Hi,
It appeared that this was a critical issue. It took a bit of time to
track it down, and Lukas patiently assisted me in the process :). The
problem was a Glamour one and it was exhibited specifically in the
case of the MoosePanel like this:
Because MoosePanel registers for the announcements from MooseModel
root, even if you close it, this instance of MoosePanel will still be
referenced by the MooseModel root announcer. So, it does not get
garbage collected. And if you opened already a model with this
MoosePanel, this model will still be referenced by the Panel, and so
it won't get collected.
The current solution is that when the Glamour Morphic window is
closed, the announcements are unregistered from all enclosed
presentation.
So, it means that in the case of MoosePanel, the garbage collector
will do its job only after you close the MoosePanel. It's not nice,
but it is the current solution.
See
http://code.google.com/p/moose-technology/issues/detail?id=314. I
will keep it open for a while. Please let me know if you find more
leakage.
Cheers,
Doru
On 27 Jan 2010, at 00:23, Tudor Girba wrote:
Hi Fabrizio,
You could probably reproduce this problem in a new smaller image as
well. No need to spend the time doing it on a very large one.
Cheers,
Doru
On 26 Jan 2010, at 18:51, Fabrizio Perin wrote:
> Hi,
> i tried to check the pointers to the model opening that menu. My
> machine start to work at 100% and after 3 hours i didn't have the
> list of pointers visualized yet. So i decide to kill the image. The
> system was surely into an infinite loop (or there were billions of
> links so i think that is useless know which they were). I will
> anyway try to let my machine work during the night just in the case
> that i'm wrong :) Meanwhile have you other suggestions? Thanks in
> advance.
>
> Cheers,
>
> Fabrizio
>
>
> On 26 Jan 2010, at 12:20, Alexandre Bergel wrote:
>
>> You can maybe check who is pointing to the model. There is a menu
>> in the inspector.
>>
>> Alexandre
>>
>>
>> On 26 Jan 2010, at 08:09, Fabrizio Perin wrote:
>>
>>> Hi,
>>> yesterday I tried to delete a moose model from my image using
>>> the function Utilities>>Delete on the moose panel. The model
>>> disappear from the list but saving the image it has the same
>>> size. Executing "MooseModel allInstances" I saw that the model
>>> was still there. I tried to close all windows and then i execute
>>> "Smalltalk garbageCollect" and again the saved image has the same
>>> size because the model has not been removed.
>>>
>>> I'm working on the pharo dev image 10508, Moose-Core 215 but i
>>> detect the same error in an older image: pharo dev image 10502,
>>> Moose-Core 208.
>>>
>>> Cheers,
>>>
>>> Fabrizio
>>>
>>>
>>>
>>>
>>>
>>> _______________________________________________
>>> Moose-dev mailing list
>>> Moose-dev(a)iam.unibe.ch
>>>
https://www.iam.unibe.ch/mailman/listinfo/moose-dev
>>
>> --
>> _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:
>> Alexandre Bergel
http://www.bergel.eu
>> ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
>>
>>
>>
>>
>>
>> _______________________________________________
>> Moose-dev mailing list
>> Moose-dev(a)iam.unibe.ch
>>
https://www.iam.unibe.ch/mailman/listinfo/moose-dev
>
> Fabrizio Perin
> Institut fuer Mathematik und Informatik
> University Bern, IAM-SCG
> Neubrueckstrasse 10
> CH-3012 Bern, Switzerland
> Tel: +41 31 631 33 13
> FAX: +41 31 631 33 55
>
> _______________________________________________
> Moose-dev mailing list
> Moose-dev(a)iam.unibe.ch
>
https://www.iam.unibe.ch/mailman/listinfo/moose-dev
--
www.tudorgirba.com
"In a world where everything is moving ever faster,
one might have better chances to win by moving slower."
--
www.tudorgirba.com
"The coherence of a trip is given by the clearness of the goal."