@sys doen't work on RW volumes

[mmarion@qualcomm.com]

On 19 Jul, Tomas Olsson wrote:

> The automatic sysname guessing code (in lib/go/gensysname.c) needs
> updating, we just need to figure out what it should do. There seems to be a
> lot of different naming schemes around. Originally, the Arla team decided
> on i386_linux6 and similar. 6 is the libc version. For practically all
> binaries, this really does make more sense than depending on the kernel

Man.. I would love it if the afs sysname were based on libc instead of kernel.
In fact, I'd go one better and do the glibc version myself.. i.e.
amd64_linux22 or amd64_linux23 (glibc 2.2 vs 2.3 is our biggest issue right
now).  BTW, is something built under glibc2.3.x still libc6? 

> major version. But maybe we should add a configure option or smth to allow
> for the Transarc/OpenAFS way too. And then there is the i386_rh9 way.

Yeah, unfortunately we're using all transarc servers, so we'd probably want to
match those.  Though if we have all the clients working the same, I guess the
server really doesn't matter.

I personally don't like the distro way, since say, rh7.3-rh8 and sles8 really
match kernel/glibc wise so can share the same stuff.

-- 
Mike Marion-Unix SysAdmin/Staff Engineer-http://www.qualcomm.com
SCCS, the source motel!  Programs check in and never check out!  - Ken
Thompson