Axel Thimm wrote:
> On Mon, Oct 24, 2005 at 09:16:06AM +0200, Jan Iven wrote:
> 
>>Dear all,
>>
>>at the "Spring" Hepix conference in Karlsruhe, we had complaints that SL
>>wasn't "scientific" enough, i.e. was missing several packages.
>>
>>I had contacted several developers to see whether they could make
>>+maintain SL versions (RPM+SRPM). Examples where there was some
>>willingness are OpenScientist and GSL, with ROOT and GEANT4 at least not
>>against the idea. I also have some things I prepared myself (xmgrace or
>>(eventually) a recompilable CERNLIB), which currently would go into the
>>CERN repository but are useful outside.
>>
>>The question is where to store these - I don't want such things to
>>become "CERN"-additions, but to be available to the wider community. We
>>could put them into the CERN "extras" repository and then mirror (I
>>think Connie already takes things from there), but I think some specific
>>section on scientificlinux.org would be more appropriate (including
>>instructions for package authors).
> 
> 
> How about an orthogonal project like scirpm.org, that would provide
> packages for SL as well as Fedora Core and other non-Enterprise
> distros?
> 
> I know that it is very difficult to introduce Enterprise distros on
> end-user desktops in academic research/universities at least in
> Germany, and there was already some small (momentum) within the Fedora
> Core Community in the past to create the above scientific oriented
> repo. So extending to/embracing Fedora Core would perhaps make sense.

I think this would be a good idea, having a separate scientific oriented 
repo.  I haven't replied to either of these, not because I don't think 
they're good ideas, but because each time I think of them I keep 
thinking of different ways to do it.

I'm curious about uploading things to www.scientificlinux.org.  I like 
the idea, but the problem is that that machine has no connection at all 
with ftp.scientificlinux.org other than being in the same room.  But 
different scenarios are going through my head about how to get the 
packages from one to the other.

There is also the idea of someone else hosting the scientific oriented 
repo that Axel talked about.  This could be at some lab, university, or 
private organization.  If we go this route then there has to be 
discussion about bandwidth, permissions, contributors, rpm format. 
Again different scenarious keep going through my head for this.  But if 
it was actually someplace else ... I wouldn't have to do as much, other 
than mirroring it, and giving my opinions where applicable.

Troy
-- 
__________________________________________________
Troy Dawson  [log in to unmask]  (630)840-6468
Fermilab  ComputingDivision/CSS  CSI Group
__________________________________________________