[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: IDL crashing badly on my solaris 8 system



Eureka!

Well, I had the notion that it might possibly have something
to do with hardware rendering (which I didn't realize was
turned on by default).

Once I figured out how to toggle to sofware rendering
(in IDLDE preferences) BINGO,  it worked beautifully.

So, 2 questions:

1) is there a similar way to turn off hardware rendering
   from the IDL prompt (not in IDLDE mode)?

2) what can I do to get hardware rendering working on this machine?  and is
   it worth it?

OK, that's really 3 questions.

Thanks a bunch.

-Jonathan

Jonathan Joseph wrote:
> 
> Still crashing the CDE session - but I now have more info.
> 
> Both Peter Clinch and RSI very promptly (thank you) suggested that
> I select C Posix from the language menu at the login screen.
> Unfortunately, this did not solve my problem.
> 
> It turns out that the font error messages may have been
> a red herring.  After questionnig the student, I
> discovered that his problem resulting in a bunch of font error messages
> followed by a crash of IDL, and the IDL crash that crashed the whole CDE
> session were two distinct problems.  Perhaps the C-Posix idea
> will fix font problem crash, but I'll worry about that later.
> 
> We found that we could easily reproduce the CDE session crash by
> typing the following command:
> 
> IDL> xplot3d,[1],[1],[1]
> 
> So, I did a little digging to see if I could find the
> exact line in xplot3d.pro that caused the crash.  And I could.
> It turned out to be the line:
> 
> oWindow->Draw, oScene
> 
> I then tried running one of my own object graphics programs, and
> it also killed the CDE session (presumably at the draw command,
> but I didn't verify that).
> 
> So, the problem seems closely tied to using object graphics.
> Can anyone out there make an educated guess as to how this
> might result in a crash of the CDE session?
> 
> I've crashed the session using either IDL 5.3 or 5.4, so if it's
> an IDL bug it's not a new 5.4 problem.
> 
> Thanks for any input.
> 
> -Jonathan