chat.freenode.net #tryton log beginning Mon Jan 11 00:00:02 CET 2010 | ||
2010-01-11 00:18 <CIA-5> vengfulsquirrel roundup * #1372/Performance problem with sorting one2many.: | ||
2010-01-11 00:18 <CIA-5> [new] This change reduced a 10 minute delay into a 15 second delay. | ||
2010-01-11 00:18 <CIA-5> In my case the ids2 list had a length of around 32,000. | ||
2010-01-11 00:18 <CIA-5> http://coderevie ... | ||
2010-01-11 00:18 <CIA-5> http://bugs.tryton.org/roundup/issue1372 | ||
2010-01-11 00:24 <cedk> vengfulsquirrel: good | ||
2010-01-11 00:26 <cedk> vengfulsquirrel: we must use a kind of cache to search index | ||
2010-01-11 00:28 <cedk> vengfulsquirrel: oops, I did not see that you wrote a patch | ||
2010-01-11 00:29 <vengfulsquirrel> Yeah I'm not sure if there is a faster way to make a list into a dict but that worked for me. | ||
2010-01-11 00:32 <cedk> vengfulsquirrel: I made a comment | ||
2010-01-11 00:33 <cedk> vengfulsquirrel: but it is false :-) | ||
2010-01-11 00:34 <cedk> vengfulsquirrel: I'm still looking | ||
2010-01-11 00:38 <cedk> vengfulsquirrel: it is really painful that enumerate return value in this order | ||
2010-01-11 00:42 <vengfulsquirrel> dict(zip(ids2, range(len(ids2)))) ? | ||
2010-01-11 00:44 <cedk> vengfulsquirrel: not sure it is faster | ||
2010-01-11 00:45 <vengfulsquirrel> Using my state of the art datetime.now() performance testing they seem to be the same. I think actually making it doesn't take that long so maybe the zip would just be prettier? | ||
2010-01-11 00:46 <cedk> vengfulsquirrel: zip scale less than: dict((y, x) for x, y in enumerate(a)) | ||
2010-01-11 00:47 <cedk> vengfulsquirrel: from my test | ||
2010-01-11 00:47 <cedk> vengfulsquirrel: with a list of 10000 | ||
2010-01-11 00:49 <vengfulsquirrel> Oh right you mean use a generator list comprehension instead of list comprehension. | ||
2010-01-11 00:49 <cedk> vengfulsquirrel: you can propose a patch with my last comment | ||
2010-01-11 00:49 <cedk> vengfulsquirrel: yes with list you waste memory | ||
2010-01-11 00:49 <cedk> vengfulsquirrel: test script: http://dpaste.com/143420/ | ||
2010-01-11 00:51 <cedk> vengfulsquirrel: with a list of 1000000 the difference is really big | ||
2010-01-11 00:51 <vengfulsquirrel> Yeah seems like the generator is the best. | ||
2010-01-11 00:55 <cedk> vengfulsquirrel: do you know how to submit patch? | ||
2010-01-11 00:56 <vengfulsquirrel> http://codereview.appspot.com/186061/patch/1004/5 | ||
2010-01-11 00:56 <vengfulsquirrel> What do you mean ? | ||
2010-01-11 00:56 <vengfulsquirrel> I updated it | ||
2010-01-11 00:57 <vengfulsquirrel> I took out the extra line too, I can add it back though if you wanted it there. | ||
2010-01-11 00:57 <vengfulsquirrel> Or I guess I stopped adding the new line. | ||
2010-01-11 00:58 <vengfulsquirrel> *blank | ||
2010-01-11 00:58 <cedk> vengfulsquirrel: no porblem for the line | ||
2010-01-11 00:59 <cedk> vengfulsquirrel: new comment | ||
2010-01-11 00:59 <cedk> vengfulsquirrel: by submit patch, I mean in roundup when it will be finished | ||
2010-01-11 00:59 <cedk> vengfulsquirrel: to be integrated in hg with your name as author | ||
2010-01-11 01:02 <vengfulsquirrel> http://code.google.com/p/tryton/wiki/HowtoContribute | ||
2010-01-11 01:02 <vengfulsquirrel> I just follow that rihgt ? | ||
2010-01-11 01:02 <vengfulsquirrel> I usually mess it up the first 4 times though ha. | ||
2010-01-11 01:03 <cedk> vengfulsquirrel: yes | ||
2010-01-11 01:11 <cedk> ACTION going to sleep | ||
2010-01-11 01:14 <CIA-5> vengfulsquirrel roundup * #1372/Performance problem with sorting one2many.: [chatting] Attached a patch | ||
2010-01-11 01:14 <CIA-5> http://bugs.tryton.org/roundup/issue1372 | ||
2010-01-11 01:15 <vengfulsquirrel> cedk: Thanks, talk to you tomorrow. | ||
2010-01-11 04:07 -!- panthera(n=daniel@static.88-198-196-34.clients.your-server.de) has joined #tryton | ||
2010-01-11 04:08 -!- ChanServ(ChanServ@services.) has joined #tryton | ||
2010-01-11 04:08 -!- vengfulsquirrel(n=ian@c-69-181-194-95.hsd1.ca.comcast.net) has joined #tryton | ||
2010-01-11 04:08 -!- udono(n=udono@dynamic-unidsl-85-197-25-39.westend.de) has joined #tryton | ||
2010-01-11 04:08 -!- yangoon(n=mathiasb@p549F6A91.dip.t-dialin.net) has joined #tryton | ||
2010-01-11 04:08 -!- juanfer(n=juanfer@190.144.69.234) has joined #tryton | ||
2010-01-11 04:08 -!- rednul(n=rednul@host-98-127-54-98.bln-mt.client.bresnan.net) has joined #tryton | ||
2010-01-11 04:08 -!- CIA-5(n=CIA@208.69.182.149) has joined #tryton | ||
2010-01-11 04:08 -!- fil(n=phil@blue.hands.com) has joined #tryton | ||
2010-01-11 04:08 -!- sharkcz(n=dan@plz1-v-4-17.static.adsl.vol.cz) has joined #tryton | ||
2010-01-11 04:08 -!- preC(n=preCTWO@orkan.Informatik.Uni-Oldenburg.DE) has joined #tryton | ||
2010-01-11 05:18 -!- yangoon(n=mathiasb@p549F7436.dip.t-dialin.net) has joined #tryton | ||
2010-01-11 06:41 -!- sharoon(n=sharoont@89.243.21.157) has joined #tryton | ||
2010-01-11 07:33 -!- Timitos(n=timitos@88.217.184.172) has joined #tryton | ||
2010-01-11 07:39 -!- bechamel(n=user@host-85-201-159-186.brutele.be) has joined #tryton | ||
2010-01-11 08:16 -!- paepke(n=paepke@p4FEB2AF0.dip0.t-ipconnect.de) has joined #tryton | ||
2010-01-11 08:21 -!- cedk(n=ced@gentoo/developer/cedk) has joined #tryton | ||
2010-01-11 08:38 -!- enlightx(n=enlightx@static-217-133-61-144.clienti.tiscali.it) has joined #tryton | ||
2010-01-11 09:18 -!- mfladischer(n=fladisch@sh018.meduni-graz.at) has joined #tryton | ||
2010-01-11 09:56 <CIA-5> paepke roundup * #1350/Error msg on start without python or oo.org: | ||
2010-01-11 09:56 <CIA-5> freshly installed win-xp (german) | ||
2010-01-11 09:56 <CIA-5> only installed vmware-tools and cause of that windows-installer 3.1. | ||
2010-01-11 09:56 <CIA-5> nothing else. | ||
2010-01-11 09:56 <CIA-5> maybe there are depend ... | ||
2010-01-11 09:56 <CIA-5> http://bugs.tryton.org/roundup/issue1350 | ||
2010-01-11 11:25 -!- cedric_b(n=cedric@ANantes-158-1-159-244.w90-59.abo.wanadoo.fr) has joined #tryton | ||
2010-01-11 12:19 -!- udono(n=udono@dynamic-unidsl-85-197-17-24.westend.de) has joined #tryton | ||
2010-01-11 12:21 <CIA-5> Ian Wilson <ian@laspilitas.com> default * 2129:1f45f1707f9e trytond/trytond/model/fields/one2many.py: Add index to improve performance of sorting subsequences for issue1372 | ||
2010-01-11 12:21 <CIA-5> http://hg.tryton.org/trytond/rev/1f45f1707f9e | ||
2010-01-11 12:21 <CIA-5> ced roundup * #1372/Performance problem with sorting one2many.: [resolved] Applied | ||
2010-01-11 12:21 <CIA-5> http://bugs.tryton.org/roundup/issue1372 | ||
2010-01-11 12:25 <CIA-5> Mathias Behrle <mathiasb@mbsolutions.selfip.biz> default * 540:75d5dca1ac55 account/journal.py: | ||
2010-01-11 12:25 <CIA-5> Remove surplus argument for issue 1363. | ||
2010-01-11 12:25 <CIA-5> (transplanted from 78254a0c1336465b459882fa1c412c3dabde7254) | ||
2010-01-11 12:25 <CIA-5> http://hg.tryton.org/1.4/modules/account/rev/75d5dca1ac55 | ||
2010-01-11 12:25 <CIA-5> C?dric Krier <ced@b2ck.com> default * 1456:76de1b10567b tryton/tryton/gui/window/view_form/view/list_gtk/parser.py: | ||
2010-01-11 12:25 <CIA-5> Add missing import of _LIMIT | ||
2010-01-11 12:25 <CIA-5> Add missing window in WinSearch | ||
2010-01-11 12:25 <CIA-5> Fix rename go function of WinSearch into run | ||
2010-01-11 12:25 <CIA-5> (transplanted from 1ce6ee084c9678de9083be09576c923bcb9b06c8) | ||
2010-01-11 12:25 <CIA-5> http://hg.tryton.org/1.4/tryton/rev/76de1b10567b | ||
2010-01-11 12:26 <CIA-5> Mathias Behrle <mathiasb@mbsolutions.selfip.biz> default * 492:c30b32b38c6a account/journal.py: | ||
2010-01-11 12:26 <CIA-5> Remove surplus argument for issue 1363. | ||
2010-01-11 12:26 <CIA-5> (transplanted from 78254a0c1336465b459882fa1c412c3dabde7254) | ||
2010-01-11 12:26 <CIA-5> http://hg.tryton.org/1.2/modules/account/rev/c30b32b38c6a | ||
2010-01-11 12:26 <CIA-5> C?dric Krier <ced@b2ck.com> default * 1337:ff7421d8d8af tryton/tryton/gui/window/view_form/view/list_gtk/parser.py: | ||
2010-01-11 12:26 <CIA-5> Add missing import of _LIMIT | ||
2010-01-11 12:26 <CIA-5> Add missing window in WinSearch | ||
2010-01-11 12:26 <CIA-5> Fix rename go function of WinSearch into run | ||
2010-01-11 12:26 <CIA-5> (transplanted from 1ce6ee084c9678de9083be09576c923bcb9b06c8) | ||
2010-01-11 12:26 <CIA-5> http://hg.tryton.org/1.2/tryton/rev/ff7421d8d8af | ||
2010-01-11 12:30 <CIA-5> C?dric Krier <ced@b2ck.com> default * 1485:14bc8e7cc780 tryton/tryton/gui/window/view_form/view/form_gtk/textbox.py: Improve speed of textbox with gtkspell | ||
2010-01-11 12:30 <CIA-5> http://hg.tryton.org/tryton/rev/14bc8e7cc780 | ||
2010-01-11 12:32 <CIA-5> C?dric Krier <ced@b2ck.com> default * 2130:f1bfdb67faaa trytond/ (5 files in 5 dirs): Add comment on table and field for postgresql backend | ||
2010-01-11 12:32 <CIA-5> http://hg.tryton.org/trytond/rev/f1bfdb67faaa | ||
2010-01-11 12:35 <CIA-5> C?dric Krier <ced@b2ck.com> default * 1486:aeddd0e04b60 tryton/tryton/gui/window/view_form/view/form_gtk/parser.py: | ||
2010-01-11 12:35 <CIA-5> Handle always states on group for issue1366 | ||
2010-01-11 12:35 <CIA-5> Allow to override states on label linked to field for issue1367 | ||
2010-01-11 12:35 <CIA-5> http://hg.tryton.org/tryton/rev/aeddd0e04b60 | ||
2010-01-11 12:35 <CIA-5> ced roundup * #1367/<label> does not honor 'states': [resolved] Fix with changeset aeddd0e04b60 | ||
2010-01-11 12:35 <CIA-5> http://bugs.tryton.org/roundup/issue1367 | ||
2010-01-11 12:35 <CIA-5> ced roundup * #1366/<group> does not accept attribut 'states': [resolved] Fix with changeset aeddd0e04b60 | ||
2010-01-11 12:35 <CIA-5> http://bugs.tryton.org/roundup/issue1366 | ||
2010-01-11 12:41 -!- LucaSub1(n=LucaSub@host240-173-dynamic.14-87-r.retail.telecomitalia.it) has joined #tryton | ||
2010-01-11 13:44 -!- essich(n=essich@p4FCF98EA.dip0.t-ipconnect.de) has joined #tryton | ||
2010-01-11 14:00 -!- cedk(n=ced@gentoo/developer/cedk) has joined #tryton | ||
2010-01-11 14:51 -!- cedric_b(n=cedric@ANantes-158-1-159-244.w90-59.abo.wanadoo.fr) has joined #tryton | ||
2010-01-11 15:44 <CIA-5> paepke roundup * #1350/Error msg on start without python or oo.org: | ||
2010-01-11 15:44 <CIA-5> there is a solution on that site: | ||
2010-01-11 15:44 <CIA-5> http://www.py2exe.org/index.cgi/Tutorial#Step5 | ||
2010-01-11 15:44 <CIA-5> http://bugs.tryton.org/roundup/issue1350 | ||
2010-01-11 16:07 -!- sharoon(n=sharoont@89.243.21.157) has joined #tryton | ||
2010-01-11 16:27 <CIA-5> ced roundup * #1350/Error msg on start without python or oo.org: | ||
2010-01-11 16:27 <CIA-5> [testing] Could you try without installing Python nor Ooo but with | ||
2010-01-11 16:27 <CIA-5> http://www.microsoft.com/downloads/details.aspx?FamilyID=9b2da534-3e03-4391-8a ... | ||
2010-01-11 16:27 <CIA-5> http://bugs.tryton.org/roundup/issue1350 | ||
2010-01-11 16:28 -!- udono(n=udono@dynamic-unidsl-85-197-19-135.westend.de) has joined #tryton | ||
2010-01-11 16:32 <CIA-5> paepke roundup * #1350/Error msg on start without python or oo.org: | ||
2010-01-11 16:32 <CIA-5> yes, that works. sorry i didn't mention i have tested that solution. | ||
2010-01-11 16:32 <CIA-5> it should be integrated somehow in the installer or at least at the wiki pag ... | ||
2010-01-11 16:32 <CIA-5> http://bugs.tryton.org/roundup/issue1350 | ||
2010-01-11 16:35 <CIA-5> ced roundup * #1350/Error msg on start without python or oo.org: | ||
2010-01-11 16:35 <CIA-5> Could you test with this one: | ||
2010-01-11 16:35 <CIA-5> http://www.b2ck.com/~ced/tryton-setup-1.5.0-20100111.exe | ||
2010-01-11 16:35 <CIA-5> http://bugs.tryton.org/roundup/issue1350 | ||
2010-01-11 16:36 -!- sharoon(n=sharoont@89.243.21.157) has joined #tryton | ||
2010-01-11 17:19 -!- juanfer_(n=juanfer@190.27.146.29) has joined #tryton | ||
2010-01-11 17:41 -!- sharoon(n=sharoont@89.243.21.157) has left #tryton | ||
2010-01-11 19:09 -!- enlightx(n=enlightx@host-84-220-92-50.cust-adsl.tiscali.it) has joined #tryton | ||
2010-01-11 19:11 -!- vengfulsquirrel(n=ian@c-69-181-194-95.hsd1.ca.comcast.net) has joined #tryton | ||
2010-01-11 20:35 -!- LucaSub1(n=LucaSub@host153-175-dynamic.14-87-r.retail.telecomitalia.it) has joined #tryton | ||
2010-01-11 21:07 <cedk> vengfulsquirrel: what is the query that takes 5-6 seconds? | ||
2010-01-11 21:42 <vengfulsquirrel> http://paste.pocoo.org/show/164419/ | ||
2010-01-11 21:43 <vengfulsquirrel> Im just printing it from execute in database.py for postgresql. | ||
2010-01-11 21:43 -!- paepke(n=paepke@p54A2A4A2.dip0.t-ipconnect.de) has joined #tryton | ||
2010-01-11 21:45 <vengfulsquirrel> the query is not valid though do you know what happens to en_us ? | ||
2010-01-11 21:49 <vengfulsquirrel> cedk:* | ||
2010-01-11 22:19 <cedk> vengfulsquirrel: do you have an index on column inventory_line.inventory ? | ||
2010-01-11 22:19 <vengfulsquirrel> Yes, but I was thinking maybe I need one on inventory_line.product. | ||
2010-01-11 22:20 <vengfulsquirrel> "inventory_line_inventory_index" btree (inventory) | ||
2010-01-11 22:21 <cedk> vengfulsquirrel: could give the ouput of an explain of the query | ||
2010-01-11 22:21 <cedk> vengfulsquirrel: I can not run it because I don't have the right tables | ||
2010-01-11 22:21 <vengfulsquirrel> http://paste.pocoo.org/show/164451/ | ||
2010-01-11 22:23 <cedk> vengfulsquirrel: yes I think an index on inventory_line.product will improve the situation | ||
2010-01-11 22:25 <vengfulsquirrel> Oh right sorry, I already added it before I ran explain so that is with that index, do you think it needs to be a composite index? Can you tell from the explain output? It looks like gibberish to me. | ||
2010-01-11 22:26 <vengfulsquirrel> -> Hash Left Join (cost=963.30..3516.01 rows=2137 width=128) | ||
2010-01-11 22:26 <vengfulsquirrel> Hash Cond: ("outer".product = "inner".id) | ||
2010-01-11 22:26 <vengfulsquirrel> That is the join on the product right ? | ||
2010-01-11 22:26 <cedk> vengfulsquirrel: because of this line: "Seq Scan on product_product ..." | ||
2010-01-11 22:39 <cedk> vengfulsquirrel: did you see how the reduce_ids function make a better use of index | ||
2010-01-11 22:40 <vengfulsquirrel> It converts a large IN into something like this right ? http://paste.pocoo.org/show/164461/ | ||
2010-01-11 22:42 <cedk> vengfulsquirrel: yes and instead of being a test id by id, it can use the index for a range of id | ||
2010-01-11 22:44 <vengfulsquirrel> Yeah that's a good idea, is it faster in all cases? Like is there any sort of trade off between the length of the IN and the number of AND conditions ? | ||
2010-01-11 22:47 <cedk> vengfulsquirrel: yes, I set the threshold to 5 elements | ||
2010-01-11 22:48 <cedk> vengfulsquirrel: but it is en empiric value :-) | ||
2010-01-11 22:54 -!- tekknokrat(n=gthieleb@dslb-088-075-227-054.pools.arcor-ip.net) has joined #tryton | ||
2010-01-11 23:00 <vengfulsquirrel> Ha yeah sounds good | ||
2010-01-11 23:08 <yangoon> cedk setup.py of trytond still points to openoffice.interact, but openoffice-python is on pypi | ||
2010-01-11 23:19 <CIA-5> C?dric Krier <ced@b2ck.com> default * 2131:b8e52e4d3be1 trytond/setup.py: Rename openoffice.interact into openoffice-python | ||
2010-01-11 23:19 <CIA-5> http://hg.tryton.org/trytond/rev/b8e52e4d3be1 | ||
2010-01-11 23:39 -!- tekknokrat(n=gthieleb@dslb-088-075-227-054.pools.arcor-ip.net) has left #tryton |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!