Kernel Version Explanation

Kernel Version Explanation

[sean]

Would it be possible to get an explanation of the new kernel version
format put onto the http://www.kernel.org website?  If it's there and
I'm missing it, feel free to call me all kinds of silly names; but can
you please send me a link?

Thank you,

-- 
Sean

Re: Kernel Version Explanation

[randy_dunlap]

On Sun, 29 May 2005 09:09:45 -0500 sean@capitalgenomix.com wrote:

| Would it be possible to get an explanation of the new kernel version
| format put onto the http://www.kernel.org website?  If it's there and
| I'm missing it, feel free to call me all kinds of silly names; but can
| you please send me a link?

Can you be more explicit in what you are looking for?

If you click on 'prepatch', 'snapshot', '-ac patch', and '-mm patch',
you can read some explanations.  What is missing there?


It looks to me like the word "stable" is overused on the main page
at www.kernel.org .  I would also prefer to see all of the 2.6.*
kernel versions together, above the 2.4.*, 2.2.*, and 2.0.* lines.



E.g. (however, this is too cluttered IMO; the prepatch/-ac/-mm
links do this well without the clutter):

The latest stable version of the Linux kernel is:  	2.6.11.11
The stable kernel patchset contains only critical and security patches
to the latest 2.6.x kernel.

The latest prepatch for the stable Linux kernel tree is:  	2.6.12-rc5
This is the mainline kernel where all new patches are added.
If they are not critical, they usually go thru the -mm patchset
before being merged here.  If they are critical, they may be
merged here first.

The latest -mm patch to the stable Linux kernels is:  	2.6.12-rc5-mm1 	
This is the development tree for 2.6 kernels.  Patches are
generally merged here for testing before being merged into
the mainline kernel tree.

The latest -ac patch to the stable Linux kernels is:  	2.6.11-ac7
(see the -ac patch explanation)


Thanks,
---
~Randy

Re: Kernel Version Explanation

[H. Peter Anvin]

randy_dunlap wrote:
> 
> It looks to me like the word "stable" is overused on the main page
> at www.kernel.org .  I would also prefer to see all of the 2.6.*
> kernel versions together, above the 2.4.*, 2.2.*, and 2.0.* lines.
> 

That's because there isn't an odd-number series right now.

	-hpa

Re: Kernel Version Explanation

[Schneelocke]

On 30/05/05, H. Peter Anvin <hpa@zytor.com> wrote:
> > It looks to me like the word "stable" is overused on the main page
> > at www.kernel.org .  I would also prefer to see all of the 2.6.*
> > kernel versions together, above the 2.4.*, 2.2.*, and 2.0.* lines.
> 
> That's because there isn't an odd-number series right now.

Will there ever be one again (at least in the foreseeable future)?
We've had "Linus = stable, -mm = unstable" for a long time now, and it
seems pretty much official now that there won't be a 2.7 anytime soon.
The actual development of new features is happening in the relevant
maintainers' trees, anyway, so there simply seems to be no need for a
single highly development-oriented tree (like 2.5 was) anymore - quite
the contrary.

>         -hpa

-- 
schnee

Re: Kernel Version Explanation

[Fao, Sean]

Schneelocke wrote:

>On 30/05/05, H. Peter Anvin <hpa@zytor.com> wrote:
>  
>
>>>It looks to me like the word "stable" is overused on the main page
>>>at www.kernel.org .
>>>      
>>>
>>That's because there isn't an odd-number series right now.
>>    
>>
>
>Will there ever be one again (at least in the foreseeable future)?
>We've had "Linus = stable, -mm = unstable" for a long time now, and it
>seems pretty much official now that there won't be a 2.7 anytime soon.
>The actual development of new features is happening in the relevant
>maintainers' trees, anyway, so there simply seems to be no need for a
>single highly development-oriented tree (like 2.5 was) anymore - quite
>the contrary.
>  
>

My understanding was that Linus eventually decided upon something in the 
middle.  I understood that there still wouldn't be a 2.7.x branch 
(unless major changes occurred, which would severely risk breaking the 
stable tree).  However, it was also my understanding that Linus would 
return to the even/odd version system; but, would do so in the rev.  In 
other words, 2.6.even would be stable, while 2.6.odd would be 
development.  I did, however, become slightly confused when I connected 
to http://www.kernel.org and noticed that the latest stable kernel was 
2.6.11.11 because it's both odd and contains four version numbers rather 
than the  three, which we've usually seen.

Hope that clears up what my confusion is.

Thank you,

-- 
Sean E. Fao

Re: Kernel Version Explanation

[randy_dunlap]

On Tue, 31 May 2005 12:45:43 -0400 Fao, Sean wrote:

| Schneelocke wrote:
| 
| >On 30/05/05, H. Peter Anvin <hpa@zytor.com> wrote:
| >  
| >
| >>>It looks to me like the word "stable" is overused on the main page
| >>>at www.kernel.org .
| >>>      
| >>>
| >>That's because there isn't an odd-number series right now.
| >>    
| >>
| >
| >Will there ever be one again (at least in the foreseeable future)?
| >We've had "Linus = stable, -mm = unstable" for a long time now, and it
| >seems pretty much official now that there won't be a 2.7 anytime soon.
| >The actual development of new features is happening in the relevant
| >maintainers' trees, anyway, so there simply seems to be no need for a
| >single highly development-oriented tree (like 2.5 was) anymore - quite
| >the contrary.
| >  
| >
| 
| My understanding was that Linus eventually decided upon something in the 
| middle.  I understood that there still wouldn't be a 2.7.x branch 
| (unless major changes occurred, which would severely risk breaking the 
| stable tree).  However, it was also my understanding that Linus would 
| return to the even/odd version system; but, would do so in the rev.  In 
| other words, 2.6.even would be stable, while 2.6.odd would be 
| development.  I did, however, become slightly confused when I connected 
| to http://www.kernel.org and noticed that the latest stable kernel was 
| 2.6.11.11 because it's both odd and contains four version numbers rather 
| than the  three, which we've usually seen.
| 
| Hope that clears up what my confusion is.

The even/odd-ness of 'x' in the 2.6.x version number discussion
was ultimately rejected as far as it having any stable/development/test
meaning.

Mostly Linus's 2.6.x tree is for stable work and Andrew's -mm
patchset is for development work.  However, as I often say,
everything is relative.  There are no absolutes in this.


---
~Randy