Al Hopper wrote:
> On Sat, 2 Jun 2007, Roland Mainz wrote:
> > Alan Burlison wrote:
> >> Roland Mainz wrote:
[snip]
> > Unfortunately this is not the finish line, only the first checkpoint. We
> > have lots of other things on the ToDo list, including "shcomp" (shell
> > script compiler (and the "shbinexec" kernel glue)), kshdbx, Dtrace
> > support for ksh93, switching over the consumers of alias.sh,
> > /usr/bin/sleep&&/usr/bin/test switchover, more builtin command issues,
> > pfksh93 bickering (or better: RBAC vs. POSIX behaviour), /usr/bin/ksh
> > migration to ksh93, development of a libshell API which matches the
> > stabilty requirements of OS/Net, killing more bugs, threading support in
> > ksh93 (e.g. make libshell's interface threadsafe and add support for
> > multiple worker threads in a subshell instance), aiding others to use
> > ksh93 features in their shell scripts (e.g. each time I look at webrev I
> > could scream. And scripts like "bfu" and /opt/onbld/bin/acr are
> > canidates for the worst mixture of Bourne/Korn-style coding ever seen in
> > history, right after some of the SMF scripts). And somewhere between
> > these items I need a real-world job and a faster build machine because
> > my Ultra5 _sucks_ (sorry for using explicit languauge but waiting two
> > days for a build to finish with "yet another build error" ("... please
> > fix and rebuild from scratch, see you with another error in two
> > days...") isn't really something which makes me happy right now... ;-(
> > ).
>
> What is your preferred hardware built environment to help you complete
> the ksh93 integration project. Please describe 3 configs:
>
> 1) maxxed out config
> 2) reasonable, middle-of-the-road config
> 3) minimally useful config (and I know a U5 does not meet this
> requirement)
>
> Are you looking for SPARC or x86/AMD64 hardware or both.
About [2]: Right now the only thing I need is one machine which is
really fast on one platform - the other can be "done" on a slower
machine (I wouldn't complain about the Ultra5 a lot if my laptop would
be faster...). My current "weak" point is 64bit AMD64 since my laptop is
32bit-only (and trying to build anything AMD64-bit related via QEmu on a
32bit machine is hopeless). Therefore a reasonable 64bit AMD64 machine
(>= 3GHz) which can run SuSE Linux 10.1 or 10.2 (yes... I know...
blasphemy... but there isn't a VMware kernel module for Solaris...) with
enougth memory to run VMware[note_1] (e.g. >= 1GB) would be more than
enougth (I am using this VMware setup to use the "snapshot" feature of
VMware to have a way to quickly roll-back any experiments without the
need for a full rebuild of the tree).
[note_1]=(Or drop the idea with VMware and directly install Solaris
11/B65) ...
[1] would be the same as the AMD64 box above but >= 2 CPUs and 4GB
[3] would be any AMD64 box with enougth swap space to compile OS/Net
A faster SPARC may be nice, too (but see the note above) only optional.
Per above categorisation:
[1] E10k (ok, that's hopeless =:-) ), alternatively a T1000[note_2], 8GB
main memory and >= 80GB disk)
[2] Blade 1000/2000 2x900MHz, 4GB main memory, 2x18GB disks
[3] Blade 1000 1x750MHz, 1GB, 1x18GB disk
[note_2]=(the 60day "Trial" program of Sun won't work - I'll depends on
having enougth credit on the credit card (and I don't have that much
room on it))
> Please detail your requirements and I'll see what I can to to allieve
> the pain.
See above...
----
Bye,
Roland
--
__ . . __
(o.\ \/ /.o) [EMAIL PROTECTED]
\__\/\/__/ MPEG specialist, C&&JAVA&&Sun&&Unix programmer
/O /==\ O\ TEL +49 641 7950090
(;O/ \/ \O;)
_______________________________________________
opensolaris-code mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/opensolaris-code
