[OpenAFS-devel] Re: OpenAFS: conventions for new @sys? (rs_aix43 s390_linux22)

[Russ Allbery]

Nathan Neulinger <nneul@umr.edu> writes:

> How bout a much more capable facility than @sys... Extend that concept
> into what the old apollo workstations had - dynamic symlinks. Why not
> make that part of AFS. i.e. support more than just @sys.

> For example, alot of sites would probably find '@acsys' to be useful to
> do the autoconf determined sys type.

Er... for what?

The output of config.guess is useful for FSF build tools and for doing
some system-specific checks in configure, and that's pretty much it.

> I dunno exactly, but it seems like making a more general purpose
> mechanism would be very useful.

Hm.  My immediate reaction is that we really don't want to break things
that aren't broken.  I find @sys quite useful for managing a large shared
software installation on basically every platform Transarc currently
supports, and with a few minor tweaks it could be even more useful
(currently, we're using fs sysname to do that tweak for AIX ourselves).
I'd really rather not add a bunch of complexity to an idea that's already
working fine.

If we start trying to track things like ABI models and whatnot in @sys
rather than some rough idea of "stuff compiled on a system mapped to this
@sys value will probably work on other systems also mapped to this @sys
value in the absence of ABI linking issues and deep kernel magic," I think
we're going to end up with too fine of granularity.

-- 
Russ Allbery (rra@stanford.edu)             <http://www.eyrie.org/~eagle/>