Re: [MTT users] Actual releases?

2015-12-10 Thread Gilles Gouaillardet
+1 too Ralph, just to be clear ... from a mtt point of view, do you expect only a release tag and a tarball ? or do you also expect a .spec file (iirc, OpenHPC is RPM based even if they use their own .spec file) Cheers, Gilles On 12/10/2015 11:34 PM, Jeff Squyres (jsquyres) wrote: +1. We did

[MTT users] Python client

2015-12-10 Thread Ralph Castain
Hey folks I'm working on the Python client and it is coming along pretty well. The code is completely separate from the Perl-based client and doesn't interact with it, so I would like to push it into the repo on an on-going basis so others can look at it and comment as I go rather than hold it unt

Re: [MTT users] Actual releases?

2015-12-10 Thread Jeff Squyres (jsquyres)
+1. We did do a release branch at one point (which is likely completely stale), but we've never invested any release engineering efforts into MTT to make a packaged/perfect tarball -- mainly because, at least in the beginning, MTT was just for Open MPI core developers, and we were comfortable t

Re: [MTT users] Actual releases?

2015-12-10 Thread Josh Hursey
I think that would be good. I won't have any cycles to help until after the first of the year. We started working towards a release way back when, but I think we got stuck with the license to package up the graphing library for the MTT Reporter. We could just remove that feature from the release si