[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#1065416: Scheduling a meeting/BoF at DebConf 25



Hello,

On Thu, Jun 26, 2025 at 01:14:12PM +0200, Stefano Rivera wrote:
> I'm inviting Bastian, Matthias, and Aurelien to participate in a structured
> discussion of the pending issues here.

I happen to have met Bastian already and we had a wonderful time 
exchanging ideas and reasons. Thank you, Bastian! Note that I have a 
subtly different focus from the issue referred to the CTTE due to my 
work on rebootstrap. For the benefit of others, I'll be summarizing some 
of the insights I gained.  The conversation and this summary was on 
happening with CTTE hats.

A minor aspect I was not aware of is that Canonical approached Bastian 
about turning linux-libc-dev Arch:all. I don't know the exact details, 
but we might reach out there if we want to understand it in more depth.

As multiple of us already suspected, package/installation/archive size 
is not an aspected that seemed important to Bastian. Indeed, he 
recognizes that the typical installation grows as a result of 
linux-libc-dev being Arch:all without seeing that as an advantage.

He explained the use of linux-libc-dev in Debian ports to me. It seems 
like dak's architecture validation lags behind the one of dpkg in 
unstable severely to the point where dak does not know about the some of 
the architectures that we presently track as ports architectures. As a 
result, you cannot mention them in an Architecture field to be uploaded 
to unstable and this poses a significant problem to ports wishing to 
have a working linux-libc-dev package when it still was 
architecture-dependent. Switching to Arch:all enabled them to get their 
patches into src:linux as dak would no longer reject the package.

I do relate to this argument, but I think it should be solvable on a 
technical level whilst the social side to it may be challenging on its 
own.

We also briefly went into the ownership of the linux-libc-dev-ARCH-cross 
matter that is the subject of this CTTE matter, but we quickly concluded 
that it would be better to discuss that aspect in the larger setting 
scheduled by Stefano.

So much from me

Helmut


Reply to: