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

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

Geoffrey Furnish writes:
 >  > No idea.. I've never seen this.  BTW it may be informative to run including
 >  > the -debug option.  Then again it may not help.
 > 
 > I've demonstrated that the behavior is the same with snapshots from
 > 1/1/02 and from 9/1/01.  Therefore, I think it is clear that it is at
 > least not a recently introduced bug in PLplot.  Whether or not it is a
 > bug in PLplot at all, is yet to be determined, it may still be in my
 > code.  However, I haven't been able to find the cause yet.  Somehow it
 > seems that when running the tk driver, somehow my code cores and the
 > message is coming from plserver, saying the host app is dead.  When I
 > run the host code in the debugger, it sometimes claims to have ended
 > with a SIGPIPE.  Doesn't make any sense to me.  yet.  Using any other
 > device seems to work fine.

No, that's wrong.  It's plserver that's dieing, leaving the host app
up, but unable to communicate with it.  That explains why the host app
gets a sigpipe.  So now that makes sense.  But I don't understand
what's killing plserver.

Is there an easy way to run an app against a seperately started
plserver?  I seem to remember traffic on this list about this, but I
wasn't paying enough attention then to remember how to do it now.

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