You are not logged in.

#1 2006-09-28 15:19:16

Romashka
Forum Fellow
Registered: 2005-12-07
Posts: 1,054

Looking at 2.6.19

Here is the huge list of patches that are planned to be merged into 2.6.19: http://kerneltrap.org/node/7153

There should be quite big number of fixes in hardware support (mostly IDE & SCSI), memory management, platform support, swsusp, UML, different filesystems, networking, SELinux, paravirtualization, scheduler, framebuffer, device mapper, RAID, Generic IRQ layer. Added encryptfs, started work for namespace virtualisation. Reiser4 is planned for 2.6.20.

And an interesting message about PATA to libata move is here: http://marc.theaimsgroup.com/?l=linux-k … 104283&w=2
(progress here: http://zeniv.linux.org.uk/~alan/IDE/STATUS.txt )

What do you think about this?


to live is to die

Offline

#2 2006-09-28 15:25:14

tom.deb
Member
From: manchester/UK
Registered: 2005-06-20
Posts: 42
Website

Re: Looking at 2.6.19

the dev process modifications allso seem interesting. basically, .19 will have lots of new features and .20 will be the same kernel but less buggy and more stable.

I personnally think it's time for reiser4 to get in there .


t o m d e b
_______________________________________
"the urge to destroy is a creative urge."
                                                 Mikhail Bakunin.

Offline

#3 2006-09-28 15:31:52

Dusty
Schwag Merchant
From: Medicine Hat, Alberta, Canada
Registered: 2004-01-18
Posts: 5,986
Website

Re: Looking at 2.6.19

tom.deb wrote:

basically, .19 will have lots of new features and .20 will be the same kernel but less buggy and more stable.

Yeah... I liked it better when all released kernels were stable and the development happened in another tree. Seems there's been a lot more... variability in stability... since Linus chose not to go with a 2.7 tree.

Perhaps your observation is an indication they will eventually return to an even/odd dev/stable release system, but on minor revisions rather than major.

Dusty

Offline

#4 2006-09-28 15:35:37

Romashka
Forum Fellow
Registered: 2005-12-07
Posts: 1,054

Re: Looking at 2.6.19

tom.deb wrote:

the dev process modifications allso seem interesting. basically, .19 will have lots of new features and .20 will be the same kernel but less buggy and more stable.

Yes, thanks for pointing this, I forgot to mention this.
Linus proposed that 2.6.odd will be 'development' and 2.6.even - 'stabilizing' releases.

tom.deb wrote:

I personnally think it's time for reiser4 to get in there .

There is some problems with getting Reiser4 in 2.6.19. But if 2.6.20 will be 'stabilizing' release then Reiser4 will most probably appear only in 2.6.21.

EDIT: Dusty, you are so fast! (finished writing your post right before mine). This thread seems to be quickly growing. :-)


to live is to die

Offline

#5 2006-09-28 21:27:03

tom.deb
Member
From: manchester/UK
Registered: 2005-06-20
Posts: 42
Website

Re: Looking at 2.6.19

Seems there's been a lot more... variability in stability... since Linus chose not to go with a 2.7 tree.

I know.  It's weird how i tend to have version I liked and haen't updatd for quite some time. I stuk to 2.6.11 for ages, just because i didn't feel the need to update. 

It's also interesting to know that Red Hat Enterprise Linux that I use at work is stuck on 2.6.9.


t o m d e b
_______________________________________
"the urge to destroy is a creative urge."
                                                 Mikhail Bakunin.

Offline

#6 2006-09-28 22:20:04

allucid
Member
Registered: 2006-01-06
Posts: 259

Re: Looking at 2.6.19

Does this mean I should stick with 2.6.18 if I want to stay stable? Will security patches be backported to 2.6.18 while work on 2.6.19 is being done?

Offline

#7 2006-09-28 22:35:49

phrakture
Arch Overlord
From: behind you
Registered: 2003-10-29
Posts: 7,879
Website

Re: Looking at 2.6.19

If this is the case, I think the arch stable kernel will remain at 2.6.18, with 2.6.19 in testing, then 2.6.20 going to current... it's the only logical way about it if 2.6.19 will be testing

Offline

#8 2006-09-28 22:59:56

Romashka
Forum Fellow
Registered: 2005-12-07
Posts: 1,054

Re: Looking at 2.6.19

I don't know if 2.6.19 will be official 'unstable' release. IIRC there was no final decision yet.

BTW, did anyone notice this message:

From: Andrew Morton [email blocked]
Subject: Re: 2.6.19 -mm merge plans
Date:    Thu, 21 Sep 2006 11:55:45 -0700

On Thu, 21 Sep 2006 14:33:41 -0400
Jeff Garzik [email blocked] wrote:

>  I think it's more than fair for Andrew to loudly shame said
> developers into action.  "The following nincompoops are holding up the
> release:  Jeff Garzik [bug #1222, #3391], Greg KH [bug #9987, #4418], ..."

I don't have the bandwidth to do that work, alas.  I know how to do it, and
have tried to do it, but a) it's dull and b) even I get tired of whining at
people all the time.

The good news is that google is prepared to hire someone to sit next to me
and do it
, but I haven't got off my ass and written the job description
yet.

?


to live is to die

Offline

#9 2006-09-29 12:25:39

filoktetes
Member
From: Skien, Norway
Registered: 2003-12-29
Posts: 287

Re: Looking at 2.6.19

I think Andrew Morton works at Google.

Offline

#10 2006-10-11 11:41:05

iphitus
Forum Fellow
From: Melbourne, Australia
Registered: 2004-10-09
Posts: 4,927

Re: Looking at 2.6.19

Dusty wrote:
tom.deb wrote:

basically, .19 will have lots of new features and .20 will be the same kernel but less buggy and more stable.

Yeah... I liked it better when all released kernels were stable and the development happened in another tree. Seems there's been a lot more... variability in stability... since Linus chose not to go with a 2.7 tree.

Perhaps your observation is an indication they will eventually return to an even/odd dev/stable release system, but on minor revisions rather than major.

Dusty

The 2.odd release cycle worked quite differently though. It occurred when there were major changes that could not be safely integrated within the standard release cycle. Within the 2.6 series, there have been no such major changes.

libATA, while big, is something that has been merged in over a period of time, and has received a lot of testing. It exists alongside the old drivers -- piix is still there if you want it, so if there are problems, the original isnt available. It has been getting to the stage where a majority of the development has been occurring within libATA and not the in kernel drivers, so it was a neccesity to merge it sooner or later. 
It is true however that in the past, new things, and things like libATA would have gone into a 2.odd at an earlier stage if there was one in progress, and maybe backported later.

-mm has in effect currently, replaced the 2.odd being the place where new and unstable things go. The way things have been going recently have created an interesting new development process, that manages to get things pushed in quicker. Months in -mm, then to a 2.6.odd kernel. While using 2.6.odd for unstable, and 2.6.even for stable changes is an interesting, and slightly.... messy? concept, I can see how it will work very effectively.

The option for a super stable, 2.even style kernel is still available anyway. 2.6.16.y is still being developed, with Adrian Bunk posting 2.6.16.30rc1 just over a week ago, and another 2.6.17.y coming out just the other day.

So yeah, in summary, the new processes have been shaky at first, but should smooth out quite nicely as they're adapted to and tweaked.

James

Offline

Board footer

Powered by FluxBB