24 Jun
2009
24 Jun
'09
6:27 p.m.
Dear Moosers
As you may know, identity hashes attributed by the virtual machine
ranges up to 4096. This may be problematic when storing objects in a
dictionary. We often have to deal with high number of entities, and
using dictionary is out of question for performance reason.
One solution (inspired from Lukas) would be to use the ID that each
MooseEntity has instead of the hash.
I will try to see whether hash can be overriden in MooseEntity to
return its ID instead.
Will let you know.
Cheers,
Alexandre
--
_,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:
Alexandre Bergel http://www.bergel.eu
^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
24 Jun
24 Jun
8:31 p.m.
Alex,
You might consider using the Hash analysis tool to see if your new
hash function is better.
It's available on the Cincom store (ok... you need to try that in
Visualworks...)
cheers
Johan
On 24 Jun 2009, at 18:27, Alexandre Bergel wrote:
Dear Moosers
As you may know, identity hashes attributed by the virtual machine
ranges up to 4096. This may be problematic when storing objects in a
dictionary. We often have to deal with high number of entities, and
using dictionary is out of question for performance reason.
One solution (inspired from Lukas) would be to use the ID that each
MooseEntity has instead of the hash.
I will try to see whether hash can be overriden in MooseEntity to
return its ID instead.
Will let you know.
Cheers,
Alexandre
--
_,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:
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
----------------------------
Johan Brichau
johan.brichau(a)uclouvain.be
10:50 p.m.
yes this looks like a good idea.
but the hash function should spread well too I guess
On Jun 24, 2009, at 6:27 PM, Alexandre Bergel wrote:
Dear Moosers
As you may know, identity hashes attributed by the virtual machine
ranges up to 4096. This may be problematic when storing objects in a
dictionary. We often have to deal with high number of entities, and
using dictionary is out of question for performance reason.
One solution (inspired from Lukas) would be to use the ID that each
MooseEntity has instead of the hash.
I will try to see whether hash can be overriden in MooseEntity to
return its ID instead.
Will let you know.
Cheers,
Alexandre
--
_,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:
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
25 Jun
25 Jun
10:56 a.m.
I do not know whether the fact that hash values is increasing by 1 for
each element is a problem or not.
Best is to try. I will give a shoot soon.
Cheers,
Alexandre
On 24 Jun 2009, at 22:50, Stéphane Ducasse wrote:
yes this looks like a good idea.
but the hash function should spread well too I guess
On Jun 24, 2009, at 6:27 PM, Alexandre Bergel wrote:
--
_,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:
Alexandre Bergel http://www.bergel.eu
^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
Dear Moosers
As you may know, identity hashes attributed by the virtual machine
ranges up to 4096. This may be problematic when storing objects in
a dictionary. We often have to deal with high number of entities,
and using dictionary is out of question for performance reason.
One solution (inspired from Lukas) would be to use the ID that each
MooseEntity has instead of the hash.
I will try to see whether hash can be overriden in MooseEntity to
return its ID instead.
Will let you know.
Cheers,
Alexandre
--
_,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:
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
_______________________________________________
Moose-dev mailing list
Moose-dev(a)iam.unibe.ch
https://www.iam.unibe.ch/mailman/listinfo/moose-dev 
26 Jun
26 Jun
8:45 a.m.
On Jun 25, 2009, at 10:56 AM, Alexandre Bergel wrote:
I do not know whether the fact that hash values is
increasing by 1
for each element is a problem or not.
Best is to try. I will give a shoot soon.
You might want to use Hashtable from squeaksource, take a look at the
wiki page.
Could be enough if the number of entities is not too large.
Stephan
8:57 a.m.
Yes this is something on my todo for some times now.
Stef
On Jun 26, 2009, at 8:45 AM, Stephan Eggermont wrote:
On Jun 25, 2009, at 10:56 AM, Alexandre Bergel wrote:
I do not know whether the fact that hash values
is increasing by 1
for each element is a problem or not.
Best is to try. I will give a shoot soon.
You might want to use Hashtable from squeaksource, take a look at
the wiki page.
Could be enough if the number of entities is not too large.
Stephan
_______________________________________________
Moose-dev mailing list
Moose-dev(a)iam.unibe.ch
https://www.iam.unibe.ch/mailman/listinfo/moose-dev
1:16 p.m.
Ok, I will check
Cheers,
Alexandre
On 26 Jun 2009, at 08:45, Stephan Eggermont wrote:
On Jun 25, 2009, at 10:56 AM, Alexandre Bergel wrote:
I do not know whether the fact that hash values
is increasing by 1
for each element is a problem or not.
Best is to try. I will give a shoot soon.
You might want to use Hashtable from squeaksource, take a look at
the wiki page.
Could be enough if the number of entities is not too large.
Stephan
_______________________________________________
Moose-dev mailing list
Moose-dev(a)iam.unibe.ch
https://www.iam.unibe.ch/mailman/listinfo/moose-dev 
1:30 p.m.
Apparently, it does not load in Pharo. The code is not closure aware.
I will contact Philip and let you know.
Cheers,
Alexandre
On 26 Jun 2009, at 08:45, Stephan Eggermont wrote:
On Jun 25, 2009, at 10:56 AM, Alexandre Bergel wrote:
--
_,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:
Alexandre Bergel http://www.bergel.eu
^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
I do not know whether the fact that hash values
is increasing by 1
for each element is a problem or not.
Best is to try. I will give a shoot soon.
You might want to use Hashtable from squeaksource, take a look at
the wiki page.
Could be enough if the number of entities is not too large.
Stephan
_______________________________________________
Moose-dev mailing list
Moose-dev(a)iam.unibe.ch
https://www.iam.unibe.ch/mailman/listinfo/moose-dev
2:27 p.m.
what is the license because it should be MIT and in that case the
closure problem should just be a store in a block arg:
On Jun 26, 2009, at 1:30 PM, Alexandre Bergel wrote:
Apparently, it does not load in Pharo. The code is not
closure aware.
I will contact Philip and let you know.
Cheers,
Alexandre
On 26 Jun 2009, at 08:45, Stephan Eggermont wrote:
On Jun 25, 2009, at 10:56 AM, Alexandre Bergel wrote:
--
_,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:
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 I do not know whether the fact that hash values
is increasing by 1
for each element is a problem or not.
Best is to try. I will give a shoot soon.
You might want to use Hashtable from squeaksource, take a look at
the wiki page.
Could be enough if the number of entities is not too large.
Stephan
_______________________________________________
Moose-dev mailing list
Moose-dev(a)iam.unibe.ch
https://www.iam.unibe.ch/mailman/listinfo/moose-dev
5411
days inactive
5413
days old
8 comments
6 participants
participants (6)
-
Alexandre Bergel -
Alexandre Bergel -
Alexandre Bergel
-
Johan Brichau -
Stephan Eggermont -
Stéphane Ducasse