RE: [Plplot-devel] Deadline for 5.1.0 is Monday 8 AM PST = 16:00 UT

[Geoffrey F. <fu...@ga...>]

Alan W. Irwin writes:
 > On Sat, 26 Jan 2002, Geoffrey Furnish wrote:
 > > [...]Now, you might suspect its just my code.  And that could well be.
 > > However, it runs to completion with no upset using many other devices,
 > > including xwin, ps, and plmeta.  With a metafile, I can render to
 > > device tk.  Also the plplot C examples work with dev tk.  So I realize
 > > this is a pretty inconclusive set of data.
 > 
 > The C examples with tk driver may not be adequately exercising the code.
 > Therefore, I think you should exercise every Tk example to see if there are
 > any obvious problems on your system.  Here is my script for doing this from
 > plplot/tmp:

Thanks.  The script was helpful, things seem to be working in the core
distribution.  But I agree with your suspicion that my particular
client code is probably doing things wityh the API that the examples
don't.  However, I haven't been able to pin it down yet to a single
specific problem, and realistically, do not expect to have such
closure in the next 14 hours.  So, unless someone else announces
similar problems, I think you should just go ahead.

I did manage to build my code with a completely different compiler,
just to try to eliminate the "bad code gen" prospect.  The same
failure occurs even when the client code is built with the latst gcc,
3.0.3.  So, the plot thickens.

Realistically, I think its going to take a timescale of days for me to
get to the bottom of this, so you should just proceed.  The problem
may well turn out to have nothing to do with PLplot.

More when I know more.

-- 
Geoffrey Furnish         fu...@ga...