I think RHEL got to where it is by being a good choice for students that was similar to the most relevant UNIXes. More total users with UNIX like experience should have been good for the UNIXes but I wouldn't consider any of them healthy.
Giving Universities labs of RHEL was a way to get more Intel based machines in and therefore undergrad students used Fedora.. Research students would end up with RHEL where I went based on costs to their department.
Debian does not work as a replacement for RHEL for enterprises.. Only Ubuntu LTS is an option but it is not as good..
First because enterprises need a company behind offering support contract.
But more important, many enterprises are slow moving and unfortunately Debian lifecycle is not long enough for enterprises because of that..
I have at least one customer that migrated from RHEL 6 into newer versions last year, RHEL 6 was released on 2011, regular support ended on 2021 and extended support goes until next year. So it is currently still supported under their extended support.
The Debian version from 2011 was Debian 6 (Squeeze) that had an end of life in 2015, and long term support ended on 2016, it did not had extended LTS.
Just on the lifecycle of RHEL 6 Debian released versions 6, 7, 8, 9, 10, 11 and 12 and with exception of the last two, 11 and 12, all other versions already reached end of life, additionally version 6 to 9 reached end of LTS, and version 7 reached end of ELTS, older versions did not had ELTS as far i could find.
Is is also worth noting that Debian LTS and ELTS are not handled by Debian security team, so anything after end of life would be a solid no from most enterprises.
Even Ubuntu is behind RedHat in this aspect, the oldest Ubuntu LTS release still supported by next year, when extended support for RHEL 6 will end, is Ubuntu 14.04 from 2014. Both Ubuntu 10.04 from 2010 and Ubuntu 12.04 from 2012, that is the last LTS release on or before 2011 and the following release, already reached end of life and support before RHEL 6 extended support ended.
(And there are many more organisations and individuals who support the OS as well.)
Otherwise, this harkens back to "nobody ever got fired for buying IBM", Microsoft's "Linux is Cancer" and their "Get the Facts" FUD campaign (See: <http://news.bbc.co.uk/1/hi/technology/3600724.stm>), and much spew from SCO against IBM during the former's lawsuit against the latter.
The fact that Red Hat are now IBM and the Circle Has Closed is only all the more ironic.
(Red Hat was in fact my own first Linux over a quarter century ago, though I switched to Debian shortly afterward after realising that the latter was far, farr, far easier to install and maintain, for reasons intrinsic to the Debian Project's orientation and both social and technical structure. Incentives are very curious things.)
> First because enterprises need a company behind offering support contract.
I used to think this too, but it is not as big a deal as you may think. I have supported linux in the enterprise since the late 90's. The only place that ever bought support was an Oracle shop and we had 5 support licenses for our production OEL severs.
My current company is 100% CentOS and has never had support. I was not there when the choice was made or I would have pushed Debian. Some of my colleagues came from an all Debian shop.
Why is SUSE so seldom talked about? I don't know much about the distro except it has been around a long time, has professional support, supposedly is popular in Europe, and isn't it an RPM distro?
I can't speak for SUSE's corporate use, but I use opensuse on my laptop, server, router, and vps. Very nice distribution. That's not to say that there aren't sources/moments of friction and/or concern between the corporate and community, but it comes with the territory I guess.
I use OpenSUSE a lot for both desktop and servers, and it’s great.
The biggest downside though is that it probably has the least friendly and helpful community of any Linux or BSD distro I’ve ever used. Questions on their forums are frequently met with hostility and derision, mostly from one particularly active user
Ah yes, you’re worried about open source so you should use Ubuntu, the distro that is pushing Snaps, the most open source unfriendly solution in probably the most open source part of the entire Linux ecosystem.
> It's not Red Hat's fault, I blame IBM for the company's recent Linux code licensing changes and all the misery it unleashed.
Even if we were to agree 100% that nobody at Red Hat wants to do this and it's all IBM's fault, surely we can fault Red Hat for selling to IBM in the first place? I mean, everyone and their dog predicted the decay of Red Hat once the acquisition was announced. Surely the fact that they still went through with it, and therefore these all-too-predictable consequences, is on Red Hat?
…and Whitehurst was pretty explicit that he left because the board gave the CEO job to Arvind Krishna[0]. He also forfeited a large stock award when he did that.
I don't understand what you're saying. I said it's on Red Hat that they decided to sell to IBM. I didn't say that the CEO personally made that decision, the board of directors is obviously the part of the company which has the final say in such decisions.
op00to is just pointing out that being a public company (like taking VC money) has consequences that can't be avoided. Those consequences include being acquired, if the offer is good enough. Everybody knew that IBM buying Red Hat was the end of the Red Hat that had existed up to that point.
But accepting the buyout wasn't the decision; the decision was to go public in the first place. Which, of course, was Red Hat's decision. So everybody is correct.
'Red Hat's open source rot' = they are no longer giving non-customers the RHEL sources for free, something that is completely within their right under the GPL, but it means I no longer benefit from those great free security and other patches republished by the rebuilders after stripping out the RHEL trademarks, so it's inconvenient for me.
RedHat has been trying to figure out how to both benefit from open source and also avoid helping others at the same time.. if you think of RedHat the OS as a combination of the linux kernel (worked on by many companies and individuals well beyond RedHat), tons of packages in userspace (of the millions of lines of code in these packages, probably a very very small fraction are from RedHat -- sure yes some important, some less important).
Examining the fact that RedHat in principle does some huge amount of testing of the packages and puts them all together -- lets think about that-- do they really? in many cases the upstream package maintainer(s) probably do most of that. So are they going to go build a ton of test vectors for each package and "test" them all .. I highly doubt that.. likely they have a set of tests that they do to try to hit a set of important things that "should work" and if a package happens to cause that to fail, then they roll it back. In the case that RH pushes a patch to the package maintainer -- does that mean they effectively "own" that software and people that want that updated version are trying to get stuff for "free"?
From the beginning people have questioned whether they should be able to hold back patches from other people -- in reality, they can try, but they also can't stop people from doing that same patch "somehow".. so effectively unless people can copyright a new patch to open source software and prevent others from doing that same patch, then they have no recourse.
Redhat should just lean really hard into the "support" side of things, and whatever margin exists with that, then fine. Sure, does it piss you off that other companies are making money off of your hard work, yes it does, but look at basically all of the other people who were ahead of you who you are also apparently ripping off RedHat..
So...which is it? Is Red Hat not really doing all that much work? Which means that people should be happy to use CentOS Stream, which is the free and open source upstream of RHEL?
Or are they actually doing some valuable work? In which case they are justified in charging money and providing source code only to their customers, under the provisions of the GPL?
Before replying, please go ahead and read up on the GPL and especially the parts where they confirm that it's completely OK to charge money for the software as long as you provide users the source code on request.
I'm familiar with the GPL -- one of the key provisions is that users can redistribute the code -- whether RedHat decides to release one version or zero versions of their code should not prevent anybody from redistributing that code when there is an associated binary.
Yes you can charge people for that binary and sell the code as well, but if somebody takes the code and redistributes it and even sells it, then they also have to allow others to distribute it -- its like a river of open source that everybody benefits from. Maybe I'm missing the magic sentence in the GPLv2 or v3 license that says something like "if your user takes the code you provided that was under GPL license and sells it, you can sue to stop them."
If RHEL succeeds in creating a model that other companies clone where because they provided valuable work they get to stop others from getting a copy of their source code, then the GPL license is basically dead.
> Maybe I'm missing the magic sentence in the GPLv2 or v3 license that says something like "if your user takes the code you provided that was under GPL license and sells it, you can sue to stop them."
You missed the actual events that happened and made up a narrative of 'Red Hat is suing to stop their customers from redistributing the RHEL source code'. They are not doing that, they are deciding to part ways with customers who do that. The customer still has the right--per the GPL--to keep distributing the copy of the RHEL source code that they got from RHEL, and no one is stopping them from doing that.
You’re ranting based on internet trolls. GPL doesn’t state that you need to provide source code in any particular form. You can ship a CD to folks who ask and be compliant.
They release all of the code for RHEL in git. The just stopped publishing SRPMs so you cannot trivially duplicate RHEL.
Exactly, so: fuck them. Everyone is not only completely within their rights, but doing what they are supposed to do.
Red Hat is saying "we don't care about people who use a rebadged clone of our OS and pay us" (hard to find fault with that) and open-source users who don't pay for Red Hat are saying "fuck a Linux distro that hides source code" (ditto).
And we'll see who was right, in the sense of actually making a good decision, within a year or two. It might even be "both". (Although I personally suspect that Red Hat effectively removing RHEL (including its not-directly-money-making clones) from contention as a mainstream OS will have worse long-term effects on Red Hat than it will for anybody else... but we'll see.)
I have zero problems with people who vote with their feet and walk away from RHEL over RH deciding not to share source code with non-customers. What I do find cynical and hypocritical is people making up all sorts of wild accusations with no basis in reality, like elsewhere in this thread someone accused RH of violating the GPL by suing customers to stop them publishing RHEL source code.
That's just one example, the online forums and social media are full of people making up all sorts of stories. The one thing that's clear to me is that people never actually bothered to read the GPL, or didn't actually bother to find out exactly what RH is doing. And they can't be bothered to while they go on their moral crusade. God forbid that facts get in the way.
I can say for certain that this started even before IBM took over Red Hat. All these decisions have been driven by Red Hat sales team in co-ordination with the upper management. They experimented with these ideas with the entrerprise edition of JBoss Application server (later called WildFly). There was a lot of internal friction when that happened and yet they moved ahead with similar decisions as what is repeated with RHEL today.
Autor, pleas... There is no RedHat anymore. There is only blob with tentacles not feeded and cut off when dry. And name is collected - absolutely no surprise when suddenly "RedHat" will start to sell dolls...
39 comments
[ 2.9 ms ] story [ 101 ms ] threadThe future is unclear for the RHEL/CentOS clones
Nothing has really changed for the companies that pay for RHEL. They do so for a good reason.
https://en.wikipedia.org/wiki/Fedora_Linux#History
First because enterprises need a company behind offering support contract.
But more important, many enterprises are slow moving and unfortunately Debian lifecycle is not long enough for enterprises because of that..
I have at least one customer that migrated from RHEL 6 into newer versions last year, RHEL 6 was released on 2011, regular support ended on 2021 and extended support goes until next year. So it is currently still supported under their extended support.
The Debian version from 2011 was Debian 6 (Squeeze) that had an end of life in 2015, and long term support ended on 2016, it did not had extended LTS.
Just on the lifecycle of RHEL 6 Debian released versions 6, 7, 8, 9, 10, 11 and 12 and with exception of the last two, 11 and 12, all other versions already reached end of life, additionally version 6 to 9 reached end of LTS, and version 7 reached end of ELTS, older versions did not had ELTS as far i could find.
Is is also worth noting that Debian LTS and ELTS are not handled by Debian security team, so anything after end of life would be a solid no from most enterprises.
Even Ubuntu is behind RedHat in this aspect, the oldest Ubuntu LTS release still supported by next year, when extended support for RHEL 6 will end, is Ubuntu 14.04 from 2014. Both Ubuntu 10.04 from 2010 and Ubuntu 12.04 from 2012, that is the last LTS release on or before 2011 and the following release, already reached end of life and support before RHEL 6 extended support ended.
[0] - https://wiki.debian.org/DebianReleases [1] - https://ubuntu.com/about/release-cycle
(And there are many more organisations and individuals who support the OS as well.)
Otherwise, this harkens back to "nobody ever got fired for buying IBM", Microsoft's "Linux is Cancer" and their "Get the Facts" FUD campaign (See: <http://news.bbc.co.uk/1/hi/technology/3600724.stm>), and much spew from SCO against IBM during the former's lawsuit against the latter.
The fact that Red Hat are now IBM and the Circle Has Closed is only all the more ironic.
(Red Hat was in fact my own first Linux over a quarter century ago, though I switched to Debian shortly afterward after realising that the latter was far, farr, far easier to install and maintain, for reasons intrinsic to the Debian Project's orientation and both social and technical structure. Incentives are very curious things.)
That's quite literally why Ubuntu exists… forked Debian with enterprise support.
https://ubuntu.com/support
I used to think this too, but it is not as big a deal as you may think. I have supported linux in the enterprise since the late 90's. The only place that ever bought support was an Oracle shop and we had 5 support licenses for our production OEL severs.
My current company is 100% CentOS and has never had support. I was not there when the choice was made or I would have pushed Debian. Some of my colleagues came from an all Debian shop.
Is there any downside to its fully free spins?
The biggest downside though is that it probably has the least friendly and helpful community of any Linux or BSD distro I’ve ever used. Questions on their forums are frequently met with hostility and derision, mostly from one particularly active user
Even if we were to agree 100% that nobody at Red Hat wants to do this and it's all IBM's fault, surely we can fault Red Hat for selling to IBM in the first place? I mean, everyone and their dog predicted the decay of Red Hat once the acquisition was announced. Surely the fact that they still went through with it, and therefore these all-too-predictable consequences, is on Red Hat?
[0] https://wraltechwire.com/2021/07/07/whitehurst-stepping-down...
Jim had to take Ginny’s offer to the Board, and they decided to sell. That’s life in the big city.
op00to is just pointing out that being a public company (like taking VC money) has consequences that can't be avoided. Those consequences include being acquired, if the offer is good enough. Everybody knew that IBM buying Red Hat was the end of the Red Hat that had existed up to that point.
But accepting the buyout wasn't the decision; the decision was to go public in the first place. Which, of course, was Red Hat's decision. So everybody is correct.
RedHat has been trying to figure out how to both benefit from open source and also avoid helping others at the same time.. if you think of RedHat the OS as a combination of the linux kernel (worked on by many companies and individuals well beyond RedHat), tons of packages in userspace (of the millions of lines of code in these packages, probably a very very small fraction are from RedHat -- sure yes some important, some less important).
Examining the fact that RedHat in principle does some huge amount of testing of the packages and puts them all together -- lets think about that-- do they really? in many cases the upstream package maintainer(s) probably do most of that. So are they going to go build a ton of test vectors for each package and "test" them all .. I highly doubt that.. likely they have a set of tests that they do to try to hit a set of important things that "should work" and if a package happens to cause that to fail, then they roll it back. In the case that RH pushes a patch to the package maintainer -- does that mean they effectively "own" that software and people that want that updated version are trying to get stuff for "free"?
From the beginning people have questioned whether they should be able to hold back patches from other people -- in reality, they can try, but they also can't stop people from doing that same patch "somehow".. so effectively unless people can copyright a new patch to open source software and prevent others from doing that same patch, then they have no recourse.
Redhat should just lean really hard into the "support" side of things, and whatever margin exists with that, then fine. Sure, does it piss you off that other companies are making money off of your hard work, yes it does, but look at basically all of the other people who were ahead of you who you are also apparently ripping off RedHat..
So...which is it? Is Red Hat not really doing all that much work? Which means that people should be happy to use CentOS Stream, which is the free and open source upstream of RHEL?
Or are they actually doing some valuable work? In which case they are justified in charging money and providing source code only to their customers, under the provisions of the GPL?
Before replying, please go ahead and read up on the GPL and especially the parts where they confirm that it's completely OK to charge money for the software as long as you provide users the source code on request.
Yes you can charge people for that binary and sell the code as well, but if somebody takes the code and redistributes it and even sells it, then they also have to allow others to distribute it -- its like a river of open source that everybody benefits from. Maybe I'm missing the magic sentence in the GPLv2 or v3 license that says something like "if your user takes the code you provided that was under GPL license and sells it, you can sue to stop them."
If RHEL succeeds in creating a model that other companies clone where because they provided valuable work they get to stop others from getting a copy of their source code, then the GPL license is basically dead.
You missed the actual events that happened and made up a narrative of 'Red Hat is suing to stop their customers from redistributing the RHEL source code'. They are not doing that, they are deciding to part ways with customers who do that. The customer still has the right--per the GPL--to keep distributing the copy of the RHEL source code that they got from RHEL, and no one is stopping them from doing that.
They release all of the code for RHEL in git. The just stopped publishing SRPMs so you cannot trivially duplicate RHEL.
L
Red Hat is saying "we don't care about people who use a rebadged clone of our OS and pay us" (hard to find fault with that) and open-source users who don't pay for Red Hat are saying "fuck a Linux distro that hides source code" (ditto).
And we'll see who was right, in the sense of actually making a good decision, within a year or two. It might even be "both". (Although I personally suspect that Red Hat effectively removing RHEL (including its not-directly-money-making clones) from contention as a mainstream OS will have worse long-term effects on Red Hat than it will for anybody else... but we'll see.)
That's just one example, the online forums and social media are full of people making up all sorts of stories. The one thing that's clear to me is that people never actually bothered to read the GPL, or didn't actually bother to find out exactly what RH is doing. And they can't be bothered to while they go on their moral crusade. God forbid that facts get in the way.
Autor, pleas... There is no RedHat anymore. There is only blob with tentacles not feeded and cut off when dry. And name is collected - absolutely no surprise when suddenly "RedHat" will start to sell dolls...