[elrepo] Can we rename kernel-ml to something more correct?

[Phil Perry]

On 05/09/18 22:46, Dave Chiluk via elrepo wrote:
> I had to explain this again today.  @alan, or @akemi, thoughts?
> 
> Thanks,
> Dave.
> 

Hi Dave,

It's unlikely to be renamed, given the long history now. We did consider 
the naming very carefully when the packages were originally conceived 
and we have revisited that decision again now but do not feel renaming 
is warranted. The kernel-ml packages are built from the sources 
available from the "mainline stable" branch, i.e the results of Linus' 
-rcX "development tree" release candidates.

I also feel it is a question of perspective. You may view (and prefer 
the name) kernel-ml as -stable, but from an Enterprise Linux 
perspective, someone else may view the distro kernel as the "stable" 
kernel and kernel-ml as a cutting edge alternative. Hence personally, I 
do not feel "stable" is a good fit to describe kernel-ml for the 
Enterprise Linux world in which we operate.

We are quite happy to add some further documentation to either the Wiki 
or release announcements to help clarify exactly what it is that is 
being offered if folks are genuinely struggling with this. T be honest 
though, I think this is the first time it's been raised. Most confusion 
seems to arise around kernel-lt versions.

Phil


> 
> On Thu, Aug 30, 2018 at 2:06 PM Dave Chiluk <dchiluk at indeed.com 
> <mailto:dchiluk at indeed.com>> wrote:
> 
>     If I'm not mistaken in kernel-ml, ml stands for mainline.  This is
>     actually quite a misnomer as kernel.org <http://kernel.org> uses
>     mainline to refer to Linus' development tree.  In actuality
>     kernel-ml is instead built off of the stable tree.  I've seen a
>     number of people get confused by this nuance.
> 
>     I know it would suck to do, but it really feels like it should be
>     named something like kernel-stable *(because of kernel.org
>     <http://kernel.org> conventions), or possibly kernel-st (for
>     short-term).  I'm not really sure what the correct name should be,
>     but I'm certain it should not be kernel-ml.
> 
>     We could even create a transitional kernel-ml package that depends
>     on the newly named rpm to provide a no-brainer upgrade path.
> 
>     Thanks,
>     Dave.
> 
>