1. Ctl-Break Crashes
Paul Martin <dmnetr at gte.net> writes (on alt.lang.euphoria):
:I use the ee.ex editor that is posted at euphorias main web site. It is
an
:excellent editor except for one thing. The problem is that when I am
writing
:a program with this editor under Windows 95 and I use CTRL-BREAK in
:conjunction with the abort(i), it crashes Windows 95 and some times
reboots
:my computer, even Crash Shield does not stop it. Is there another way
EASY
:way of exiting a program without using abort(i) or crashing my
computer.
:Thanks
:Paul
It crashes mine as well. Any ideas?
Irv
2. Re: Ctl-Break Crashes
>It crashes mine as well. Any ideas?
It crashed mine, too, just till recently. There are just two things that I
can think of that fixed the problem, but I don't know for sure. Suddenly the
problem was gone.
1. Use the most recent version of EE.EXE (it _could_ be that Cuny
accidentally solved the problem)
2. I've installed "Windows 95 Service Pack 1". Don't remember where I've got
it from, think it was a mag CD.
I can give NO garantie that it would work, but my system does not crash
anymore.
Einar Mogen
3. Re: Ctl-Break Crashes
Einar Mogen writes:
> It crashed mine, too, just till recently. There are just two things
> that I can think of that fixed the problem, but I don't know for sure.
> Suddenly the problem was gone.
That's good if it's ok now on your system.
Here's what I posted yesterday on alt.lang.euphoria:
Paul Martin writes:
> I use the ee.ex editor that is posted at euphorias main web site.
> It is an excellent editor except for one thing. The problem is
> that when I am writing a program with this editor under
> Windows 95 and I use CTRL-BREAK in conjunction with
> the abort(i), it crashes Windows 95 and some times reboots
> my computer, even Crash Shield does not stop it. Is there
> another way EASY way of exiting a program without using abort(i)
> or crashing my computer.
There's a known problem (see TROUBLE.DOC)
with hitting control-C or control-Break
when you run your program from ed. I assume the same
problem holds for ee (you might want to ask David Cuny
who wrote ee). The solution is to exit the editor and run
your program normally, from the command-line, whenever you
might want to use control-C or control-Break.
Remember that with ex.exe, control-C/control-Break will only
stop your program at a point where keyboard input
is being requested. It won't necessarily get you out of
an infinite loop. (With exw.exe it *will* get you out.)
There shouldn't be any problem with using abort()
in the absence of control-C/control-Break.
Regards,
Rob Craig
Rapid Deployment Software
http://members.aol.com/FilesEu/
4. Re: Ctl-Break Crashes
Robert wrote:>There's a known problem (see TROUBLE.DOC)
>with hitting control-C or control-Break
>when you run your program from ed. I assume the same
>problem holds for ee (you might want to ask David Cuny
>who wrote ee). The solution is to exit the editor and run
>your program normally, from the command-line, whenever you
>might want to use control-C or control-Break.
It's an effective way to toast your machine, all right. I got the vertical
pinstripe pattern on my screen (indicitive of a runaway loop filling memory
with the same value), quickly followed by a Windows error, and my machine
locking up.
EE uses a non-standard method of reading the keyboard. It looks in the
circular keyboard buffer before attempting to read from it. If the next
character is a Ctrl-C, it zaps the buffer, and returns with a Ctrl-C -
without actually *reading* the buffer. Otherwise, it tries to read the
buffer normally.
The result of this is that Ctrl-C is never processed by Euphoria, so none of
the traps can happen.
If someone were able to supply me with the scancode and/or keycode of
Ctrl-Break, I could easily add the same sort of logic for the Ctrl-Break,
and try to prevent these kinds of crashes.
I'll see if I can determine the value of Ctrl-Break myself. If I can, I'll
send Robert a fix.
However, if Ctrl-Break causes some sort of hardware interrupt to do it's
dirty work, I don't know that it would make any difference.
Thanks!
-- David Cuny
