LHC Computer Grid - Rollout
> [mailto:[log in to unmask]] On Behalf Of Brew, CAJ (Chris)
said:
> So as a experiment we are probably going to want to certify
> each of these OSs individually anyway.
Where do you stop - if someone takes the Fermilab sources and recompiles
them do you want that classed as a different distribution?
> It's not too difficult to put "or" clauses in the JDL or is there a
> maximum length for the JDL.
I'm not sure, but expressions can be pretty long. However, do the
experiments really want to maintain complicated expressions, which might
need to be updated when a site comes in with a new flavour?
Looking to the future, what do we expect to happen when there's
another release of RHEL? It's also worth bearing in mind that there are
three attributes defined in the schema, called Name, Version and
Release, and the other two are not really used at the moment - it might
well make sense to put finer-grain information there.
Stephen
|