Date: Sun, 17 Oct 2010 16:09:20 -0400 (EDT)
From: Jun-Hui Zhao 
To: Peter Teuben 
Subject: Re: miriad beta trouble

     Thanks, Peter.
     Jun-Hui


On Sun, 17 Oct 2010, Peter Teuben wrote:

>   one technique you can try is overlaying a tar file on top of a fresh
> miriad CVS tree, and use tkdiff on the files that show either 'U' or 'M'
> (or 'C' :-),
> that's what I would do.
>
>
> On 10/17/2010 03:56 PM, Jun-Hui Zhao wrote:
> >
> > Hi Peter,
> >        Good suggestion.
> >        This is actually the next thing that we plan to do in merging
> > SMA Miriad back to Miriad-CVS after we sort out our internal problems.
> >        Jun-Hui
> >
> >
> >
> > On Sun, 17 Oct 2010, Peter Teuben wrote:
> >
> >>    since we're using CVS, another suggestion I would have for next time,
> >> is to actually use the branching feature of CVS
> >> It is trivial in CVS to branch, do the development and testing in the
> >> branch, and then merge back. It simplified
> >> comparing two version of the trunk and the branch etc.etc.
> >> peter
> >>
> >> On 10/17/2010 02:47 PM, Dan Marrone wrote:
> >>> Is there an active end to end test effort? I'm using the code and it's
working, I've reported bugs and you've fixed them very quickly. It seems to me
to be fairly functional, and considering that the non-beta version of SMA
miriad is completely useless for almost all current SMA data, keeping it intact
doesn't seem like a very important priority. Furthermore, my experience with
the current CVS version of smalod, at least as it is accessed through the CVS
miriad release (rather than the tarball you post on the SMA website) is that it
is not functional for reading SMA data. So there is actually not very much to
lose by doing the hard work of integrating with the official release, and a lot
to gain, including the utility for the very large number of astronomers who now
use macs, and the greater ease of install with the miriad CVS release.
> >>>
> >>> I will continue to be testing this, and can officially do so if you tell
me how to be official, but my unofficial results are that it works pretty well
now. To me, the highest priority seems to be merging SMA miriad into the rest
of miriad, just as ANTF is. What do you think?
> >>>
> >>>
> >>> On Oct 17, 2010, at 12:26 PM, Jun-Hui Zhao wrote:
> >>>
> >>>> Hi Dan,
> >>>> We are doing so right now(?). We need to debug the beta with the testing
> >>>> team's feedback from end2end tests and tell us it is OK to replace the
> >>>> standard SMA miriad in CVS with the beta.
> >>>>         Jun-Hui
> >>>>
> >>>> On Sun, 17 Oct 2010, Dan Marrone wrote:
> >>>>
> >>>>> Hi Junhui,
> >>>>>
> >>>>> I think that the key to getting this to work for mac and everything
else is to get the sma beta stuff into the main CVS distribution. once that
happens, everything else will be fine. Is there a reason that this should
remain beta at this point?
> >>>>>
> >>>>> Dan
> >>>>>
> >>>>> On Oct 17, 2010, at 10:29 AM, Jun-Hui Zhao wrote:
> >>>>>
> >>>>>> Hi Dan and Peter,
> >>>>>>      Thanks for exploring the SMA Miriad on Mac. We now can
> >>>>>> be justified to request a Mac for the increasing Mac users.
> >>>>>> Please let me know if you have solutions for the Mac compilation.
> >>>>>> I recalled we had users' help to fix the Mac problems for the old SMA
> >>>>>> Miriad. Here are the links,
> >>>>>>
http://www.cfa.harvard.edu/sma/miriad/manuals/SMAuguide/smauserhtml/install.html
> >>>>>>      One of tips there is linking to Peter's page
> >>>>>>      Jun-Hui
> >>>>>>
> >>>>>> On Sun, 17 Oct 2010, Peter Teuben wrote:
> >>>>>>
> >>>>>>> I had some code with that duplicate symbols via an
> >>>>>>>
> >>>>>>>      int foobar;
> >>>>>>>
> >>>>>>> in .h file that was included multiple times.... We solved it this
way:
> >>>>>>>
> >>>>>>>
> >>>>>>> In a defs.h file we would say
> >>>>>>>