of government we are back in the "natural" state that we invented up government to avoid in the first place!
So, yes I think that Libertarians and Greens share some goals, but I think they do for different reasons. I'm not here to run down Libertarianism. I think it is a worthy philosophy. Certainly better than the business-as-usual Democrats and Republicans. I think we have some decent third party choices here. My pick is the Green party. I don't think there is any reason to bicker over the differences of these parties, as they are already both vastly different and superior choices to the status-quo parties.
It's 10 PM. Do you know if you're un-American?
Bob, what the heck is the point of that letter?? Was it to justify the release of 7.0 or just to divert the attention from the real issue of a buggy release by using the ever_successful "Microsoft" name ??
/. I was expecting a good rebutal, trying to support the decision to release 7.0 and maybe a list of what all is good about 7.0.
.Why ?? Well, I tried to install 7.0 . Everything was ok until the dhcp part. 7.0 ships with a dhcp client with a version 1.3 or so , when the latest is 3.x !! How more outdated can you be.
I really dont get it when people just tend to use MS anywhere and everywhere. When I read the main page of
But now, when I have finished reading the letter, I feel cheated. And for all those who are keen to know, I am using 6.2 and not 7.0
That was enough to put me off..
Thanks but no thanks for 7.0. I am waiting for 7.1 (hopefully I will be able to use it).
... "follow me" the wise man said, but he walked behind
It's on the second CD, in the /preview directory.
You have a bunch of 2.4-testXX kernels.
MAN! Just go to a mirror site and take a look. It takes 10 minutes.
By posting you managed to unmoderate it yourself. :)
--GnrcMan--
So if you want to criticise us for shipping gcc 2.96, you have every right to do so - you'd be wrong, but it is at least a legitimate debate and I'd respect your opinion. But to compare Red Hat to Microsoft indicates an ignorance of what is driving our success.
I am sorry, but the simple fact that exit(0) breaks in certain programs really upsets me. I do not have this problem with RH6.2...
The GNU team released a statement as to why this is bad juju... defend this one all you want, but I feel the majority of us are fairly unhappy with this move.
RH's defense on using a snapshot compiler for a better transistion to the 2.4 kernel is not a worthy defense. It isn't out yet... and still has some time to go. Why not address kernel 2.4 issues when, well, kernel 2.4 is out?
RedHat is not Microsoft, I do believe they are a very good company, but others and myself still believe this snapshot inclusion of GCC to be a Bad^Idea(tm). It may be an opinion, and of course opinions really can't be surmised as wrong or right.... they just reflect our views.
"You should always go to other people's funerals; otherwise, they won't come to yours." -- Yogi Berra
I don't think that is valid. It would be 2007 before all the bugs are fixed. No one waits to deliver a "perfect" product, at least not of this size, and they shouldn't. You correct all the major ones, publish all the rest for people to see to let them determine how it would effect their deployment, and put it out there. ......
It's all a judgement call, whats major, whats minor, even what is a bug.
Should they have waited? I don't know, I haven't had a problem with my install.
Including "beta" releases of components in a final release is very questionable, but hey, they weren't keeping it a secret. You don't have to deploy it. Stay with 6.x or shit go over to Debian, Slackware, SUSE,
- I like pudding.
If there was a head on the proverbial nail of Slashdotters, you've hit it. Folks around here just don't to see success from anything other than their select group of interests. While the whole "one click shopping" patent is a little obtuse, Slashdotters instinctively think they should hate Amazon.com. When the issue about Amazon charging different prices to different customers came up, most of /. jumped all over it. What's the problem here? If you don't like it, don't shop there. "Amazon.com has sold out!" No shit. They're in business to make money.
/.'ers hate for no good reason other than the fact that they're successful.
Lately I've seen a lot of pro-Nader posts. I prefer Browne, personally, but Nader makes some good points. But it seems to me that a large number of Slashdotters heard Nader was anti-corporation and jumped on the bandwagon. No, I don't think large corporations should run the country. Yes, I do think that a corporation, no matter what size, deserves to make money if they provide a useful service. I'll spare you the list of companies that
And now we're at a new stage ever since Big Money entered "our" world (meaning OSS). Slashdot is aquired by Andover who files for an IPO, Red Hat goes public, etc., etc. Now all of the sudden it's "Slashdot sucks!" and "Red Hat is too commercial!" Fine. Don't go to Slashdot and use a different distro. It's just a matter of choice and opinion.
-- "Complacency is a far more dangerous attitude than outrage." -Naomi Littlebear
Duplication of effort is what open source software's about though! Well, not quite but:
- Developers work on what they want to work on. You can't reassign them as though doing OSS development is their job or something.
- Parallel development is a good thing. The way Open Source development coordinates tens of thousands of developers (or however many) is by saying "go do what interests you", and
.. well, no two people write the exact same programme. If it's all open source eventually the best ideas get integrated.
I percieve that "why are you wasting your time working on similar projects" argument as a whining remark that not every developer in the world is focusing their energy directly on making your life better.In fact, I would say that open source development is a good example of mimetic evolution. Multiple programmes sharing the same niche allows more vectors for improvement. The more journaling filesystems that bring ideas to US the better off we are! (given that we're working with a modular kernal that can support multiple filesystems formats easily).
Trees can't go dancing
So do them a big favor
Pretend dancing stinks!
Mr. Young is obviously right on track with Microsoft as evidenced by his blatant threats to those who may oppose (i.e. You will be punished - now start writing!).
Interesting article, but I'm a little surprised that he skirted the RH 7.0 stability issues as skillfully as a politician.
They call me the working man. I guess that's what I am.
I totally agree with bob on this letter. Comparisons with microsoft & stuff like that are just echoes of the way people sometimes do this stuff simply to "play catch up to microsoft", ie, see themselves as something that can be compared with it, and limit themselves and their software by seeing proprietary versions of what they do as some kind of perfect and unobtainable goal.
Open source is a completely different thing, from the nature of the software to the way it's made and you can't compare the ideal microsoft world to the ideal open source world, because they are in many ways complete opposites.
Red hat is only trying to do what open source stuff is meant to do: release early and often, get the eyeballs to report (and maybe fix) the bugs, and release a followup version with less bugs. It's not the same as when a commercial company that writes proprietary softwear releases something like that, where in some cases you could argue that they are meant to deliver a working product. Red hat 7.0 users are taking part in the actual process of delivering a good followup to 6.2
On the other hand, it can sometimes be worthwhile to use our knowledge of the workings of proprietary methodologies when working on oss: that mention of bob's of microsoft's bug database made me think: if microsoft's bug db is consistent with the way they usually write apps, they probably share a lot of the core stuff they're made up of. I'm sure their bug db, and the internal communication tools they have between programmers, and the content revision system they use must be integrated to some extent. Maybe it's all the one app? I have no idea.
I heard somewhere that cvs is 10 years behind it's commercial equivalent. ALso bugzilla has no strong integration with developer mailing lists or the cvs tree (but please prove I'm wrong here) so when are people like red hat, with large scale stuff going on like distros, going to have a suitable tool to make the debugging process more manageable. I wonder if the tools available now are adequate for things like distros?
Maybe the communication between developers should be integrated with the code itself somehow, so that people can have an easy way to find out about design decisions and how bits of code evolved over time.... But I ramble...
Ale
Very insightful, IMHO. I agree that someone needs to be the MS of the Linux world, so to speak. You can't really be like MS in the Linux world, since the community and development model completely prohibits that.
I wish Bob Young would realize that Red Hat is pretty much the MS of the Linux world, and just admit it.
As for having them define a standard and just have everyone else follow it, we all know that that is not a good thing. Provided the new "standard" really is technologically superior, it makes sense for other vendors to stop competing and just jump on the bandwagon. Compatibility is very important. But just because Red Hat says it's good, doesn't mean that it is. For instance, I still think this kgcc/gcc thing was a bad idea.
--
And the men who hold high places must be the ones who start
To mold a new reality... closer to the heart
This is a problem with many grass-roots efforts. It's a big problem in the OS/2 world. For instance, there are a number of OS/2-related magazines (online and print) each of which is effectively competing against the others, although the editors of these magazines will deny that. So what happens? Instead of one awesome magazine that everyone knows about, we have a handful of smaller ones all over the place and they all have weak content. It's especially frustrating when you hear the editors complain that they don't have enough time to get the issues out promptly!!! It's so stupid.
The same is true with software development. OS/2 used to have about four commercial and shareware news readers. All of them eventually ceased development because of poor sales. Well, duh! There aren't that many OS/2 users out there, and by having three competitors with almost identical products, you've effectively reduced your potential market share by 75%!!!
--
And the men who hold high places must be the ones who start
To mold a new reality... closer to the heart
It is great to see that Bob Young is addressing these problems. It is troubling, that this problems existed in the first place. To compare RH with Microsoft would, of course, be a mistake, but to say that all that is not MS is good, would also be a mistake.
RH admitted mistakes, I think we should all take a deep breath and wait for RH to fix the problems, not address them.
-Moondog
Nothing too large to say here, however I would like to point out one thing...
/var/spool/mail/, instead of using mbox files in the user home directories.
I had problems with the IMAP server shipping with RedHat 7 after upgrading from RedHat 6.2. By being able to query the bug database, I found that my problem was not really a BUG, per se. It was the way that the IMAP server places mail (using
It would have taken an amazing amount of time to find such an issues in other OS's (and even some other Linux distros), but by RedHat having BUZILLA open to anonymous, public access, I was able to migrate user mail to the proper placed without any major hassle.
Just my 2 cents.
That's slashBot.
`ø,,ø`ø,,ø!
Free Software: Like love, it grows best when given away.
The Microsoft response was "Very strange. We've never seen this problem anywhere else. It must be something strange that your company is doing. This went on for weeks (if not months) until my friend talked to his counterpart at another large company that was having the same problems with recurrent file corruption.
The next time 'Dave' talked to his MS contact he mentioned that he'd been talking to this other support person.
Dave: "I was talking to X at Y company, yesterday." MS: "Ah yes, Him. I talk to him all the time"
Dave: "Oh, so you know about his problem?"
MS: {long guilty silence}.
(names withheld to protect the NDAed)
`ø,,ø`ø,,ø!
Free Software: Like love, it grows best when given away.
Would it help if Red Hat released a very public beta distro prior to the actual distro? Like a "Red Hat 7 Beta" distro?
They probably have a beta posted on their FTP site, but that's not what I'm talking about. I'm suggesting a much more grand release, automatic free CD's sent to customers who have purchased multiple Red Hat products, etc.
That would let these 2000+ bugs be worked on by developers before people start installing the distro on critical systems. From reading comments on RH7 the problem is not really the quantity of bugs (just over 2000 is not many, there were several times that with Windows 2000 when it came out); the problem is that several basic functions were inherently broken/altered - people had problems with *ping* and *gcc*, for crying out loud. These are not frills, they are essential to daily operations on Linux. A beta period would have captured comments from early users and supporters and (a) guided Red Hat in making decisions about what to include, and (b) given the developers time to iron out the most pressing bugs.
Obviously a beta works best with the most number of users, and a lot of people like me enjoy Red Hat but we use it mostly at work where we don't want to introduce potential problems. So you'd want the beta installed on less important systems and home/hobby systems. Perhaps Red Hat could combine this with an incentive program. For example, if you download and install the beta distro, you can run a program to register with redhat.com and your username is credited with $15 off the purchase of the full package when it comes out... or they send you a cool T-shirt or something. Whatever.
Of course "beta" packages are a huge tool of Microsoft, but in their case they use them for marketing purposes, to grab mindshare early on. It would be nice to see a company run a serious beta program with the objective of connecting with customers. This would be another way to differentiate Red Hat from Microsoft.
Just MHO.
Actually, Win 95 had a bug in the system clock that caused it to die after 45 days (integer overflow). Although a few people have heard of this, they were not drawn and quartered, and their stock price did not fall. The fact is that you have fallen for the MS marketing hype too.
Bob,
Regardless of the model, yes it works and it's great but how for the love of Linux could you box that release?
Ok now me I'm a RedHat user just because. RH6.0 with a few updates is my happy alternative home when not Windozing. I have no compelling need to get up to a more current version, so I probably won't. I really don't see myself going to RH7 though, not maybe till some time middle of next year, if at all.
Are you going to follow the MS model of shiping junk and waiting till SP(large # here) before people trust your releases to deploy critical systems on? Come on, we can do better than that!
Uptime of 3 weeks? I would hope you would at least have some test systems running that say we at least have to be able to run longer than the 49 days or so that a Windows machine would max out at before this gets RTM'd.
Bob, you've got a chance to lead here, there is no rush to release because you're afraid someone is going to be first to market with anything. You have to have some kind of quality standards if you want to be the #1 distribution. All the support engineers in the world are not enough if your released product doesn't have the big bugs shaken out.
Linus won't let 2.4 out the door till it's ready. That's why he's got the respect and confidence that you really need to get. And you can. And the community will gladly help. Get it right, show the closed shops how it's done.
Thanks,
sunking7
"a powerful and unexpected ally..."
When it happens, we'll see. What I do know is that I will not base the future of my company on closed-source technology. If not *PL, then at least source code license.
I don't care if my word processor is Open, but servers and middleware components are Open/Free and have been since 1997.
Besides, when I do develop a killer app I will release it for a simple reason: doing so makes it better.
Now hiring experienced client- & server-side developers
-- @rjamestaylor on Ello
Now hiring experienced client- & server-side developers
-- @rjamestaylor on Ello
Stop. How can you quantify nagging bugs as less when a company (no, an industry) employs NDAs?
I've seen the future and it is Open.
Now hiring experienced client- & server-side developers
-- @rjamestaylor on Ello
Yeah, you could view it as an attempt to save face. Whatever you want to call it, what he said is the truth. Maybe I dont like open source software. Maybe I dont like Closed Source software. But the things he said *ARE* true.
Maybe it should be more self evident, but for some reason people like to complain a lot.
Why did people compare you to Microsoft?
.0 releases just makes you look like a dumbass.
Your boy, Eric Troan, referred to the idiocy of shipping a experimental compiler with RH 7.0 as "innovation". That's the only reason anyone needs.
My personal reason? RH 7.0 thrashed my machine -- absolutely refused to run XFree/DRI no matter what I did. At lest 6.2 worked after some massaging.
Debian 2.2 may be tougher to install, but it is obviously a higher quality product for the experienced user. They don't experiment with my machine the way you do.
Stop playing "race to market" with Suse, and start making quality releases...these stupid letters defending your low quality
I strongly urge anyone at RedHat who will listen -- DROP YOUR DISTRO AND USE DEBIAN. Add helix-gnome, and stop putting your lame logo all over the screen, and actually test something before shipping.
This is not about Sun, Microsoft, closed or open source. It's about a shabby product being defended by someone who refuses to admit they hosed the release. That does sound like MSHAFT.
We expect/demand better from RHAT -- deal with it, dumbass.
Treatment, not tyranny. End the drug war and free our American POWs.
See my user info for links.
You see this sentance fragment comes off as very cocky and self-assured. But with the proper smiley insertion:
So if you want to criticise us for shipping gcc 2.96, you have every right to do so - you'd be wrong :),
Your point comes across as the semi-joke you meant it to be.
-Zane
This sig is worse than my last.
Personally, I'ma bit disappointed with RedHat this time. Not because of GCC (though I think it was a bad idea), but because they have tended to stray from the standard. Here's my personal story:
I wrote a tool to make configuring LPRng easy. I really like it. Several other people like it too. RedHat started using LPRng in RedHat 7.0. But, while the LPRng rpms available on the company's site (as well as their tarballs) set the deafult user/group for spool directories to daemon/daemon, RedHat decided to make the default lp/lp. Sure, the change makes sense, but now I've had to patch my program to work with their new Linux release. Why didn't they just keep the original rpms? The original LPRng rpms worked fine, but RedHat's change has forced me to waste several hours working out a way to fix this. Why must they be different? If it was a major improvement, maybe I could understand it (or even a minor improvement, for that matter).
Just my $.02.
I am surprised Bob mentioned Sun as a negative example (regardless of the merits), given that Sun's contributions to the Linux world (Java, StarOffice) must be very helpful to RedHat (putting aside licensing concerns re Java).
I for one agree with Bob Young. Large scale software products alway have bugs and always will. The difference is the business model which i think in Redhat's case is heads above M$. Dot 0 releases in almost any product have general problems and it is not until the product has been put through the extensive tests of production do most of these bugs show up. If you want bug free code wait for 7.1 or 7.2. If you want a company that truly doesn't care about its customers go buy Windoze. But just quit busting on a company that is working hard to do the right thing.
--Billwashere
So, let me see if I have this straight--
A company publishes a software product riddled with bugs. Customers complain. The company's response is, "Well, look how much worse the other guy is! At least we let you see the source code!"
What compelling defense of RedHat 7.0 am I missing here?
-- He's fantastic, made of plastic....
This is true to a large degree, but glosses over a significant difference between consumer products from The Software Industry and the mass produced output in other categories. Many corporations, large and small, engage in practices that many would consider shady, unethical, or similarly negative. Company X may be exploiting workers in OtherCountry, but that's not what Joe Consumer sees. He sees a Widget from Company X that comes with a money back guarantee. Granted, if Joe isn't satisfied, he's probably not going to take the time to actually get his money back (depending on how expensive the Widget is), but he still doesn't think too badly of Company X because they have a standard practice of "customer satisfaction".
On the other side of things, Company Y in The Software Industry has wonderful internal policies and practices. The employees there get tons of benefits, etc. But the flagship product, SoftWidgets, ships with no warranties, no "customer satisfaction" policies and a "license" that often amounts to legalized censorship. If Joe Consumer is unhappy with SoftWidgets and wants it fixed, or his money back, or any sort of customer satisfaction, he's out of luck. The end result? Joe Consumer feels like he's being exploited, and paying for it.
With practices typified by Company Y being endemic for over 10 years, many people have become reactionary to dealing with any problems with software. If the normal approach to dealing with customer problems doesn't work, turn up the heat. And this is applied categorically.
In this specific case, RedHat shipped a product that made many customers unhappy. And a large number of these people are still fed up with the hostile practices of places like Company Y, so they're dealing with RedHat the same way they would deal with places like Company Y, where such tactics are the only way to get anywhere.
T. M. Pederson
"...and so the moral of the story is: Always Make Backups."
T. M. Pederson
"Lies, Damn Lies, and Documentation"
Bob, if your listening:
I completely disagree with this guy - keep up the good work. I dont think he 'gets it'. By putting 7.0 'out there' you will push the implementation of the solutions... moving us all forward -quicker-. Release early - release often. The reality is we dont even have to pay for 7.0 - we can all wait for 7.1 then buy a boxed cd. So if the 'stability' is an issue, stay @ 6.2 until 7.1 then buy yourself a box... otherwise simply make the 'features' you _need_.
Oh, and Ummm Bob, thanks for not filtering your letter through the mindless-pr-markatroid-bs department. This group can smell pr-speak a mile away, and you did the right thing by not pushing that BS on us.
If you think that Red Hat's is a stinkpile of doo-doo don't use it. Try Debian, try Slackware. That is the whole point here.
In principle, that's great. But when Joe P. Accountant, sick of his Windows box crashing all the time, walks into a computer store and buys a copy of Linux, Red Hat is likely to be what he'll buy. He's heard of Red Hat.
And when he installs Red Hat 7 and it crashes his computer in 50 new and interesting ways, he's going to blame Linux, not Red Hat. New users don't understand the distinctions between distributions.
He doesn't know or care that Red Hat is different from Debian is different from Corel is different from SuSE... As far as he's concerned, Linux is Linux.
So, a bad commercial distro, especially from a recognized name in Linux, is going to drive away lots of new users who try out that distro as their first Linux.
More than any other form of advocacy, Red Hat, Corel and all the other big names have a responsibility to Linux users to ensure that their distributions work at least as well as the competition's operating systems (Windows 2000/Me). And I would argue they must work significantly better than Windows, since Linux has to overcome the Windows inertia.
I like Red Hat. I've tried several other distros, and I love the information that's available to me as I run a popular distro. I like Red Hat's website, and hell, I even like their logo. I don't like it when people bash Red Hat. But let's face it, if RH 7.0 is anywhere near as bad as RH 6.0 was (and I'm led to believe it's a lot worse), then they really dropped the ball.
Fire and Meat. Yummy.
I AM CANADIAN!
So am I. I feel your pain.
Being a market leader in Canada is like being a market leader in Kansas.
Fire and Meat. Yummy.
Hi Lawrence, my name is Jonathan. I started using Linux with Redhat 5.0 about 2 1/2 years ago and my first installation was on a Compaq.
Welcome to our little 12-step program.
This was one of the most difficult things I have ever done with a PC; and one of the most rewarding.I agree.
I encountered the "Ll" problem and so many others that I cannot recall them all.LILO problem aside (not Red Hat's fault, but it needed to be addressed earlier than it was), how was a later non-x.0 installation?
After a while I realized that I needed to R T F M. It was an uncomfortable realization - I actually had to learn something before I could *use* this OS.R T F M, *S*. I did. And yet, I was completely unprepared for the *loads* of problems with RH6.0.
That installation took weeks. WEEKS dammit! But it finally worked. And it has gotten easier all the time.I agree. One of my boxes still runs RH6, mostly because I haven't been able to accomodate the downtime required to shut it down and reinstall everything. But, I restate: RH 6.0 was an unacceptable product; too many problems. More problems, in fact, than a 1985 Hyundai Stellar I bought for $100 as a winter beater one year. I appreciate that Red Hat works hard for our community, but I also ask that they show that by not releasing stuff that's not yet ready for the big leagues.
I value an OS that is user-installable as Redhat now is, but I learned to love Linux because I first hated so badly how little I knew about it.A well-equipped neophyte can install DOS 5.0 and up and have it work first shot. It's far from perfect, to be sure, but it works. Any Linux distro, aimed at the new Linux user (as, I would argue, anything sold in shrink-wrap at Fry's, is), should actually work first shot.
Maybe that is a distinction between a Windows user and a Linux user, maybe we are just more stubborn.Oh, in my case, that's not an issue. That's how it is that my Linux systems, even my RH6.0 system, all work now.
Our common, mostly unstated goal as a community, is to usurp Windows, right? Screw the server market; anyone with any sense is already running Linux/*BSD/UNIX there. But to get the desktop, *most* installations, especially on low to mid machines (most people I know seem to try out new operating systems on the machine collecting dust in the closet), will have to work properly, first shot. Or else users less stubborn than you and me will simply go back to Windows. That which is easy (known) is comfortable, even if it blue-screens twice a day.
Out of the gate, any Linux distribution in shrink-wrap should work out of the box significantly better than Windows or indicate prominantly on the box that it's for advanced Linux users, or we're just shooting ourselves in the foot.
I applaud Red Hat for their installation program. I boo them for the fact that when it's done, the system (RH6.0 and probably RH7.0) still doesn't work.
This is unacceptable.
Fire and Meat. Yummy.
Mr. Young you're talking to an askew view of the gripes being hurled at you. The sentiment is, RH implements major tool changes, such as your nominal gcc 2.96, that are way outside the logic of most Linux users. I know it bothers me when I read the GCC steering commitee's announcement chiding such a move without any =general= public disclosure as to why such a change is being implemented. Further, it irks me to read RH spokesmen try sophistry when explaing your rationale for such kinds of changes. But, what gives rise to the cake is that by the de facto influence of RH market share RH possesses great leverage in Linux policy. Thus, because I greatly like Linux, Red Hat Linux engenders no similar loyalty in me. IOW, you have the greatest market share, you have (undeniably) therefore great influence, you need to be doing *it* better than the other guy or anybody else. The 'it' is concensus -- you must be doing soemthing wrong to have a bunch of serious users ticked-off at you, accept that fact already.
Lastly, I'm a RHL6.1 user, but I have to tell you, in all sincerity, this distro has given me nothing but grief. From xterm not working properly, to being unable to mount CDs because X has started first (a very quirky bug; Deja mentions several occurences; RH Web support at last look {
I don't hate RH but I love Free Software(TM). BTW, I am surprised ni one else is mentioning this, but it is my theory that RHL ditros are released too soon (Again, I disagree with you on this point, B Young) because RH is trying to surpass SuSE as the highest release-number distro available first -- point in case, RH marketroids suceeded in the silly ommission of the dot 0 in the Official name of "Red Hat 7", the latest RH distro. Isn't that something goofy.
--
Me pican las bolas, man!
Thanks
--
Me pican las bolas, man!
Thanks
Jaco
I want to second this thought and add to it. Not only is the reason why Red Hat is a company that is worthwhile in the world, but this is exactly what is so great about Open Source and Free software. I'm just a user, for the most part, and while I think Microsoft has written a few really great applications over the years, I've never once had an experience with them that made me (as a customer/donor/user) feel like I would really want to buy anything from them ever, because they are so incredibly dodgy in the licensing and CYA areas. But when I've taken the time to subscribe to a mailing list for an OS application or distribution, I've found that the developers themselves read those lists and while they might not answer newby questions, they certainly participated in the discussion. I've yet to see an OS developer/company/foundation take the easy way out of anything. This is a highly ethical way of doing business that we could all learn from. And it's a hard business to be in, especially when many on Slashdot seem to expect to get their Free software without ever giving back in some way (either by donating/buying or by coding something useful). Developers gotta eat, and I know I'd much rather give my money to developers who valued my freedom and me as a customer-- something that Red Hat puts a lot of effort into (in my opinion).
I do not have a signature
My answer to them, Use KGCC, it's EGCS 1.1.2!
For all of you who don't believe gcc 2.96 is useful, you're dead wrong. Lots of heavy commercial shops building inhouse C++ software NEED gcc 2.96. I work at one. Our company will now make much better and faster leaps towards using linux for real corporate computing (as opposed to just sendmail/apache servers) because of gcc 2.96. And the more corps jump on the bandwagon, the more support Linux as a whole gets...
Which of course will lead people like IBM and SGI to keep turning over key technologies to the open source community, and feed the whole process.
11*43+456^2
In the interest of good open source-dness, I'd like to know how it's "wrong" to object to the presence of a compiler that's not a public release, not a stable beta, and not even considered a "stable development version" by the team that designed it. No offense intended, sir, but that's a lot to swallow, and even more to defend in front of a client or executive.
I understand innovation. I appreciate some dominant distro taking a bit of a technological leap to push the progress of the whole Linux market. That's actually a good thing, and RH has done it before in *.0 versions. I would prefer, however, that 7.1 greet my machine with a rock-solid gcc 3.0 and recompile my kernel in the process. I'd prefer that my inet superdaemon be the industry standard, Berkley's inetd, with a single config file that I can scan at a glance, instead of "xinetd" with its whole directory full of config files. I don't consider that an improvement. I'd rather the exciting new Red Hat Network daemon be, at least, a debugged product before finding out on the internet that it crashes the system after three weeks uptime. I understand improving RPM to 4.0, but making it completely backwards-incompatible is a real leap. It makes experimentation with 7 more of a wholesale leap into the deep, red fog. (BTW, thanks to RH for the cd full of apps already packaged with RPM 4, which eases the transition a LOT!)
Bugs in a *.0 are not unknown, and IT pros know better than to go mission-critical with a first releast of *any* OS. The whole business of unsupported, unfinished, unknown, unstable ingredients that forces a soup-to-nuts upgrade cycle is made even worse when the non-compatible nature of that upgrade is not (at the very least) advertised clearly beforehand. It brings to mind the two-year life cycle of MS Office 95 and all its file formats. Educated users should be equipped to make qualified decisions about upgrading from a top-notch 6.2 to a pre-release quality 7.0. How difficult does this make the job of Linux professionals who have to explain why their bosses should *not* respond immediately to the hot new release from Red Hat? Red Hat's not Microsoft, never intended to be, but the unidirectional nature of the development of key components is a big pill to swallow, especially when they're not Unix industry standard.
No offense intended, but we're not just "wrong", sir. We're concerned.
John Beamon, RHCE
-j
The legacy software industry is built on the proprietary binary-only model where not only does the user not get the source code he needs to make changes, but worse he receives the product under a license that essentially says that if you make any improvements to the technology you are using, if you solve a bug that is causing your systems to crash, or add a feature that your users or customers desperately need the vendor can have you thrown in jail. (If you don't believe me, just read any shrinkwrapped software license). This kind of business model, where the customer is completely beholden to his supplier exists in no other industry in any free market that I know of. It harks back to the old feudal systems of 12th century Europe.
This is what it is all about, folks ! The only thing to get upset, serious and exited about !
This is the forest, which you can't see for all your little quarrels about bugs, premature releases, jealousy and greediness about how much a commercialized corporation RH has become (oh, oh how naughty).
I am a somewhat older lady and my brain is starting to fall apart, but I can still see the forest, I believe.
Want to hear a joke ? I had open source code apps on my Linux box, before I even knew that software could come any other way than "open sourced". Yup. I remember very well when I opened up for the first time a Perl program file and thought it looked like a nice chinese-like wallpaper pattern. Well, my heartily unwelcomed questions I posted to my consultant, made me understand, believe me, click by click. :-)
I am the most loyal open source distro whore there is, in spring I feel like redhaddish, in summer I need slackware, in fall I cuddle up with Debian and in winter I just can't live without my SuSe-chen. C'est la vie. Chaqu'un a son plaisir. :-)
So, don't pollute the atmosphere. I want open source code companies to be successful, a lot of them, small and large, and I want to be able to read your code the way I can read chemical formulas, math formulas and study biochemistry.
I want transparency of what is going on in this technology in a way, that any person can learn about it and influence it with its own input.
I think Mr. Young has all the reasons to remind you all about it. Your in-house dirty laundery washing on a public forum like slashdot may be an honorable effort to prove your sacred freedom, but it would be quite a dumb thing, if you end up throwing out the baby with the bathtub's water.
You forget whom and what you are up against. And you have by no means won the battle. Even though open source code is IMHO an absolute necessity, so far you guys haven't found the code solution yet, which would make each copy/clone of an open source code program sellable the same way the proprietary software industry can sell their's.
Suffice to say that's all there is to it in the end. May be you all can survive on service-based open source business models only, but one needs a LOT of faith, stamina and stubborness to hold on and through with it. RH has proven and stated that they are determined to do so. Why not just be grateful for it and that's it ?
Only now have they started their journey towards fixing the nefarious bugs in RH7.
As has been said before, the release of 7.0 by RH was quite probably a marketing decision. Was it a good one as far as the programmers/linux geeks are concerned?
Probably not.
But that doesn't matter. What matters is that they are correcting the existing issues in a timely fashon. Yes, I agree, some of them shouldn't have been in there at all, but that's life.
As the old saying goes, when life gives you lemons, make lemonade.
And as the new saying goes, when RH gives you a buggy kernel, patch it, and recompile. Just let everyone else know what you did.
I like you, Stuart. You're not like everyone else, here, at Slashdot.
Most people don't fully understand why Microsoft is evil. It seems to me most Microsoft bashers are young techies, trying to vent their frustration at the most obvious target. I'm not arguing that Microsoft's products are good, or that the company isn't evil reincarnate. I hate Microsoft as much as the next guy. But it is most important to understand why Microsoft is evil (and the reason isn't because they make crappy software). Given this understanding, it's also much easier to see why RedHat will never become Microsoft of Linux world. The way I see it, Microsoft's products aren't targeted at the technically gifted. The programs are buggy and performs poorly. Despite the faults Microsoft's products are perfectly suited for most of their users, e.g. secretaries, middle management, pre-meds, etc.) Normal users don't mind rebooting every day or two, or having some unknown, cutting edge innovations muffled. They are happy simply having 90% of the things working 90% of the time. What makes Microsoft evil is the fact they want to perpetrate mob rule using their size. They try to corrupt open standards, stifle competition, and overall do its best to ruin the day for those of us who want rapid innovations, fierce competition, and to be at the technical edge. Microsoft was successful because they provided what the customer wanted, and now that other companies threaten to outdo Microsoft, it wants to use its size to compete, rather than the quality of their product. Now this becomes a problem for normal users and techies alike. Microsoft's actions ruin innovation that the techies crave, while at the same time keeping the prices high for normal users. However things are different with Linux, RedHat can't become a Microsoft because Linux is a GPL OSS. Unlike Microsoft RedHat cannot maintain tight control over their product. No matter what RedHat does, it is simple (relatively speaking) to create another distro of Linux, that is perfectly compatible with RedHat. No matter how big RedHat gets, they won't hold the monopolistic power held by Microsoft. Also remember that the "normal user" for RedHat is highly knowledgeably, technically savvy, unwilling to deal with the frustration of Microsoft-like bugs. So the bottom line is if RedHat makes crappy products, they'll simply go out of business, rather than dragging the industry down like Microsoft.
To abandon open source is simply not in our customers interest and hence not in Red Hat's financial interest. Abandoning Open Source would also probably force you into developing your own Unix(like) environment; something your business model couldn't withstand.
-- Hob - Java Spectrum Emulator
Um, what if the standards set by the dominant distro happen to be retarded? One of the reasons I don't use RedHat is because of the awkward standards they try to set.
Microsoft makes the dominant operating system. Shouldn't we all just use Windows?
Intel makes the dominant PC processor. Shouldn't we all just use Pentium IIIs?
Nike is the dominant producer of athletic shoes. If we oppose their labor practices, we should just sit down and keep quiet about it, because they're number one, and we should buy their shoes.
McDonald's has more fast-food restaurants than anybody else in the world. What do we need Burger King for?
Coca-Cola is king of the soft drink industry. RC Cola should just give up.
--
$x='S24;r)>63/* h@<5+oZ)32"5cz';$me='phroggy'x$];
$x=~y+ -xz+\0-Tx+;print$_^chop$me for split'',$x;
I felt people were making the comparison by the stereotype of closed source projects shipping whenever it was in the best interests of the company versus open source's historical "when it's done" ship date. I think a lot of people consider Red Hat 7 to have been knowingly shipped prematurely. I'm not saying I agree, but that is what I gleaned from the discussions that have unfolded on slashdot.
We're still with you, Bob. We believe Red Hat will stay open source and give us power to control our own software. But a lot of people felt profit was driving this release, and that's a small step down the road to the dark side.
What did you eat today? http://www.atetoday.com/
I agree to a certain extent. One thing that pisses me off is watching all of the duplication of effort in the Open Source Community. For example, if all of the groups out there working on accounting projects would combine forces and agree on technologies (Perl, Apache, PostgreSQL, RPM, etc.), we could soon have an Open Source, modular, scalable accounting package comparable to something like Great Plains Dynamics. If you don't agree with me, browse around sourceforge.net sometime and look at all of the groups working on similar projects. Imagine what would happen if you combined some of those projects and focused that productivity into one kickass line of Open Source applications!! Don't re-invent the wheel!!
Now... I'm not saying that RedHat should have full reign of Linux standards. However, those of you who have read the book Net Ready know that there are four pillars to success in the modern E-conomy: Leadership, Governance, Competencies, and Technologies. RedHat is providing leadership for much of the community. We should be using the Open Source model not only to support them with the other three, but also to help them with a good system of checks and balances.
Bob's got some good points, and I think we should stop beating RedHat down and instead stick out a hand to pull him back up. Face it, we're all in this together. We might as well stop pointing fingers and try some "I'll scratch your back if you scratch mine" instead. Then, and only then, will we truly be on our way to world domination.
Personally, I'll be installing RHL7.0 as soon as I can purchase a copy. Why purchase you may ask? Because RedHat, for all they've done for the community, deserves $29.95 of my money.
--Gunfighter
-- Stu
/. ID under 2,000. I feel old now.
What they did wrong was failed to consult the gcc steering committee to get approval for a new stable so that other venders can also use it. They also failed to modify the contact numbers and emails for the compiler such that it directs problems to RedHat rather than the Gcc group. Third, they failed to roll back those changes to the C++ ABI so that their version would be able to generate code libraries which could be used across platforms. This is bad because it makes shipping C++ libraries more difficult and error prone. Some venders may only chose to ship RedHat versions and thus users of other dists are left in the cold. Also they placed the burden of handing the library problems on the free software developers such as myself without a lot of notice. I only test my software against known stable compiler versions and thus this release caused me some problems.
With all respect to Bob Young, I think that RedHat did make some mistakes. He certainly doesn't deserve the horrible Microsoft comparisons and other flames. However, on issues like shipping snapshots I don't agree with him. Summarily declaring all people who disagree are wrong with him seems quite arrogant. I think RedHat deserves to hear at least on that issue what mistakes they made and where they miscalculated.
Further, I also disagree with RedHat on the production of Inti rather than supporting the GNOME's current C++ binding, Gtk--. That like shipping a version of Gcc which is not compatible is divisive and intrusive into the workings of the Free Software community. A great many people did not see cloning an existing free product nor moving Havoc away from gtk+ development as good for the community. Obviously, RedHat felt it was in their best interests. I do hope that this situation wakes us up to the fact that what companies do is best for them and not necessarily the best for everyone. After all the point of a company is to make money. We should praise a company that does something in our best interest, and point out why we won't support them when they don't. After all it is in a companies best interest to release as little information about what they will support as possible to maintain competitive advantage. At the same time this is really against our interests as free developers such as myself have little chance to fix the bugs that shipping a snapshot compiler causes. However, flaming them like some people have choosen to do helps no one. Constructive criticism is a better approach.
--Karl
That's interesting. I find most people that bash Redhat are people that have been using Linux or BSD for a long time, way back when the kernel was in its late 1's and early 2's and before. I gave up on Redhat after 4.2, and from all reports, it's a good thing that I bailed before 5.0. RedHat has always been the leader in irritating bugs, and moronic applications that are supposed to help you administrate your box. It's the newbies that *use* Redhat because it's the biggest name in Linux right now.
Frankly, I think Redhat is giving linux something of a bad name. They're successful, sure, but their bugs always get turned into 'bugs in linux' by the press and the M$ cronies.
I'm grateful for the money that they bring in so that Alan Cox and folk like him can devote their time to the OS, but they have to do *something* about their buggy releases. I can't remember the last time anyone said 'Slackware? That buggy piece of crap? Forget it, I'm going to Redhat.'
So, to some up, I'll repeat something that my friends like to use (stolen from Something Awful):
'Use Redhat Lunix! Lunix si moar fastar!'
Generally, stuff that doesn't compile is one of three things -- the compiler (possible, but unlikely - I think there's only one confirmed this-code-is-valid-but-doesn't-compile bug), glibc (very possible -- make sure you apply the update to at least 2.1.94-3), or the code is broken with respect to the standard, and /will not compile with gcc 3.0 when it's done/. The most likely of the three, especially if c++ is involved, is the latter.
-30-
On the other hand, there are some areas where competing standards don't really hurt folks too much, given that the standard is well documented.
I do, however, like the idea of Red Hat taking charge. At worst, given the manner in which the "community" works, if the standard they attempt to implement ain't no good, it will more than likely get ignored.
----
"A fool does not delight in understanding, but only in revealing his own mind."
Five years ago, we had the same problem, only in a different fashion. Then, the debate was over libc5 versus glibc. At that time, RedHat was seen as evil because they 'forced' everyone to use glibc with their distribution. After all the flames subsided, glibc came out on top, and the world went on its merry way.
Now, granted, this isn't exactly the same thing. GCC 2.96 isn't necessarily the best choice to maintain compatibility with anything else out there. (And that's an understatement.) But without a leader, no decisions get made.
Maybe the bigger question is, should RedHat be that leader? RedHat seems to me to be the distribution taking the bigger risks and trying to get the biggest lead to show 'how far Linux has come'. So they include patches to the kernel that aren't in anyone's tree yet - AFAIK, that hasn't been a big problem to date. They use bleeding edge software they deem to be stable - hasn't seemed to backfire yet, current issue excepted.
It just seems to me that people criticising RedHat over this decision haven't taken the time to realize RedHat's overall contribution to the community, in terms of writing software, hiring developers, and taking the lead to bring Linux to the average Joe.
If you can't stand it, there are plenty of other good distributions out there that you can use. Slackware, Debian, Mandrake, Stampede, to name a few. Just because RedHat makes a decision you do or don't agree with doesn't mean they are the final word. Plenty of other people have worked long and hard hours to bring what they wanted to the table; try their solutions out.
people not knowing about the -n switch for ping (if anyone at redhat reads this PLEASE FIX PING, what it does is unexpected and has sent many people looking for deeper problems in their netoworks)
ping -n has behaved exactly the same for 15 years on every operating system except Windows.
Tell Microsoft to fix *their* broken ping, and RTFM.
-
Engineering is all about tradeoffs. If you're designing an airplane or the power system for a hospital, you can't afford any errors under any circumstances, so you over-engineer everything and spend most of your resources on the debugging to make absolutely sure there are no bugs.
If you're writing a word processor, you don't want to lengthen your development cycle by a month to fix a problem that only affects 1 in 100000 users. You don't want to double the price of your product so you can chase down the last few bugs that causes poor formatting of certain files.
In the tradeoff between bug-fixes and cost/time, you choose based on how mission-critical an app is. You'd be crazy to use Red Hat 7 to run an airplane. But if you're running someone's desktop machine, it's just fine.
It's just not possible to iron out all bugs without astronomical cost. To make that the standard would cripple the industry and make us wait twice as long for already-delayed software. In most cases I'm willing to accept buggy software with new features. If you aren't, stick with an older version. But don't demand that the rest of us be prevented from seeing a release until all bugs are released to your satisfaction.
I trust most Slashdot readers to see through that kind of stuff.
/. readers to have half a clue...shudder
whoa. I have never seen such misplaced trust.
...trusting
The basic sleazeware produced in a drunken fury by a bunch of UCBerkeley grad students was still the core of BIND. --PV
... except that it happened in 1975. And Bill's letter read:
I don't think that calling an entire community as thieves qualifies as "disagreeing"...
In short, if Bob wants to quote Levy's Hackers, he at least should do it right!
To the editors: your English is as bad as your Perl. Please go back to grade school.
I dunno . . . I'm not so keen on the idea that the dominant powers-that-be should make standards, and the rest of the world should then follow.
If there is a dominant vendor, whatever they do becomes a de facto standard. I'm not too keen on it either, but that's reality.
What I am keen on, though, is for that dominant vendor to say "Here's our product. It will become standard. Here's the source code. Copy and reimplement to your heart's content."
cut and paste dosn't count!
---
"...silence is a dangerous sound."
Okay, Bob, let's deal with this issue, then. You say that those of us that disagree with the shipment of gcc 2.96 "would be wrong". Even though the GCC Steering Committee was strongly opposed to this, and that its existance in Red Hat 7 may cause all sorts of compatibility problems.
// optimized Hamlet
I want to know -- what's the upside of 2.96?
---
question = '\FF';
--- question = 0xFF;
Including alpha components??? If you think this is the first time that RedHat or any other distribution has included an alpha-quality component, then you haven't been paying attention. Right now, I count 133 packages on my system with version numbers starting with "0."!
They included a compiler that has much better C++ support. If anything, this was done precisely to help developers, and "Joe every day windows user" will never even notice the difference. Goodness, what world are you from? Go to their FTP site. They have these things called "source RPMs". They have instructions for installing them. When you install them, you will see not only the pristine source code for the component, but also the individual patches that they apply to them. Debian is a wonderful distribution, so have fun! I certainly hope that the 2.2.17pre6 kernel that they provide doesn't give you any trouble!I'm a leaf on the wind. Watch how I soar.
I wouldn't ever compare RedHat to microsoft. I don't particularly like RedHat's distribution, bu that's just personal taste.
I Do know that last night while hanging out on IRC there we gobs of perople having problems with things specific to redhat.
Stuff not compiling on rh7, people not knowing about the -n switch for ping (if anyone at redhat reads this PLEASE FIX PING, what it does is unexpected and has sent many people looking for deeper problems in their netoworks), and linuxconf in general munging stuff.
I applaud redhat for making a mostly easy but flexible distribution... but please try not to break things that people depend on to work (ping, gcc, etc.)
Consider that Red Hat do engage is a policy of encouraging proprietary vendors to release their software "for Red Hat Linux" offering binary products only.
This only appears correct; it's not. Red Hat engages in a policy of encouraging proprietary vendors to release their software for Red Hat Linux, yes--but I suspect they also encourage people to consider releasing the source to their software as well, if not going the full open-source route.
The real incorrectness comes from the implication, that Red Hat encourages vendors of proprietary software to only ship for Red Hat Linux. Nothing could be further from the truth. Red Hat encourages vendors to make sure that their software works with Red Hat, which is pretty sane business advice--Red Hat has an extremely large market share.
Vendors will (often) choose only to officially support one distro, because QA testing is a long and costly process. Making sure that their software also works on Debian will cost them a small fortune, and the Debian market is small in comparison to the Red Hat market.
If you want to blame anyone for contributing to the non-support of non-RH distros, blame the software vendors and not Red Hat. Red Hat only encourages vendors to support Red Hat (a policy which is as unobjectionable as it is eminently sensible); they don't encourage vendors to get locked into a Red Hat solution (a policy as objectionable as it is monopolistic).
The former is the Red Hat way of doing things; the latter is the Microsoft way of doing things.
Oh, and by the way--I run Debian, and I still like Red Hat a heck of a lot. There's an awful lot of software on my Debian system which came from Red Hat, and one of the books on my shelf is Linux Application Development, written by some Red Hat guys.
I bought the Deluxe version of Red Hat 7.0, even though I'm a Debian user, just to say "thank you" to Red Hat for the software they've contributed to my Debian system.
What's the problem here? If you don't like it, don't shop there.
Well, except that I think differential pricing is illegal. If it isn't, it is at least unethical.
And no, not everything can be solved by "if you don't like it, don't shop there". Hey, maybe I don't like the practices of Archer Daniels Midland, or Monsanto. Try and buy food that *hasn't* passed through their hands. You pretty much can't. How can you tell? You can't use that excuse against monopolies. And while there are many corporations out there, many are equally as bad as the others, so there is effectively no "choice" to be made.
It's a crime that corporations are treated as individuals under law, because in many cases they don't have to hold up to their responsibilities and are unnaccountable for their actions. Yes corporations and businesses should be allowed to exist. But it should *also* be understood that a corporate charter is a mandate given on the good graces of the hosting society, which can pull it when the corporation does not behave well.
And for the record, I sympathize with many of the Libertarian ideas, but think that "pure" libertarianism verges on irresponsibility. I believe government can and should play a positive role in providing fundamental services to citizens, if it is architected to resist corruption, and open to a real democracy. I feel the Green party, and Nader's candidacy, is the best of many worlds, by mixing some good libertarian ideas, a passion for social justice, and a good helping of civic responsibility and "daily democracy". That's why I'm voting for Nader.
It's 10 PM. Do you know if you're un-American?
freedom & personal responsibility good, serfdom & tyrannical control bad, :)
freedom & personal responsibility good, serfdom & tyrannical control bad,
freedom & personal responsibility good, serfdom & tyrannical control bad,
freedom & personal responsibility good, serfdom & tyrannical control bad
...
I installed the Inti class library [...] because the gtk library that got installed [...] is a devel one that's prone to crashing
Inti is not yet beta code. Neither is the gtk version it uses. And that's quite clear even from the packagename (gtkbeta).
Red Hat Linux 7 also includes gtk 1.2.8, which is the most stable gtk release yet, and that's the version of gtk that's used unless you want to play with experimental stuff.
This message is provided under the terms outlined at http://www.bero.org/terms.html
There has been a problem (check the Qt source RPM for our patch).
We've had a close look at it, and we think it's non-standards compilant code in Qt.
With the patch, Qt works without problems.
Writing this from Konqueror 2.0RC2...
This message is provided under the terms outlined at http://www.bero.org/terms.html
...RedHat Linux is something that his company has created/assembled - and shared with the world. Only to have some people say "you suck".
His point is that, any innovation is going to have its faults, and that they have to be worked through. RedHat publishes publishes its distro's faults to the world - something that MS and Sun do not fully disclose. That data is being used to say compare Redhat to MS and Sun - an inherently flawed concept.
Now, comparing RedHat to MS because they are trying to make money from Linux is also wrong (IMHO). RedHat gives a lot back to the Linux community - and heck, they even employ Linux geeks to do just want they want to, add to Linux and open source software.
People are too willing to throw stones at RedHat - helping them find the bugs, and correct them would be more in the spirit of Open Source, and definitely more productive.
BlackNova Traders
Doug Alcorn
Well it appears that Red Hat has been hit by the growing popularity of Linux.
/., the news groups, a web site etc. So everyone can make the changes themselves if they want. Then Red Hat might even be a leader in the community and post the info or code to their own site for users to get to. Or someone else might just find a new company that does this for them.
There was a time when this would have been a non issue. If you didn't like what the Distro contained, you'd replace it yourself. Maybe compile some source code and stick it in, or install a new binary. It would be no big deal, something everyone did to use Linux.
With the growing popularity of Linux, the user base may no longer have the time or skills or both to do this and some definitely don't have the tolerance for any need to do it. Is this a slam against users? No, just recognizing the fact that it's not just the type or kind people it used to be.
All the rants, open letters, diatribes, posts, etc, against Red Hat certainly starts to look like the reactions against Microsoft. To that degree it starts to look like Red Hat is the next MS. It's not all their fault, it's the demands of all the users. The can certainly fall into Microsoft's error of trying to be all things to all users.
This is still Linux though. Everything is still open, the code hasn't been hidden. We can still change it ourselves if we need or want to.
Why not stop all this stuff and start acting like a community of users again. Someone replace gcc 2.96. Then post the code or changes to
We can act more as a community and realize, on all sides, that this won't be the last time this happens and be better prepared to solve the problem when it arises again.
I noticed that you used the term "Slashdotter" in your letter several times. This is incorrect.
The correct term is "Slashbot".
Please take note of this in your future correspondence.
Regards,
AC
However I think that there is something in the worries expressed by many about Red Hat's decision to include a gcc that the gcc people have publicly stated is not up to the standard of an official release due to incompatibilities.
Consider that Red Hat do engage is a policy of encouraging proprietary vendors to release their software "for Red Hat Linux" offering binary products only.
Now consider what happens should one of those binary only products is compiled against Red Hat version 7.0 and the product makes use of C++ in it's development.
That product will then only work on either Red Hat 7.0 or another platform that chooses to discard the views of the GCC steering committee.
Red Hat is the biggest player in the Linux market. They control the biggest share of the biggest market for commercial Linux use and they know it. By including "GCC 2.96" in Red Hat Linux 7.0 they are creating a situation of vendor lock in for the users of any proprietary software that is released "for Red Hat Linux" and compiled on the latest version.
I am sure that this is not a deliberate policy at Red Hat - were it deliberate it would certainly suggest a certain leaning towards the position of companies like Microsoft - however the effects remain the same: vendor lock in.
For Bob to reply to such critisisms with "You're wrong" without offering any explanation suggests arrogance. That is the kind of thing one expects of Microsoft. Red Hat need to offer a better explanation than "you are wrong".
Dear Bob,
I admire your company greatly and have had very productive relationships with Red Hat and Cygnus engineers going back many years. I am a member of the GCC steering commitee. I wish you nothing but success.
However, you do have a problem with openness that you are not acknowledging. There is one sense in which your practices do resemble those of Microsoft: your practice of keeping outsiders in the dark about upcoming plans that will affect them. To be specific: your management ordered its employees, including those who were members of the GCC Steering Committee, not to discuss anything about your plans for Red Hat 7.0 with other members of the committee. Advance discussion could have led to improved quality in 7.0, better relations with the outside developers you depend on, better planning by your customers and a whole lot less anger against you.
Joe Buck
They're shipping 2 compilers because their userland compiler doesn't work
Untrue. It works almost perfectly. There is, so far, one known problem that fails to compile code that is ok (and that's currently being fixed).
It can't compile 2.2.x kernels because of bugs in the kernel code (ask Alan), that's why the two compilers are needed.
This message is provided under the terms outlined at http://www.bero.org/terms.html
Please send bug reports on these, including sample code.
FWIW, all of Red Hat Linux 7, including Powertools and the Extra CDs for the European edition have been compiled with 2.96 using -O2 -mcpu=i686 -march=i386 without problems.
KDE 2.0 is completely c++, doesn't cause any problems with 2.96 -O2 -mcpu=i686 -march=i386.
This message is provided under the terms outlined at http://www.bero.org/terms.html
Quoted from Bob's article:
But if you don't like something about Red Hat Linux you don't have to use that feature or function.Hi Bob. My name is Lawrence Wade, and I have root on four Red Hat Linux boxes.
I'm still a relative Linux newbie. Though I was on the Internet (Arpanet) in 1988 and am therefore very familiar with UNIX heirarchy and commands, my first Red Hat Linux system was my first attempt at administering a UNIX(-like) box.
So, looking around, everyone told me how wonderful Linux is, and how light and easy on resources it is.
At the time, Red Hat 6.0 was current, and I installed it on a 486DX2-66 that I had kicking around. (Linux doesn't need much hardware, right? One would expect that the geniuses at Red Hat wouldn't expect that knowledgeable computer users are going to give a new OS a spin on their main computer, right? One would *think* then that a 486 would be a reasonable place to try out Linux.)
My first problem came when the installer detected that I had a monochrome VGA monitor, and set my text to the same color as my background. I plugged in a color monitor, and still couldn't read it. A reboot with a color monitor on, then a swap back to the monochrome monitor after booting, and the installer was still legible.
Next thing was, RedHat 6.0's installer asked me if I had PCMCIA card slots. This was a VESA-bus 486. I indicated NO.
The installation continued, and then finished. I restarted the machine:
"LI"
Stuck in the rescue disk, booted off that. Eventually found out about the LILO >1024 cylinder BIOS issue. Oops. Not Red Hat's fault, sure, but new users don't understand enough to distinguish that.
After I finally got the machine to start up, the machine hung at "Bringing up PCMCIA services". Still being a complete neophyte at the time, I had no idea how to go and kill that from the machine's startup. You'll note again that when the installer asked me if I wanted to install support for PCMCIA services, the response was probably sent to /dev/null; the installer apparently did what it wanted to, independent of my input. It took instructions about as well as my cat. With a cat, that's cute. With a computer, it's not.
Frustrated to all hell by this point, having wasted a weekend farting around, I took my RH6 disk and threw it across my home office. It landed behind a desk. Windows 95B went back onto that old, occasionally-used 486.
A few months later, I decided to get DSL internet service, and there was no way that I was going to use Windows for my gateway/firewall. So, I hit the Linux websites and got a list of supported network cards.
I rooted around under the desk, and found my RH6 disk under a dustbunny and an empty coffee cup. This time I knew that the LILO bug really only affected 486-vintage machines. I had a Pentium 133 ready to go. With 2 identical and supported ISA ethernet cards installed.
I'm not an idiot. First off, these network cards were set properly so that they didn't have any conflicts. Even Windows 95 was happy with them. Yet, every time RH6 tried to use the second card (eth1), a kernel panic happened. This continued until, in frustration, I replaced eth1 with a PCI NE-2000 card. Immediately, things worked.
After RH6.2 came out, I put it onto that machine, and then swapped back to my original matched-pair of ethernet cards. The machine worked like a million bucks, and has ever since.
So, Bob, what is this, a rant without a purpose? No.
I run Red Hat because, from the perspective of a newbie, Red Hat makes the most sense. Information is readily available. And 6.2 has been very good to me.
And while I realize that you have shareholders who will lose interest if you don't frequently bring out new releases, I'm afraid that most people, upon going through what I went through with RH6, would decide that Linux wasn't worth the trouble.
All the advocacy in the world isn't worth squat if the impressionable Windows user goes out to the local software store and buy a copy of Red Hat 7 to discover that nothing works as it should.
It undermines the Linux movement as an alternative to the scourge of Windows. And while I'm very sensitive to the fact that you have to keep your shareholders happy, discouraging new users by releasing buggy software in colorful boxes only serves to hurt Linux.
We simply are not pursing a business model that bears any resemblance to Microsoft's, so just quit it.New users don't distinguish between distributions - they don't know the difference between SuSE, Caldera, Debian, Slackware and Red Hat. To them, Linux is Linux. If Red Hat is flaky ("and golly, they're a big name!"), then *all* Linux must be flaky.
So, don't be surprised when someone falls asleep at the switch or buckles to shareholder pressure, releases a shitty version of your operating system, and Linux users all around the world start to compare you to Microsoft.
No one asks for perfection. That's simply impossible. But even Windows 95 Upgrade was more stable, reliable and functional than RH6. If RH7 is anywhere near as bad as I'm led to believe it is from here, let the comparisons begin.
I run RH 6.2. I like RH 6.2. I look forward to when your firm releases 7.2; until then, I won't be upgrading any systems. And I'll keep on burning CD copies of 6.2 for my friends, with the explanation that Red Hat bowed to shareholder pressure and released 7.0 before getting it to work properly. <sigh> Sounds just like any product Microsoft has ever released.
Fire and Meat. Yummy.
I understand your frustration with this scenario - many of these bugs are not with your code but with others, but there appears to be frustration all around. I applaud you for publishing your errors and fixing them, as well. However, there are certain things I can't abide by.
I'll give you a personal example of what's griping me. I installed the Inti class library because I was looking for a decent object oriented framework under which to develop Linux/GNOME apps. I can't even compile a basic tutorial program with it, because the gtk library that got installed with RedHat 7.0 is a devel one that's prone to crashing, deprecation problems, etc. The error message advised me to go and get a different version of the GTK in order to compile a basic "Hello World" style program.
I'd understand it if I found a complex bug. Really, I would. But what I found was something anyone could have found by taking a blank machine, installing RH7.0 and then trying to compile a program that RedHat has on its website as a tutorial.
A packaged, paid-for distribution is NOT the place for devel code, for alpha code, use at your own risk code, strange new worlds of compilers that can't even build the kernel, etc. Make that an option if you wish.
The best way to continue in your business is to get out of the CD world entirely - except to publish one on ISO and at nominal cost as a baseline, and then have a subscription-based service whereby when I log on to my RedHat machine I'm advised of new developments and offers to upgrade given packages.
--- Jump!! Fire!! Bullet time!! - Lego version of the Matrix
Ok, This Is gonna get a lot of people pissed off. In My opinion Red Hat Is doing the right thing by taking charge. They might end up looking like the new microsoft of the linux world, but someone needs to. The standards that the dominant linux distro makes should be followed. I've made comments before about netscape giving in to IE's standards. Why Not? This would make everything a lot easier for the rest of the world. But Oh Well, Just My opinion.
Btw, Make A Reply to this, i really want to know other people's opinions.
I have found that those who bash Red Hat the most are often newcomers to Linux who have maybe been part of the community for a year or perhaps even less. I am a bit confused about why they do this, perhaps they think Red Hat bashing is cool. I have been using Linux since 1994 and I think Red Hat is a great company! They pay people like Havoc Pennington to work fulltime on GTK 2.0, Alan Cox to hack on the kernel and many many other. Also, since when does Slashdot represent the community? Most people here don't contribute anything, the core community has always and still consists of the various mailing lists (the kernel, debian, mandrake, redhat and many other lists). Just MHO.
It's always nice to see an open letter, and even nice to see that Red Hat are actually watching the flow of conversation on Slashdot.
When something goes wrong, it's so usual for someone to sit around and say 'not my fault', that it's nice for someone to sit down and honestly say 'We may have made a mistake, perhaps.. But we made decisions we believed to be right, and you're quite at liberty to disagree..', and actually take the time to read the arguments against, and weigh them up.
I guess there's no progression without controversy and dissent over the paths to follow, and somebody has to spark controversy.
Maybe I speak just for myself, maybe for many, but thanks Bob, for paying attention to what we have to say, and for letting us know that you are, indeed listening to our many voices.
Cheers,
Malk
Well, there is a lot here to digest, both in terms of what Mr. Young wrote, and in terms of the larger debate about RedHat. Personally though, I feel that there is, in fact, a lot of paranoia in the Open Source Community - and rightly so, for we effectively LOST our community back in the early 80's. So here we are, in fear of RedHat becoming another Microsoft, or at the very least, a Sun (which I find to be an infinitely more likely situation, and a totally different debate altogether). Well, I can say in all honesty that I have used RedHat's products before, and I have kept up with all the various claims against them - and as a Debian user, I can say that I believe RedHat has done NOTHING to deserve the kind of claims against them that people have levied against them.
Now, this issue is moot in my mind - Young, in effect, seems to be feeding the trolls more than anything else here. No one who had actually researched this debate would claim RedHat to be acting like Microsoft. Again, as a Debian user, I feel that there is a hell of a lot WRONG with RedHat's distro - for ME. This more of a personal taste though - and don't anyone dare tell me that Debian hasn't had it's share of crippling bugs.
In the end, this is all just incredibly silly for me, and I have finally made my decision about RedHat. RedHat is a company, like SuSE and the Turbolinux guys... they are selling a product. RedHat has supported it's product admirably, and though it isn't in my decision the product that I want to use, that doesn't mean that when something goes wrong we should all point our fingers and scream "REDHAT IS BECOMING MICROSOFT!!!" RedHat, as a company, had done a great deal of work for the Linux community and for the acceptance of Linux in places that in turn have made Linux a better OS (IBM, Dell, etc...). So I think we should back off and turn down our flamethrowers here - if you want to criticize RedHat, fine, but at least find a good reason.
Know ye not that ye are Gods???
Most of the criticism of RedHat seems to be coming from college students with a selective dislike for what they see as large, successful businesses. I say "selective" because these same critics also fawn all over corporately produced mass market products, like Mountain Dew, The Simpsons, and anything Star Wars related.
One great thing about Linux distributions is competition. Don't like RedHat? Buy SuSE or Turbo or Corel or Stampede or Debian or one of several others. If you want to jump on corporations, then there are some pharmaceutical companies and food processing companies that really need public vehement public criticism of their practices. But don't bother pointing the same kind of fire at RedHat.
From: bob@redhat.com
Dear Slashdotter,
"it appears that you're trying to write an Open Letter. Would you like some help in choosing defensive words and phrases?"
--
--
"It is now safe to switch off your computer."