[1222] in NetBSD-Development
Re: List of problems
daemon@ATHENA.MIT.EDU (Greg Hudson)
Sat Feb 3 03:08:12 1996
To: Yoav Yerushalmi <yoav@MIT.EDU>
Cc: netbsd-dev@MIT.EDU
In-Reply-To: Your message of "Sat, 03 Feb 1996 02:56:32 EST."
<199602030756.CAA04901@i-see-everything-twice.MIT.EDU>
Date: Sat, 03 Feb 1996 03:07:41 EST
From: Greg Hudson <ghudson@MIT.EDU>
> We can either upgrade to 5.001e, or go back to 5.000, but we
> can't leave it in this state. (also, sys/file.ph is in one way or
> another broken.. but that's minor).
sys/file.ph is the same under 5.000 and 5.001e, and there's been no
change to h2ph to make it different; apparently the warnings you get
trying to evaluate it come from a change in the perl parser.
> 2) We shouldn't have a program named "xlocal". If you want to follow
> athena model, you should use "mkserv xlocal" or something
> similar. In fact, maybe we should create our own mkserv and have it
> use /afs/sipb/project/mkserv for its services? that way, we could
> get linux, netbsd and other services without getting ops support?
I did in fact look at mkserv, and I concluded that it would be too
difficult to fit xlocal into the mkserv model as things stand. I also
felt that because we aren't doing release engineering on /usr/X11R6,
the Athena model need not apply to our practices in that area. (The
primary benefit of mkserv is that your local changes are propagated
through releases; if we don't do releases of /usr/X11R6, there's very
little point.)
I agree that we need our own services directory.
Right now I need to work out how to do a patch release, since we
haven't done that before and don't have the infrastructure in place
(except for the necessary hooks in reactivate.) Items for the next
patch release include:
emacs (currently in /srvd/patch)
xdsc (we should get it to work first, though)
perl 5.001e
adding ahe* devices to the kernel configuration
removing the IRQ 11 ed* device from the kernel configuration
(it causes too many machines to fail to boot)
