Squeaking trouble

Donald Major Donald.Major at sas.com
Mon Nov 14 18:14:38 UTC 2005


Sorry, but you missed one of the key facts -- ALT-. DOESN'T interrupt
the loop -- once I reproduced the endless loop condition, I spent over a
minute trying ALT-. and later even CTL-. to try and interrupt the
process -- it didn't respond.  When I took the "extreme" measure of
killing the Squeak X window, which usually catches even this phenomena,
causing Squeak itself to shutdown within another second or two, this
particular runaway process STILL kept chugging along -- I actually had
to kill the Squeak process via an OS shell-issued kill command.  The
suspected, unreliable bug mentioned in someone else's post regarding
ALT-. behavior not always being 100% operational IS correct.

That said, I still like the idea of changing size to signal subclass
responsibility, as Andreas suggested (not that I'm the one to do it --
I'm still too far back on the learning curve regarding the current
mechanisms for entering changes, no matter how minor, to the system).

Bert Freudenberg wrote:
> Am 14.11.2005 um 16:13 schrieb Bob.Cowdery at CGI-Europe.com:
> 
>> x := SequenceableCollection new, then try to print x. Lock-up +  100%
>> CPU. The only response is from the main window, quit without  saving
>> then kill squeak as it's still running.
> 
> As others pointed out, Alt-. would have interrupted that "endless"  loop
> quite nicely.
> 
> SequenceableCollection is an abstract class and thus is not meant for 
> direct instantiation. Which is not a problem at all because you would 
> find out very soon if you try to use it. Which in fact you did ;-)
> 
> - Bert -
> 
> 
> 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Donald.Major.vcf
Type: text/x-vcard
Size: 272 bytes
Desc: not available
Url : http://lists.squeakfoundation.org/pipermail/squeak-dev/attachments/20051114/bfbbe50f/Donald.Major.vcf


More information about the Squeak-dev mailing list