Will systemd make FreeBSD more popular?

[Carbide]

In reply to the question originally posed in the thread's title, yes and the other BSDs too. I had sampled the BSD world some time back. While I generally liked what I found, it was different enough that I went back to Linux in order to avoid some of the pain (read work) necessary to make a full transition. That and the fact that I found a Linux distro that did not have systemD and planned to stay that way. All was fine until the hammer dropped.

A few weeks ago, one of the IT journals (sorry, I don't remember which one) published an analysis of how systemD was affecting Linux in general and what we could expect in the future. The author postulated that systemD was mutating and grabbing functionality at such a rate that the hold out distros only had about one year left before they would be forced to capitulate completely or lose major features because of more forced dependencies on systemD.

A few days ago, the hammer fell the rest of the way. Poettering announced that systemD was now taking over all network functions. That was it. Linux is dead to me now. I wish the hold outs well. I hope they can win in the end, but have little faith that they will. I'm in BSD land to stay.

 

[walterbyrd]

I'd suggest to hide. If disappointed Linux devs came to "improve" FreeBSD (or other BSDs without a definite "that won't ever happen" strategy), we can all imagine how FreeBSD will continue. Everyone can be the next Poettering.

2016, the year of the FreeBSD desktop? For some of us, indeed. I'd miss it.

IMO: it is not the Linux devs that are ruining Linux, it's Red Hat. For years now, changes have been made to Linux which have no technological advantage, but will give Red Hat increased control.

I suppose that might seem like some crazy conspiracy theory, but if you look at it from the perspective of a $10 billion corporation which would benefit substantially by being able to control the Linux ecosystem, it may not seem as crazy as all that.

It is not the popularity of Linux that causing the problems, it's corporate interests. To be fair, those corporate interests have been beneficial in the past, but I think things are changing.

Were it not for Red Hat, Poettering would be an easily dismissed nobody.

 

[Carbide]

Were it not for Red Hat, Poettering would be an easily dismissed nobody.

I agree with that. Red Hat seems to also have managed to place its own employees into critical decision making positions within competing distributions and critical projects, which gives it the power to effect control over the direction of those distros and projects. At a minimum, Poettering and his cohorts are proceeding with the blessings of Red Hat, tacit or otherwise, if not at its outright behest and direction. As his employer, Red Hat is responsible for what he is doing in the Linux arena, including his conduct on mailing lists and forums.

 

[Beastie7]

The Linux ecosystem was historically broken to begin with, when you have two entities with two different approaches to the same ideology, chaos is bound to happen with people splitting off into different realms of the ecosystem. I honestly believe "GNU"/"Linux" just happened because both parties didn't have the momentum to complete their objective. Redhat sought to unify everyone (and they did actually), but Red Hat has always been a company since the inception of Linux, so of course they're going to have more control over the ecosystem given the cash they have. Hell, even canonical had the ability to do it, but they're too busy trying to be Apple. In retrospect, with Red Hat being Linux's biggest contributor, they have did a lot of good for advancing Linux as a whole though. I believe Red Hats greediness began when they acquired JBoss, and started on their huge "Cloud" fantasy trying to compete with Oracle/IBM/Google, and knowing the Linux ecosystem, the wild west doesn't care for Market trends.


Just imagine if 386BSD released in time, we wouldn't be here having this conversation.

 

[Cthulhux]

Of course we would, as 386BSD was started after Linux.

 

[drhowarddrfine]

The previous two posters need their history corrected but I don't have the time.

 

[Cthulhux]

The history is: There was no free Unix back in 1991, so Linus started developing Linux modelled after Minix. In 1992, free replacements for almost all AT&T code was ready.

If they had just hurried a bit.

 

[drhowarddrfine]

If they had just hurried a bit.

You are misinformed but, again, I have no time to go into it. Today is my birthday and I must appear in front of my millions of admirers.

 

[BSDBernd]

I forgot to answer the question of the thread. http://distrowatch.com/index.php?language=DE seems to suggest an opposite trend, at least when it comes to normal users, general citizens like you and me . The interest seems not to have grown, I regularly look on that site. In fact, popularity seems to get down. The same I get when I look at google trends. Apart from the frightening diagram http://www.google.de/trends/explore#q=FreeBSD , even when you look at more recent time spans, f.e. the last 90 days, there seems to be no upwards trend. For the Distrowatch thing, I believe a simple explanation could be that the Linux folk try to get more reliable when it comes to Desktops, Laptops, mobile devices, and they care about Clouds etc. Ubuntu has these long time support versions which seem to run stable. I know several people who run these and they seem not to have problems with stability. Also these more stable Linux distros care about power consumption, boot up times, about getting their OS on mobile devices, about the things Hubbard mentioned in his talk.
May be the upwards trend will come, but it seems not at the moment to come.

 

[Beastie7]

You are misinformed but, again, I have no time to go into it. Today is my birthday and I must appear in front of my millions of admirers.

You've replied twice, and yet you have no time to enlighten us... interesting.

 

[scottro]

Meh, it's easy enough to track down, I think. I don't really know the timeline, nor care all that much, but AT&T did Unix, Berkeley did some stuff, there was court stuff.

http://en.wikipedia.org/wiki/FreeBSD#Birth_of_FreeBSD

Now what that has to do with Linux, I dunno. Linus Torvalds made a kernel, somewhere along the line it got mixed in with the Stallman stuff and you got a Unix like system.
I don't care enough to google which was first, but it should be a quick look.
Anyhoo, during the lawsuit time, I think, more people started using Linux or LInux mixed with the GNU stuff. Gnu is not Unix, part of Stallman's whole thing about software being free.

So, how does this affect all of us. IT HAS ABSOLUTELY NO MEANING LEGAL OR OTHERWISE to what we're using today, save for a few, Oh, if this, then maybe that. I'd have to to spend more effort to google the time period than I care to do. I can affirm that either Linux or the BSDs came first, and the other one came second.

I'm a very selfish person. I even have a vid up--the situation is that the man and the woman are being given a head start before a drug lord hunts them as human prey. Here is the video.

http://srobb.net/selfish.mp4

So, has anyone job hunting seen a lot more FreeBSD sysadmin ads since CentOS (and RHEL) 7 started using systemd? I haven't, I've just seen people on mailing lists or forums saying, Gee, I'm going to FreeBSD.

Has Amazon, for example, in shock and disgust at systemd, or Google, suddenly changed their servers to FreeBSD or made plans to do so?

Anyway, this has, to be a bit trollish, begun to remind me of the 100 post plus threads on Linux forums about how Windows is dying. And yet, somehow, most people still use it. Shucks, I was watching Grimm the other day, where they probably have a product placement deal with Apple, and one of the main characters goes to check his girlfriend's iMac. You see the desktop and its Windows. Of course, on the Linux forums (fora?), many of the posters will be young kids posting about the end of Winblow$ and so on. But...

While there may be some users leaving Linux for this, and while it may even be the beginning of an exodus, I doubt it will have the effect that many of us hope it will. It may have the negative effect of making things harder to port, as it grabs more and more things to do, and I believe that Mr. Poettering has stated that he doesn't care about POSIX standards (but that might be FUD on my part--too lazy to google, see the selfish thing above.)


TL;DR
The history and which came first isn't going to matter practically. Whether some people leave Linux because of systemd or whatever, I don't yet see noticeable amounts of sysadmin jobs changing from requesting Linux to requesting BSD expertise.

 

[retrogamer]

I forgot to answer the question of the thread. http://distrowatch.com/index.php?language=DE seems to suggest an opposite trend, at least when it comes to normal users, general citizens like you and me . The interest seems not to have grown, I regularly look on that site. In fact, popularity seems to get down. The same I get when I look at google trends. Apart from the frightening diagram http://www.google.de/trends/explore#q=FreeBSD , even when you look at more recent time spans, f.e. the last 90 days, there seems to be no upwards trend. For the Distrowatch thing, I believe a simple explanation could be that the Linux folk try to get more reliable when it comes to Desktops, Laptops, mobile devices, and they care about Clouds etc. Ubuntu has these long time support versions which seem to run stable. I know several people who run these and they seem not to have problems with stability. Also these more stable Linux distros care about power consumption, boot up times, about getting their OS on mobile devices, about the things Hubbard mentioned in his talk.
May be the upwards trend will come, but it seems not at the moment to come.

That's the point I was trying to make, but you did a much better job of illustrating it. Big corporations that may not be a fan of Red Hat's current direction with Linux - say Google - have already created Android and Android derivatives that are completely free from systemd dependence, new Linux based OSes using Android's init could be developed into a server or workstation OS (Android is already pretty usable as a desktop OS now), to think they are suddenly going to take interest in BSDs isn't very likely. There was a great article about that (a workstation OS based on Android) I saw at http://www.dragonflydigest.com/ last year, but I couldn't find it again.

Anyway, if we leave out the giant corporations, who does that leave that might switch? The smaller company, or the Linux power user. They might switch over systemd at some point, but there is at least a 5 year window before systemd is going to be the only supported option in several major distros, I doubt that a huge number of people are planning 5 years ahead. Even then, look at the number of unsupported Linux servers out there. These are the companies that will switch over systemd? Doubtful, in my opinion. Now, ZFS, bhyve, Clang, OSS are very present reasons that you might switch to FreeBSD. That said, I wouldn't trust Distrowatch at all to gauge interest in BSDs, that is a completely unscientific ranking of popularity (based solely on DW page hits). I'd imagine most people who migrate to a BSD quit using Distrowatch with any regularity, since it's a Linux focused site.

 

[Carbide]

They might switch over systemd at some point, but there is at least a 5 year window before systemd is going to be the only supported option in several major distros, I doubt that a huge number of people are planning 5 years ahead..

Where in the world do you get five years? At the rate systemD is gobbling up functionality, in five years it will be the Poettering OS and Linux will be a shell of its former self. Users will have been forced to jump one way or another long before that.

 

[retrogamer]

Where in the world do you get five years? At the rate systemD is gobbling up functionality, in five years it will be the Poettering OS and Linux will be a shell of its former self. Users will have been forced to jump one way or another long before that.

As Oko pointed out earlier, - https://access.redhat.com/support/policy/updates/errata for Red Hat - meaning there is a supported RHEL without systemd until 2020.
Slackware-Current still uses sysvinit and will be supported for at least 5 years when it becomes the next stable release. Pat has not said (that I know of) whether Slackware will ever implement systemd, though I suspect this will be a situation like OSS and XFree86 where his hand is forced eventually.

 

[Carbide]

As Oko pointed out earlier, - https://access.redhat.com/support/policy/updates/errata for Red Hat - meaning there is a supported RHEL without systemd until 2020.

Yes, but that is mainly used by big budget corporate/government entities for servers. Individual users generally prefer something more up to date, not to mention free of charge. A close family member of mine administers Linux systems for a large organization. Red Hat is almost exclusively servers. Ubuntu and a sprinkling of other stuff is used on desktops. It is the free open source users who will decide the question posed at the top of this thread, and they don't use RH 6.

Slackware-Current still uses sysvinit and will be supported for at least 5 years when it becomes the next stable release. Pat has not said (that I know of) whether Slackware will ever implement systemd, though I suspect this will be a situation like OSS and XFree86 where his hand is forced eventually.

Forced is the key term here. It is instructive to go to distrowatch and look at the included packages listings for the various distros. Of the hundreds of distros, very, very few are systemD free. PCLinuxOS and Slackware are the only two that I would consider major league, and a lot of people would debate my inclusion of PCLinuxOS.

The bottom line is simple. With the caving of Debian and Ubuntu, the die has already been cast. And that was before the latest systemD power grab that was just announced, networking. It's over, Linux as we knew it is dead. Today, not five years from today. The only thing that is likely to reverse the trend is something like a blistering anti systemD rant from Linus (long overdue in my mind) along with some sign of a systemD free distro arising out of the Debian ashes at the Devuan project.

I'm not holding my breath.

 

[drhowarddrfine]

You've replied twice, and yet you have no time to enlighten us... interesting.

Yes. In between trips to various activities, while waiting for others, I managed to find a minute to surf here. It's now 2AM and I'll give you another brief moment.

Linus Torvalds wrote Linux because BSD was in legal turmoil with ATT over code BSD contained. As he wrote, and you can also Google for, he stated that if BSD was available back then, he never would have written Linux. Therefore, BSD was around but under hold due to legal problems. Working faster, as one of you two stated, would not have mattered.

I will let you Google for all that since I did not like your response but you can start here.

 

[Beastie7]

Yes. In between trips to various activities, while waiting for others, I managed to find a minute to surf here. It's now 2AM and I'll give you another brief moment.

Linus Torvalds wrote Linux because BSD was in legal turmoil with ATT over code BSD contained. As he wrote, and you can also Google for, he stated that if BSD was available back then, he never would have written Linux. Therefore, BSD was around but under hold due to legal problems. Working faster, as one of you two stated, would not have mattered.

I will let you Google for all that since I did not like your response but you can start here.

Yes, I'm aware of that. But the First "GNU/Linux" distro didn't come into fruition until around 1993 (SLS/Slackware), a little after 386BSD. You can't have a complete operating system without the GNU userland. That's what I was hinting at. But whatever. Let's put this one to rest shall we?

 

[ANOKNUSA]

Individual users generally prefer something more up to date, not to mention free of charge.

They also prefer not ever thinking about the internals of their operating systems.

With the caving of Debian...

The Debian developers did not "cave." It was put to a vote, as are all such major decisions. The alternatives were OpenRC and Upstart. systemd won. Repeatedly.

The only thing that is likely to reverse the trend is something like a blistering anti systemD rant from Linus (long overdue in my mind)

He doesn't care. At all.

 

[walterbyrd]

JMHO responses to the last few posts.

Whether some people leave Linux because of systemd or whatever, I don't yet see noticeable amounts of sysadmin jobs changing from requesting Linux to requesting BSD expertise.

I would not expect such a change overnight. It would take a few years, if would ever happen.

As Oko pointed out earlier, - https://access.redhat.com/support/policy/updates/errata for Red Hat - meaning there is a supported RHEL without systemd until 2020.

Again I am not sure that "supported" means practical to actually use.

I was using CentOS 6.5 (based on last Red Hat version before default systemd) up until a few months ago. The kernel in CentOS 6.5 is already about five years old, and I was coming across apps that would not install, just because of the age of the kernel. Technically speaking, non-systemd may be supported until 2020, but will stuff run on it? In particular, can we count on the latest versions Red Hat stuff like JBoss, and OpenStack, to run on it? I think "supported" may just mean security patches.

They might switch over systemd at some point, but there is at least a 5 year window before systemd is going to be the only supported option in several major distros

Systemd will be the default in practically all Linux by the time Jessie comes out.

As I see it, especially for corporate use, there are two dominating players: Red Hat and Debian. Keep in mind that Canonical is based on Debian, and OpenSuse is really based on Red Hat (OpenSuse denies this but OpenSuse uses the same package management, and the same systemd).

With Red Hat, Debian, Canonical, Oracle, IBM, and Novell all firmly behind systemd, I think systemd will be the standard by the end of 2015. Once systemd is the standard, Red Hat is firmly in control. When that happens, I expect anything optional about using systemd will be eliminated.

Slackware-Current still uses sysvinit and will be supported for at least 5 years when it becomes the next stable release. Pat has not said (that I know of) whether Slackware will ever implement systemd, though I suspect this will be a situation like OSS and XFree86 where his hand is forced eventually.

I like Slackware, but it has no real leverage in the corporate world. I read an interview by Pat, where Pat said that he had no plans to put systemd in slackware - but that it might become inevitable.

The only thing that is likely to reverse the trend is something like a blistering anti systemD rant from Linus (long overdue in my mind) along with some sign of a systemD free distro arising out of the Debian ashes at the Devuan project.

It might be possible that Linus' ego is preventing this. Linus may not want to admit how powerless he is in this. I hope Devuan works, but in terms of leverage, resources, and entrenchment, Devuan is hugely outmatched. We are talking about a $10 billion corporation, that already has 65% of the corporate Linux users, compared to what? Devuan may win out anyway, it will be interesting to watch.

The Debian developers did not "cave." It was put to a vote, as are all such major decisions. The alternatives were OpenRC and Upstart. systemd won. Repeatedly.

I have read some of the email list discussions. It seemed to me that the discussions were not about the quality of systemd, but rather: is systemd inevitable?

I get the idea that Debian is supporting systemd for the same reason that Libreoffice supports OOXML. Not because they love it, but because it is, effectively, forced on them by the dominate player.

Again, all JMHO, of course.

 

[ANOKNUSA]

OpenSuse is really based on Red Hat (OpenSuse denies this but OpenSuse uses the same package management and the same systemd).

They use the same package format. That doesn't mean the contents of those packages, nor the software to install it, nor configuration of software once installed is the same. And, what, are there other systemds I don't know about? Or are all distributions Red Hat now, because they all contain 20-odd megabytes of software developed by Red Hat?

It might be possible that Linus' ego is preventing this. Linus may not want to admit how powerless he is in this.

It might be possible that Linus has been possessed by demons summoned from beyond the realms of mortal men by Lennart Poettering, and is prevented by dark magicks from retrieving the mighty Sword of SysV in order to slay the evil systemd beast that threatens to destroy our villages and spoil our crops and defile our women and melt our ice cream cones. But such speculation contributes nothing to the discussion. He says he doesn't care, and that's all anyone has to go on.

It's fine to dislike systemd---there are valid gripes to be had---and discussions over its ramifications are fine as well. They're good, in fact---whether you like systemd or not, talking about how it may or may not affect your future is just the smart thing to do. But fallacious reasoning and random speculation and conspiracy mongering taint any and every attempt at a decent discussion on matter every goddamn time. Every time. The only reasonable thinking I've seen on the matter is a single blog post nobody of consequence will ever read. I'll shed a single tear for the well-reasoned discourse that could have been, and move on.

 

[retrogamer]

="ANOKNUSA, post: 281627, member: 44375"

It's fine to dislike systemd---there are valid gripes to be had---and discussions over its ramifications are fine as well. They're good, in fact---whether you like systemd or not, talking about how it may or may not affect your future is just the smart thing to do. But fallacious reasoning and random speculation and conspiracy mongering taint any and every attempt at a decent discussion on matter every goddamn time. Every time. The only reasonable thinking I've seen on the matter is a single blog post nobody of consequence will ever read. I'll shed a single tear for the well-reasoned discourse that could have been, and move on.

That really is a great article, I think the section breaking down what is meant by the "UNIX philosophy" in particular deserves to be referenced more often. This is where I think Poettering does deliberately feed the trolls, I don't believe for one minute he really meant it when he said:

"That all said, I am pretty sure that systemd actually brings Linux much closer to Unix than Linux ever was. All the true Unixes of today (such as FreeBSD, Solaris, ...) are maintained in central place, sharing code repository infrastructure, lifecycles and release schemes, for all their compenents, regardless if its the kernel, the libc or the rest of userspace. On real Unixes it is much easier to patch through the entire stack from kernel to userspace, because everything comes from the same source, and follows the same cycles. Real UNIXes tend to feel more uniform, since the same folks work on the whole stack, things come out of a single hand. Linux always was different: our components are independently maintained, in different repos, in different coding styles, by different people, with different release cycles. They are differently well maintained, a lot of our stack is actually traditionally very badly maintained if at all.

With systemd we kinda try to find the middle position, move to a more UNIX-like scheme, without dumping everything into the same repo like UNIX does, but at least the core userspace bits. But even beyond the procedural bits, there's a lot of areas where systemd is closer to traditional UNIX than Linux was. For example, one Unix mantra is "everything is a file" (which, btw, is a pretty broken one, because my printer is not a file, not at all), and you could say that systemd exposes one of the most core concepts of a Unix system, that of services/daemons as files via the cgroup logic. So yeah, if you claim that we are not UNIX, then I'll tell you that we are actually much closer to UNIX in many ways then we ever were.

I am pretty sure that most folks who constantly repeat that systemd wasn't Unix-like like a parrot actually have no idea what UNIX really is...”

EDIT: Also, lest anyone question "Why do you feel you know what comprises the UNIX philosophy?" I would just answer that I have read and mostly understood Basics Of The UNIX Philosophy.

 

[Oko]

OpenSuse is really based on Red Hat (OpenSuse denies this but OpenSuse uses the same package management, and the same systemd).

That is grossly inaccurate statement. SUSE enterprise Linux and its open source counter part OpenSUSE are products of
Micro Focus International which is one of Red Hat main competitors (not so much on U.S. market but certainly should be in Europe). Micro Focus International employees are second largest contributors to Linux kernel after employees of Red Hat. To give you further analogy OpenSUSE is to SUSE enterprise the same as Fedora and since recently CentOS to Red Hat. Fedora was always cutting edge unlike OpenSUSE which was more stable. Now Red Hat has CentOS which is probably as stable as OpenSUSE. Scientific Linux is dead and it is just a community version of new CentOS 7.0. The only truly genuine open source clone of Red Hat I am aware off is Springdale Linux which we use in our organization and is created by people from Princeton University and institute for Advance studies plus people from Rutgers University (New Brunswick campus) super computing centre.

The only thing SUSE and Red Hat share is package format.

As most people who are in this business in U.S. I don't care much for SUSE. When we need say Linux that typically means Red Hat or for some people Debian and its proprietary version Ubuntu(Canonical). For the record IIRC Canonical doesn't contribute to Linux kernel.

I am not sure what people who need OpenStack will do but I can certainly reiterate that many shops like mine will continue to run Red Hat branch 6.xxx until 2020 in-spite "older kernel issue" (That is typical statement coming out of Debian and other people who are not familiar with Red Hat kernel back porting policies).

So far I have been running away as far as possible from OpenStack but I have played a lot with Docker and it is pure garbage. Docker or even LXC might fool people who are not familiar with UNIX but can't fool people who used Solaris Zones and FreeBSD jails for the past 15 years.

 

[Crivens]

Maybe this will all resolve pretty fast.

I have run ldd against /sbin/init (which is systend) on my VM and it shows me that it needs glibc. Now they want networking? I am really waiting for the 0day attack on it, and maybe someone will learn then that this is not how to do it, but as usual only after someone ghosted the servers.

 

[walterbyrd]

Okay, I was wrong about Suse being based on Red Hat.

But my salient point is still just as true: all of the major commercial forces behind Linux are behind systemd. If you know about the way Linux is developed, then you know that is a very big deal. To suggest that most Linux users can avoid systemd for any length of time seems a bit naive. I think that anything Red Hat sees as a viable alternative to systemd, for most Linux users, will be gone by the end of 2015.

Again, don't get me wrong, I have a lot of respect for Slackware, Gentoo, and the Devuan project. But all of those put together are not David vs. Goliath, it's more like David vs an army of Goliaths. People like Patrick Volkerding may know the right thing to do, but he does not have that much control over most of Linux development, Pat only controls stuff unique to the Slackware distro. Even Linus only controls the kernel. For Linus to complain about systemd would only expose Linus as being powerless to stop it - and Linus has too much ego for that.

Everybody knows about boiling a frog, right? If you put a frog in hot water, the frog will immediately jump out. But, if you put a frog in cold water, and slowly heat the water, the frog will stay there until it dies.

I suspect that Red Hat's systemd strategy is analogous to frog boiling. If Red Hat does too much, too fast, users will abandon Red Hat for other Linux distros. But, if Red Hat moves slowly, if Red Hat claims that systemd is only optional, and the non-systemd will be around for years, then Red Hat waits to make sure that all viable alternatives to systemd are eliminated: then Red Hat can turn up the heat until they kill Linux as we know it, and there is not much anybody can do it.

JMHO, of course.

 

[Crivens]

It looks like this can be considered to be a fact by now. RedHat is doing a move, and the systemd crowd is playing the role which was labeled "useful idiot" by Lenin.

Linus himself can not do anything, he is forbidden to do so by his own agenda. He can not introduce code into the kernel which would forbid systemd to work because this would be a regression. And that is something the kernel team is all about to prevent. He can not break that policy, for any reason. So he can do nothing, publically. What RedHat is doing there is going for this point of maximum market control, but they are not in a position to steer this intelligently. They are a public company, and as such need to maximize the return for a quarter. Future be damned. That is how that stuff works, been there, seen it, got the tie.

The key pivot point is the applications. Who is going to depend on systemd and who will not. And I do think that a lot of the upstream from KDE for example is keeping portability in mind, and I hope that they do care about betting the farm on one platform. Because they would deliver themselves to the mercy of the suppliers of the stuff they depend on. That is true for all of the dependencies. As long as your dependencies and requirements are POSIX, you are safe. You can run your code on a lot of platforms, and no one can strong arm you into something. Leaving that behind would require a special kind of stupid, in my opinion. So my idea would be to see who is now pulling in dependencies for systemd and if they can be persuaded not to do that.

It is mostly desktop users which will be sucked into this, I can not think of any server-centric software which would need this kind of dependency to work. But my field of vision there is almost NULL, so if someone of you knows about server applications pulling in such dependencies, please speak up.

walterbyrd, maybe they want to cook the frog (which does not work, by the way) - but it is the upstream of the applications which must decide if they want to be part of that party. If RedHat is killing of Linux as we know it, then the Linux users must do something about that. We can help, but not act. And if they succeed, maybe it was time for Linux to die and be replaced by something else. Maybe it is time for HURD after all. In the end, Linux is only the kernel, and the GNU userland is running amok. Time will tell who needs whom, and who wants to join the dinosaurs. In the end, we all do. But not today.