[Bioperl-l] throwing in the towel

[Jason Stajich]

I've done all I can do today, was really hoping that we could have
something ready to release on March 1, don't know if that can be met now.
  
So far I have tested on IRIX, Solaris, FreeBSD, and Linux and all things
run well.

3 sticking points which I can actually replicate on WinNT.  These all seem
to be IO related.  

a) Tempfile creation is failing on WinNT.  I don't know why.  It only
happens in strangly particular cases, but only 1/2 the tests pass on
t/Tempfile.t.  We tried really hard to be generic here and write the code
as general as possible, but I see that we still didn't clear that hurdle.

b) Index.t is failing on Bio::Index::AbstractSeq fetch.  I can't find a
reason for this other than something to do with file handles or the DBM
stuff.  It fails on a retrieve though and since it works on UNIX I am
inclined to think that the DBM implementation is not quite up to snuff on
NT?

c) GDB.t doesn't seem to be very happy.  Again I cannot seem to trace the
problem, I attempted to make the module fairly generic, but I see that
didn't actually work either.  I'm fine with dropping it from the release
or at least disabling the tests if we cannot fix it.  I'd rather it not
hold up the release though. 



-Jason

Jason Stajich
jason@chg.mc.duke.edu
Center for Human Genetics
Duke University Medical Center 
http://www.chg.duke.edu/