On Fri, Mar 30, 2001 at 01:39:21PM +0930, Brett Lymn wrote:
> According to ctest_at_neural.dlsemc.com:
> >
> >open_log failed.
> >do_rvm_options failed
> >13:33:56 rvm_init failed RVM_EIO
> >
> 
> Jan's comments caught my eye about NetBSD - if you _are_ using NetBSD
> then be aware that the coda that is currently in the package system
> may not recognise NetBSD 1.5 and will not set up the correct offsets
> for the RVM mmap.  You may need to check the setup scripts and make
> sure the case that follows the "uname" call contains a case for 1.5.
> 
> FWIW I have made an attempt at updating the coda package but I have
> fell foul of some libtool weirdness which I need to resolve.

I didn't know there was an offset problem with NetBSD 1.5. Looking at
the NetBSD-current tree, I see no patches related to RVM offsets either.

As far as the patches in the tree, it looks like patch-aa doesn't even
apply to the used tree anymore, filcon is part of the liblwp package.
Patch-ab is interesting, we should probably test for the ncurses.h
header using autoconf rather than using ifdefs on a per platform basis.
It looks like patch-ac already got applied to our tree. Hmm, and the
version that is being built (coda-5.3.5) is already over a year old.

Jan