IRC logs of #tryton for Friday, 2012-05-04

chat.freenode.net #tryton log beginning Fri May 4 00:00:02 CEST 2012
2012-05-04 11:34 -!- scrapper(~scrapper@188-23-0-25.adsl.highway.telekom.at) has left #tryton
2012-05-04 19:13 <smarro> cedk: ping
2012-05-04 19:15 <cedk> smarro: pong
2012-05-04 19:16 <smarro> cedk: you know if something happens with codereview.tryton.org?
2012-05-04 19:16 -!- Mayank1(~mayank@122.162.78.157) has left #tryton
2012-05-04 19:16 <smarro> cedk: i get a 500 Server Error
2012-05-04 19:17 <cedk> smarro: yes, we exceed the number of read
2012-05-04 19:18 <smarro> cedk: ah, ok
2012-05-04 19:18 <cedk> smarro: need to wait for the reset counter of Google
2012-05-04 19:18 <cedk> smarro: it is one of the reasons, I would like to migrate to our own host
2012-05-04 19:19 <smarro> cedk: yes, good idea
2012-05-04 19:20 <cedk> smarro: yes but nobody works on it
2012-05-04 19:21 <smarro> cedk: you are talking about the thread "Being less Google dependent"?
2012-05-04 19:21 <cedk> smarro: yes
2012-05-04 19:26 <cedk> smarro: indeed there was a bot from NerdByNature that started crowling the website without respecting robots.txt
2012-05-04 19:28 <smarro> cedk: is there a roundup bug created for the exporter that has to be written?
2012-05-04 19:28 <cedk> smarro: I think
2012-05-04 19:29 <smarro> cedk: because i can't find it
2012-05-04 19:30 <cedk> smarro: https://bugs.tryton.org/issue2179
2012-05-04 19:33 <smarro> cedk: ok, i will take a look
2012-05-04 19:34 <cedk> smarro: I think I have explain on the mailing list the first step
2012-05-04 19:35 <smarro> cedk: yes, you said "So it is needed to write an exporter for each objects like explain here"
2012-05-04 19:35 <smarro> cedk: https://code.google.com/appengine/docs/python/tools/uploadingdata.html
2012-05-04 19:39 <cedk> smarro: yes, to write the exporter, you need to look what will be the postgresql schema of reitveld with django
2012-05-04 19:41 <cedk> smarro: if you work on it, I can give you access to GAE
2012-05-04 19:44 <smarro> cedk: ok, i will find time to work on this
2012-05-04 19:52 <cedk> smarro: the repo of codereview: http://hg.tryton.org/codereview.tryton.org/
2012-05-04 19:58 <smarro> cedk: that repo is a hg changeset patch only... correct?
2012-05-04 20:00 <cedk> smarro: yes
2012-05-04 20:00 <smarro> cedk: ok, i see the changes
2012-05-04 20:06 <cedk> smarro: I have also create a repository for roundup: http://hg.tryton.org/bugs.tryton.org/
2012-05-04 20:09 <smarro> cedk: ok, i was looking at the roundup and rietveld integration at Python project...
2012-05-04 21:13 <__efx__> hello everybody, I am creating a personal module for
2012-05-04 21:14 <__efx__> is it possible to add default value for the field ?
2012-05-04 21:20 <__efx__> for example I have in my model a date field and I want to set it to the actual date, so that when somebody creates a new record the value is by default "today". I also want the user to be able to modify it if he wants
2012-05-04 21:21 <cedk> __efx__: you must define a method with the name: "default_<field name>"
2012-05-04 21:23 <cedk> __efx__: http://doc.tryton.org/2.4/trytond/doc/ref/models/models.html#trytond.model.Model.default_get
2012-05-04 21:25 <__efx__> thanks
2012-05-04 23:37 <cedk> got unittest passed on the ActiveRecord patch \o/
2012-05-04 23:42 <bechamel> cedk: about the cache issue: I don't see in which situation it may be a problem
2012-05-04 23:42 <cedk> bechamel: with or without my proposal?
2012-05-04 23:42 <bechamel> cedk: without
2012-05-04 23:44 <cedk> bechamel: if you don't see why having local cache never clean is not a problem, I don't know how to explain
2012-05-04 23:45 <bechamel> cedk: the cache is obviously clean at the end of the transaction, no ?
2012-05-04 23:45 <cedk> bechamel: this is not enough
2012-05-04 23:46 <cedk> bechamel: it must be in sync with the DB
2012-05-04 23:48 <bechamel> cedk: ok; so a problematic situation is updating a sale line and then reading the sale amount ?
2012-05-04 23:48 <cedk> bechamel: yes
2012-05-04 23:50 <cedk> bechamel: it is clear that it must be cleared between each method calls
2012-05-04 23:50 <cedk> bechamel: the problem is how
2012-05-04 23:54 <bechamel> cedk: to be sure to understand everything: currently we don't have this cache problem because we re-read the db each time a browse record is instanciated ?
2012-05-04 23:54 <cedk> bechamel: yes
2012-05-04 23:57 <albertca> cedk: but isn't the cache invalidated whenever there's a write in that model?
2012-05-04 23:57 <bechamel> cedk: just thinking out loud: and what if we decide that it's up to the developper to tell when to clean the cache ?
2012-05-04 23:58 <cedk> albertca: yes but only the global cache, not the local one of the browserecord
2012-05-04 23:58 <bechamel> albertca: the problem is when the cache content depends of values coming from other models
2012-05-04 23:58 <cedk> bechamel: I think it is bad because to keep modularity you can not rely on the behavior of other methods
2012-05-04 23:59 <cedk> bechamel: so as I said in the email, we will finish by clearing cache at the begining of every methods

Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!