OSNews

> Seriously- do we need new linux 2.6.34.3694.5766
> version update news on frontpage?

Err... yes? OS news on OSNews sounds pretty reasonable to me.

Since the changelog(pure text) from the last version is ~5MB, I would say a lot has changed, and it is worthy of a little notice on OSNews...

Well, you are right about that vanilla kernels aren't in the distro releases, but if you got a rather "lowtech"(please do not missunderstand what I mean here) distro like Slackware, there is not much problems running a vanilla kernel. I guess you can do that on a "hightech" distro like suse/fedora/Xbuntu, but with a little more hassle than say Slackware...

If this was a 2.6.23.x relaes I could understand that you are upset because it got attention, but this is a little bigger release, so I for one think that it is nice to hear about it...

Well. In short, here is one person who's always used vanilla kernels, as long as I'm using Linux overall. And I'm not alone. So yes, new vanilla releases are indeed newsworthy for some people out there besides distro devs and packagers.

Actually Slackware comes with an unpatched vanilla kernel and particularly for Slack {and derivatives as long as they also include vanilla kernel, and there are quite a few of these distros} users such an announcement makes sense as this group tends to recompile and use the latest kernel A LOT.

(Not to say Debian etc. users don't, but most other Mandriva, PCLOS etc. users probably do not replace their kernel on a regular basis.)

Anyway, I do not recall constant kernel release announcements as claimed here on OSnews, not even every week.

BarnabyH

OOps, somebody just posted in a similar vein...

Gentoo users, too, tend to update their kernels frequently, and the initial install of Gentoo involves compiling one's own kernel. One choice is vanilla-sources, the stock kernel.

So yeah, I like to see news about these things, and especially, I enjoy the discussion, to hear about the experiences others are having.

I probably shouldn't respond to Antik's antics. But... this represents 3.5 months work of a *very* busy project. Have a look at the "short" version of the change log. It's *quite* substantial. 2.6.24 is probably the single most major kernel release in a few years. 2.6.x version releases have never been reported "several times a week" on OSNews. And I happen to be using 2.6.24 in Ubuntu Gutsy (that's a distro) right now. And it's pretty close to vanilla. I imagine the Fedora 9 dev branch is also using it. The final production version of Gutsy can be expected to be based upon this kernel version. Fedora 9 will use either this or 2.6.25 depending upon timing.

So in a nutshell, you are wrong, wrong, wrong, and wrong. I could see you maybe having a complaint which was at least debatable if you complained about an announcement of, say, 2.6.23.14. But it would be hard to have timed your complaint any more badly than you did.

And for the gods' sake simply ignore stories which don't interest you instead of wasting *everybody's* time whining.

Edited 2008-01-25 19:26 UTC

WTF are you going on about? This is a version release, not an -rc or a sub-point update release.

This kernel will be in new distro releases, such as openSUSE 11.0, which has already been using the -rc versions in development. The days of every distro rolling their own branch of the kernel tree are long gone, the vast majority use vanilla with whatever they may feel are worthwhile custom patches, but they always standardize on mainline.

So, considering the kernel forms one of the major core components of a linux distribution, and this is OSNews, I'd say a new kernel release is worthy of a mention.

Indeed. Most people refuse to call it GNU/Linux because they are lazy. Some of them refuse to call it that because they disagree with Stallman. I happen to be both, but I still call it GNU/Linux because I believe in giving credit where it's due.

Is it bad form for GNU to insist their name be part of the OS they created? Maybe, but it's their right to, they made it! And if it is, it's equally rude for Torvalds to claim sole naming rights on the work of GNU.

Ditto for bcm43xx, as it so happens NDISwrapper stopped playing ball with my v.old version of this chipset about the time kernel-2.6.23 appeared, and the contemporary bcm43xx driver had no Ad-Hoc support. They claim that their new driver version (now called b43 or sumfink) which uses mac80211 will support it and I can stop sitting on my hands at the office

And here for the rt61 update.

Do you understand that git is a distributed development environment and that Linus and everybody else will have all their work on their laptops, ready to fix anything that comes about?

More importantly, this is a "work" conference. Finally, the people running serious systems are not running the latest released kernel and the people running serious systems have support teams behind them at Canonical, Novell, RedHat and Mandriva.

Stop spreading FUD.

Everything you say is true, and was either stated or implied in my post, which is why I said the risk is trivial.

My point, a tiny one, was that this is not the normal practice in software development. Not just because of risk, but also optics. Just try to make a major release at your workplace, and then leave for any reason, even if you promise to stay in touch, take your laptop, etc. [Note too that flying time to Australia from North America is not negligible (as Linus points out), as well as effects of jet-lag.] It just doesn't happen. That's not FUD, but reality. In any event, I'm glad to see a new kernel milestone. Incremental, continuous improvement is great and produces real, tangible user benefits.