From: Kero van Gelder <kero@...4050.upc-d.chello.nl>
Date: 2001-09-06T02:12:52+09:00
Subject: [ruby-talk:20934] Re: Ruby/Tk core dump

> That doesn't make any sense to me.  Unless you call Tk.mainloop() no 
> processing of Tk stuff happens; in particular, the window isn't even 
> displayed and the program exits.
> 
> Perhaps I missed your point, though.

OK, never mind. Your remarks just matched the point where the crash
happens for me when I play with Ruby/Tk in >1 (Ruby) thread.

> My code runs absolutely perfectly under Linux.  I only have this problem on 
> the Sun box at uni, where I didn't install Ruby, which is why I'm guessing 
> it's a configuration issue.

Possibly; could it be a Tcl/Tk instalation problem? seems unlikely,
but still.

> > PS: the huge line is about callback/event from the Tk side of your
> > program.
> 
> Yep.  I guessed that much.  It's far too complex for me to delve further 
> into it without more knowledge of how things work, though ... and it's 
> probably pointless, since I still think the problem is some kind of 
> configuration issue.

It's daunting for me, too. In fact, I was looking into Tk bindings, to
find out why multithreaded use of Tk crashes sometimes, but doesn't in
other cases. What I wrote above is an experimental result, that should
not be there in the first place.

But it is very hard to understand what happens in multithreaded code,
as the Tk bindings are; they explicitely use that fact, but I can not
grasp why, nor can I change the code and predict its behaviour. In
other words, I'm stuck in understanding it.

Perhaps someone (matz?) could enlighten me.

+--- Kero ------------------------------ kero@chello.nl ---+
|  Don't split your mentality without thinking twice       |
|                          Proud like a God -- Guano Apes  |
+--- M38c ------- http://members.chello.nl/~k.vangelder ---+