On Thu, Mar 21, 2013 at 01:51:33AM -0400, S.Tindall wrote:
> On Wed, 2013-03-20 at 21:11 -0700, Yasha Karant wrote:
> >
> > why did SL -- ... -- select to use the present TUV instead of SuSE enterprise ...
>
> You're right, it is a silly question. Or is Google broken again?
>
> https://www.scientificlinux.org/documentation/faq/general1
>
The link does not really answer the question, or only answers the FermiLab side of it.
From the Brookhaven Lab (and TRIUMF) side, this "selection" happened very early on.
We have settled on Red Hat Linux (without and well before the "E" and TUV nonsense)
fairly quickly around the time the first dual and quad Pentium Pro machines
came out. This must have been around 1998 time frame.
These dual and quad Pentium Pro machines were clocked at around 200 MHz and
cost a fraction of our massive Silicon Graphics UNIX (IRIX) machines, had about the same
CPU performance for our physics applications, but with more RAM and with the 100Mbit ethernet.
Stability was about the same as the SGI machines.
So obviously, we switched from IRIX to Linux as quickly as we could port our software
to run on Linux (porting from 64-bit IRIX to 32-bit Linux, how is that for progress?)
Why Red Hat? There were other contenders at the time. We certainly had Debian proponents in house.
I think the Red Hat "graphical" installer and the "kickstart" function were the main
deal makers.
Once selected, we stayed with Red Hat Linux and in a way we still are with it.
Somewhere along the line came the split into Fedora, TUV "E" Linux, SL/SLC Linux, CentoOS, etc
Some people are not aware of the history from before this split
and think that that was the beginning of history. For us it was just one more
bump on the road.
P.S. From the CERN side, I know the story is different yet again. Maybe Alan Silverman
will write it up in a book some day.
--
Konstantin Olchanski
Data Acquisition Systems: The Bytes Must Flow!
Email: olchansk-at-triumf-dot-ca
Snail mail: 4004 Wesbrook Mall, TRIUMF, Vancouver, B.C., V6T 2A3, Canada
|
