Slashdot Mirror
Torvalds & Linux Dev Process
sebFlyte writes "Builder UK is reporting that Linus Torvalds is concerned that the Linux production kernel maintainence process might be overly taxing Andrew Morton, saying: "One issue is that I actually worry that Andrew will at some point be where I was a couple of years ago -- overworked and stressed out by just tons and tons of patches. If Andrew burns out, we'll all suffer hugely." Morton himself wants to make -mm releases more often. He sees bugs as more of a problem, rather than patches themselves. His solution is simple: "I'd like to release -mm's more often and I'd like -mm to have less of a wild-and-crappy reputation. Both of these would happen if originators were to test their stuff more carefully.""
Linux Kernel Gets Fully Automated Test
n t
2.6 stabilization project (helped a lot during 2.5.x develpment AFAIK)
http://www.osdl.org/docs/stabilization_plan.curre
Linus has declared that "git" is a six-month experiment, and the end of that time
... it is true that Linux kernel development
(which is in early October) he will review other open source options to see whether
there is a better option available now than when he looked around in April when git
was created after he found nothing that met his needs.
darcs was rejected by Linus in April, and I don't think that any of the limitations
that Linus saw in it then have been resolved (these limitations are w.r.t. Linus'
requirements for Linux development, darcs may be a fine system, it just didn't meet
Linus' needs).
The only real contender in April was monotone. I'm not sure why Linus didn't like
it enough to go for fixing it rather than rolling his own. Now the prime contender
is mercurial (by Matt Mackall).
As for git being a limiting factor
was probably stalled for about a month during the switch from BitKeeper. But in
the grand scheme of things that doesn't sound too bad compared with other horror
transitions I've gone through when companies switch propietary SCM systems. The
biggest challenge has been the moving target of learning git. In the first couple
of months so much changed from day to day, that it was challenging keeping up with
what could be done. The rate of change has slowed, and the feature set is much
better. The 1.0 release of GIT is not too far away which will put a big compatability
stake in the ground to make it much harder to make incompatible changes going
forward (though in fact there haven't been too many dead-end and retrace your steps
moments in git).
Perhaps after we pass the six month probation period and Linus announces
whether git is "it", then more people might be willing to invest the time
to learn to use git.
ACPI has to be disabled, otherwise it will either freeze or spontaneously reboot. 2.6 will crash while loading modules related to USB, network (loading the 8139too module consistently crashes), agp and hotplug system detection. The install cds of Ubuntu and Suse are stable enough to install, but once installed to the hard drive, the system consistently hangs due either to one of the errors I've already mentioned; or for reasons I haven't tracked down yet.
[rant]
I'm not a kernel programmer; I just want a working desktop. KDE works on NetBSD (which automatically detects my sound card) so until the kernel people get their shit together; I'm done with Linux.
[/rant]
GIT is pretty much a fully-featured SCM by now. It still isn't fully stabilized and there is still plenty of things to work out, but it is completely practically usable, and it's actually from a large part designed to make things easy for the kernel people (well, especially Linus itself ;). I don't know if the workload is more or less than when using BK, and that would be actually a very interesting question to ask Linus, but I *think* things are easier for Linus now. I'm not really sure about Andrew Morton because AFAIK he is primarily a quilt user, permanently juggling hundreds of patches, and I don't know to how big degree did he ever use BK or git.
Regarding the strategic plan, I think Linus and most of the developers are happy with git now and don't plan to switch, and there are interfaces between git and some other systems (e.g. Merculiar (very popular between another significant proportion of kernel developers) and Monotone) that should enable you to seamlessly use those for your development.
It's not the fall that kills you. It's the sudden stop at the end. -Douglas Adams
This was true in the past and technically, it is still true from some POV now, but in reality, what you now get as the git-core tarball from kernel.org definitely is a SCM system (with fairly crude interface, though, and I'd biasedly recommend Cogito for a nice interface to GIT ;-). GIT is indeed more general than that, and you do not necessarily have to use the SCM capabilities, but GIT _has_ those capabilities now, so it's probably less confusing if you make that clear as well.
It's not the fall that kills you. It's the sudden stop at the end. -Douglas Adams
These wouldn't solve ALL problems, or even the majority of them, but they would solve some and they would make life easier on developers in the long-run. Are these being used? Well, a glance at the Freshmeat graphs for Web100 shows that it is getting downloaded. This doesn't mean it is getting used, though. The same is true of virtually all of the other code I've mentioned. People have copies, but if the code being submitted is flakey and taking a long time to fix, then maybe the code is not being used as much as it could/should be.
The tools exist, the tools exist on people's hard drives, but unless the tools are being used in practice, that's not going to do any good.
It's a small world and it smells funny; I'd buy another if it wasn't for the money; Take back what I paid (SoM)
I believe the reason that LT discontinued the odd-even numbering was that the "development" kernels were under-tested and provided insufficient grounds for migrating the tree from development to stable.
Are you aware that the LKM team puts out a stable subversion of each release? I.E. 2.6.11 is released, then 2.6.11.1, 2.6.11.2, 2.6.11.3, etc?
I'm confused. Any clarification on this from the list that the article doesn't give?
Well, I'm not sure I understand the situation correctly, but is seems to me like a branch to 2.7 might be coming. Since there's been no separate development branch, there's been a lot more patching than usual for a stable kernel. I think the comments indicate that 2.6 might be close to "done" and should enter maintenance mode. Starting major breakage in the 2.6 branch would overwork Morton, hence the need for a "changed development process".
I still haven't even bothered to move to 2.6.x as I have no reason to.
Don't confuse user numbers with development. In fact, they are usually inversely related (the less development, the more stability and the more users.... to a point). And I'm quite sure the causality is that less development in 2.6.x leads to more adoption, not the other way around.
Kjella
Live today, because you never know what tomorrow brings
I'm still using 2.4 on most of my machines. My desktops are running 2.6.12.5 (2.6.13 gives me intermittant lockups)
.config for each type of kernel (I have two - desktop and server/firewall.) When a new kernel comes out, compile as a package, install it to a test machine to make sure its stable, then roll the packages out to the field during the scheduled maintenance window. I maintain a few dozen machines, and this keeps things very simple.
of my 4 Linux systems here, only one uses 2.6.x. The others run 2.4.x. Upgrading the kernel on some of my systems is a real pain (my firewall, for example, is a 100MHz processor w/32MB RAM.
I've installed 2.6 on a couple of firewalls (a P75 and a P133, each with 32MB RAM) and it runs fine.
Recompiling the kernel to my specifications is a real pain.
Why? One kernel tree on a dev machine, keep a separate
Are you really that dumb? I'm not even French and yet I know that the French have a glorious military history! You stupid bastards know you're repeating a lie - and why? Because France said no to attacking Iraq! At least France didn't illegally invade a foreign country (Iraq) and get stuck - losing marines by the hundreds to the locals! Go USA! Hahahahahaha!
The single largest attraction of the 2.6 series is the new driver model with its /sys filesystem. It allows not only taking driver coldplugging and, especially, hotplugging to a new level, but I also use it on servers because of the hardware introspection capabilities that it offers.
There are also some really interesting things coming in 2.6, such as SECCOMP, NFSv4 and kernel key retention. SECCOMP allows per-process syscall jails for secure execution of third-party untrusted binaries, and kernel level key retention will allow you to, for example, keep Kerberos credentials in the kernel (once it has been implemented in the Kerberos libraries, that is), which allows for much more stateful key inspection and manipulation (to do stuff like automatic renewal and creds-to-process/user mappings). Of course, it can do the same things with other cryptographic keys as well, like SSH keys, X.509 certs, saved passwords, and what not.
Then, of course, there are tidbits of interesting things all over the place, like kernel thread preemption and what not.
Of course, if you, for any reason, aren't interested in any of this, that's your choice, but I was running the test versions of the 2.6 kernel before 2.6.1 was released because I was looking forward so immensely to many of these features.
What exactly is wrong with France's military record? Dien Bien Phu? The plan was bad (OTOH, who would have thought that Vietnamese dragged artillery uphill through the jungle), but the troops fought very well indeed. WW2? They lost to world's biggest military superpower (which was itself defeated by combined efforts of USA, USSR and Britain years later). Oh the humiliation! And if we look at WW1, I don't think you can blame their courage or willingness to fight there.
If we look even further in to the past, we see the French marching all over Europe. And they did save Europe from Muslim invasion.
Lesbian Nazi Hookers Abducted by UFOs and Forced Into Weight Loss Programs - -all next week on Town Talk.