IRC logs of #tryton for Sunday, 2013-03-03 #tryton log beginning Sun Mar 3 00:00:02 CET 2013
-!- ciupicri(~ciupicri@unaffiliated/ciupicri) has left #tryton08:20
jonanIs this a possible bug ? In v2.6 when receiving an incoming supplier shipment and foreign currency exchange is involved the foreign currency amount is used in the account moves instead of company currency so inventory accounts are off.11:45
cedkjonan: probably a bug13:15
jonanjust submitted bug 303613:23
cedkjonan: I have wroten a patch13:45
jonancedk: Just tested out the patch. Working now. Thanks.14:22
albertcacedk: In the constraint changes patch we cannot use record's rec_name when rising the exception14:36
albertcacedk: because rec_name may try to render the full path (which is recursive)14:36
cedkalbertca: examples?14:37
albertcacedk: webdav collection14:37
cedkalbertca: link?14:37
albertcacedk: rec_name will try to print the full path "A / B / C"14:37
cedkalbertca: I know rec_name14:39
albertcacedk: as with consraint changes raises the exception with record's rec_name it may crash with some models14:41
cedkalbertca: I still don't understand14:42
cedkalbertca: what is the link of the faulty code?14:42
albertcacedk: the model storage one14:42
albertcacedk: model storage will raise an exception using record's rec_name if the record is recursive14:43
albertcacedk: but trying to print the rec_name will crash because it browses information to parent, grandparent, etc14:43
cedkalbertca: the code?14:44
albertcacedk: those two links are enough :)14:44
albertcacedk: the issue can be solved if in the check_recursion function we print the record's id14:45
albertcacedk: but that is not relevant for the user14:45
cedkalbertca: they point to nothing!14:45
cedkalbertca: I can not speak if I don't see the *code*14:45
cedkalbertca: I will not read all modelstorage.py14:46
albertcacedk: of course, but I don't know if I can point to a given line in a codereview14:47
albertcacedk: it is in line 85114:47
cedkalbertca: so use the id14:51
albertcacedk: that's what I did :)14:51
albertcacedk: do you think we could let the programmer use any field of the model?14:52
albertcacedk: Such as "%(name)s" and we replace with the field name of the model14:52
cedkalbertca: or improve rec_name to detect loop but it is a little bit against the constrint14:52
albertcacedk: yes, I didn't want to do loop detection in there14:52
albertcacedk: another idea would be to catch the exception and raise it correctly from the inheriting model14:53
cedkalbertca: why not having a rec_name params14:54
cedkalbertca: also I think it is good to show the name of the parent field14:54
albertcacedk: what do you mean with rec_name params?14:54
cedkalbertca: …, parent='parent', rec_name='rec_name')14:55
albertcacedk: ok, sounds good to me14:56
albertcacedk: will change that and update the patch today14:57
cedkalbertca: ok, should no more wait to be included in next release14:59
albertcacedk: yes, I know15:11

Generated by 2.11.0 by Marius Gedminas - find it at!