>>>> CONFIGURE_ARGS+= --enable-contrib-no-build=vt
>>> I have no idea how NetBSD go about resolving such clashes in the long
>>> term though?
>> I've disabled it the same way for this time, my local package differs
>> from what's in wip:
>> --- PLIST 3 Dec 2009 10:18:00 -0000 1.5
>> +++ PLIST 9 Dec 2009 08:29:31 -0000
>> @@ -1,17 +1,11 @@
>> @comment $NetBSD$
> I am surprised that you are still installing binaries and other files
> with the -vt extension after disabling the vt stuff ?
I don't commit that part since I consider it my own local problems
that I have conflicting package. Other people may have none.
You can add CONFIGURE_ARGS and regenerate PLIST in a regular way.
>> I can commit my development patches into wip right now,
>> if that helps you.
> If your stuff now works then that's ideal. If your build is still
> failing after applying my patches then probably not.
> Given that we have something that does work, it would make sense
> to try and merge the two as far as possible before proceeding any
Benchmark I use to test MPICH and OpenMPI (parallel/skampi) doesn't work
for me still. It may be that I have somewhat unusual network configuration,
I'm looking at it.
> As discussed before, there is no real reason to have two getifaddrs
> loops seperating out IPv6 and non-IPv6 - that could all be in one
Sure, I think that we can do it a bit later.
>> Some patches should be there anyway, since OpenMPI doesn't help with
>> installation of configuration files into example directory anyway.
> OK, as you are the person within the NetBSD community looking
> after OpenMPI, I'll happily work with whatever is in the NetBSD
> repository and patch locally as needed, because I have poeple here
> who want to use stuff that requires OpenMPI now.
> Are you going to upgrade the NetBSD port to build against OpenMPI 1.4
> now that it available ? Might be a good time to check the fuzzz in the
> existing patches.