:: Re: [DNG] We Must be Prepared ....
Page principale
Supprimer ce message
Répondre à ce message
Auteur: Daniel Reurich
Date:  
À: KatolaZ, Jude Nelson
CC: dng@lists.dyne.org
Sujet: Re: [DNG] We Must be Prepared ....
On 24/06/15 21:33, KatolaZ wrote:
> On Tue, Jun 23, 2015 at 09:20:37PM -0400, Jude Nelson wrote:
>> Looks like Linus weighed in. I found the whole conversation thread
>> interesting.
>>
>> http://lkml.iu.edu/hypermail/linux/kernel/1506.2/05492.html
>>
>
> Yes, it is definitely interesting. The most scary thing is the
> emergence in that thread same of the argument:
>
> "...we can't actually ship a kernel.org kernel that will fail to boot
> with Fedora Rawhide or Arch AUR or whatever unless kdbus=0 is set on
> the kernel command line.", even in the form of a critic towards strict
> requirements "imposed" by systemd.


But that's been proven bunkum because systemd won't use kdbus if it's
not there. It's Lennarts bluff. If distro's break, then it's their
stupid fault for drinking at Lennarts fountain.

>
> This is what let me think that kdbus will most probably be merged into
> the kernel sometime soon, unless systemd developers unexpectedly
> decide to refrain from their insane way of pushing things forward,
> whatever the cost. But this is something is not going to happen, I am
> afraid.


Linus will call Lennart's bluff on this and not ship kdbus until it's
good and ready .... if that ever happens.

I suspect that Linus's comments about trust in GKH is as much more a
warning to GKH then a vote of confidence and certainly isn't a vote to
pull in kdbus without review. It also serves to give the other kernel
dev's a kick in the pants to do a proper review which will no doubt weed
out the big issues. Linus won't knowingly let shit in his kernel not
even for GKH. If GKH ask's again for a pull prematurely Linus will
likely give him the same treatment as he gave Kay Sievers.

In other words this is Linus's way of deflecting another flamewar and
putting the focus on facts and code whilst putting a few people in their
place.

We won't see kdbus in 4.2 and probably not 4.3 either. By 4.4 it will
most likely have grown some real legs and become generically useful or
be a dead duck.


--
Daniel Reurich
Centurion Computer Technology (2005) Ltd.
021 797 722