Aw: Re: Any input for some talk about usage of Debian in HPC
> Gesendet: Montag, 27. Mai 2024 um 14:41 Uhr
> Von: "PICCA Frederic-Emmanuel" <frederic-emmanuel.picca@synchrotron-soleil.fr>
> An: "Peter Wienemann" <fossdev@posteo.de>
> Cc: "Carsten Aulbert" <carsten.aulbert@aei.mpg.de>, "Andreas Tille" <andreas@an3as.eu>, "Debian Science List" <debian-science@lists.debian.org>, debian-astro@lists.debian.org, debian-hpc@lists.debian.org, "Ole Streicher" <olebole@debian.org>
> Betreff: Re: Any input for some talk about usage of Debian in HPC
>
> > 2. They are thinking about building up a credible alternative to
> > RHEL/AlmaLinux to decrease vendor dependence (see slides 12 to 14).
> > Concerning this second point they stress that it still lacks management
> > approval.
>
> Do you know if the CERN will officially put some man power in official Debian packaging of there software? (ROOT, geant4, etc...)
We once had ROOT (as in https://wiki.debian.org/DebianScience/ROOT) in Debian (remnants on https://salsa.debian.org/science-team/root-system, a second effort on https://salsa.debian.org/science-team/root ).
I tend to agree that having upstream uploading to Debian would be beneficial to both sides. The problem I see is that for the science you (often, not always) want a short cycle between devs and users for the scientific parts, but (often, not always) long cycles for the libraries used for maximal reliability. Conda and brew seem to fill that gap between sharing and trusting to some degree,
Cheers,
Steffen
Reply to: