IRC logs of #tryton for Tuesday, 2010-11-30 #tryton log beginning Tue Nov 30 00:00:02 CET 2010
-!- GasbaKid(~GasbaKid@ has joined #tryton00:13
-!- johbo( has joined #tryton00:36
-!- plantian( has left #tryton00:38
-!- zodman(~andres-va@ has joined #tryton01:00
-!- jcm( has joined #tryton01:10
-!- ready( has joined #tryton01:28
-!- ikks(~ikks@ has joined #tryton02:42
-!- jcm( has joined #tryton03:10
-!- yangoon( has joined #tryton05:18
-!- johbo_( has joined #tryton06:20
-!- Timitos(~kp@ has joined #tryton07:22
-!- vladimir_(~vladimir@ has joined #tryton07:54
-!- paepke( has joined #tryton08:04
-!- cedk(~ced@gentoo/developer/cedk) has joined #tryton08:29
-!- GasbaKid(~GasbaKid@ has joined #tryton08:39
-!- pjstevns( has joined #tryton09:14
-!- jcm( has joined #tryton09:24
-!- cedk(~ced@gentoo/developer/cedk) has joined #tryton09:27
-!- JoePass(~gasbakid@ has joined #tryton09:30
-!- bechamel( has joined #tryton09:32
-!- paepke( has joined #tryton09:38
-!- vincentvdl(~vincent@ has joined #tryton10:03
cedkvincentvdl: did you check with this
cedkvincentvdl: for the d&d10:17
vincentvdlcedk: Hi! I am loading it now... will let you know in a minute if it fixed. Thanks!10:19
-!- paepke( has left #tryton10:36
cedkyangoon: could you give an example for issue1775?10:37
yangooncedk: one in in the issue, but I can you give several examples10:38
cedkyangoon: but I don't see where is the issue10:39
cedkyangoon: we already talked about that and it was ok10:39
yangooncedk: the sum calculated by addition of rounded values does not match the sum calculated directly with sum / rate10:40
yangoonit is not an issue of rounding method, rather an issue of accounting reality10:40
yangooneither we do rounding at the end or we have to provide some correction to match correct sums10:41
yangooncaused by addition of rounding errors10:41
cedkyangoon: there will never be a solution10:42
yangooncedk: we use already a solution to be correct on the side of accounting10:43
yangoonwhich for us is more important than to be correct mathematically on single values10:44
-!- paepke( has joined #tryton10:45
cedkyangoon: the current behavior is correct10:45
cedkyangoon: did you re-read this
yangooncedk: no, a total of 63.07 divided by a rate of 1.8 results to 35.04 rounded, not 35.03 as in tryton10:47
cedkyangoon: but it is impossible to do in a different way10:47
yangooncedk I think there are different methods, will to have dig in before doing the blue print (in case you are interested)10:49
cedkyangoon: you can not have one calculus based on one currency and having the same if you convert each part to an other currency10:49
cedkyangoon: I don't see what you want to do10:49
cedkyangoon: and which result you want to have10:50
cedkyangoon: i would like to have the account moves you wanted10:50
yangooncedk: one is to follow accumulated rounding error and if it exceeds one rounding step -> compensate10:52
cedkyangoon: an example please10:53
yangooncedk: sorry, just now a little bit in a hurry, have to talk to you later, or better write down the issue and proposal10:55
cedkTimitos, udono: ping10:57
cedkTimitos, udono: you seem to have the same issue?10:57
-!- udono( has joined #tryton11:16
cedkpjstevns: will you submit patches for your issues?11:24
pjstevnscedk: they're on rietveld. Do you want them attached to roundup?11:24
cedkpjstevns: it is as you want but if you want to be the author of the changeset it is simplier once the patch is reviewed to put it on roundup like described in
pjstevnscedk: roger that, will do.11:27
cedkpjstevns: and when you are sure that the patch is clear (because simple) you can directly put the patch on roundup11:30
pjstevnscedk: copy that11:30
cedkpjstevns: thx11:35
cedkudono: did you see the talk with yangoon?11:36
udonocedk: yes11:36
cedkudono: it seems you have the same issue?11:37
udonocedk: the issue described by yangoon seems a valid issue for the discussion in
udonocedk: but I need to thnk about.11:39
udonocedk: It is a pity, that all discussions/solutions I found are written in german language. There is only one academical paper, which is written in english:
cedkpjstevns: your patch has an unknown parent?11:42
pjstevnscrap. Both patches for 1772 and 1773 are stacked on top of each other. Doesn't hg apply them as-is??11:44
cedkpjstevns: I use --extact option11:44
cedkpjstevns: otherwise you will have to merge your local repo11:44
cedkpjstevns: I will apply the other one first11:45
pjstevnscedk: --extract option on what command??11:45
cedkpjstevns: on import11:45
cedkpjstevns: also the patch of issue1772 refere to a parent not in the tree11:46
cedkpjstevns: so it is as you want, I can apply it without exact or let you fix it11:47
pjstevnscedk: patches were done on the 1.8 tree :-(11:48
pjstevnscedk: I'll update the patches myself for the trunk and try to make sure they're clean11:48
cedkpjstevns: ok11:50
cedkpjstevns: patch must always be done on the upper version affected11:51
cedkpjstevns: we take care to transplant it to each series affected11:51
pjstevnscedk: understand, makes sense.11:51
-!- ikks(~ikks@ has joined #tryton11:55
vincentvdlcedk: About patch 3347041: It did not fix our sequence issue with drag&drop, we now get a different key error (and yes we doublechecked sequence is added to tree in XML)11:59
vincentvdlcedk: Traceback (most recent call last):12:00
vincentvdlFile "/tryton/gui/window/view_form/view/", line 371, in drag_data_received12:00
vincentvdlfield =[treeview.sequence]12:00
vincentvdlKeyError: 'sequence'12:00
vincentvdlcedk: We are still working on 1.8 by the way12:03
pjstevnscedk: patches are uptodate now12:04
cedkudono: do you have simple description of the algo?12:12
cedkvincentvdl: this is realy strange12:14
cedkvincentvdl: the integer field is named 'sequence'?12:14
cedkpjstevns: applied thx12:14
-!- pjstevns( has left #tryton12:14
udonocedk: I requested the Author of the paper, Jiri Sima, but he told me that he doesn't know about a real world implementation in any programming language.12:15
vincentvdlcedk: Yes, it's called sequence12:16
cedkvincentvdl: the KeyError seems to indicate that it is not in the view12:17
cedkvincentvdl: do you see it in the view?12:17
cedkudono: no need of a real implementation, but even in pseudocode12:18
udonocedk: the author is a mathematics, not a programmer...12:19
udonocedk: another way mentioned in the german wp mentioned is to use linear approximation.12:20
vincentvdlcedk: No, it's not visible on the screen... but the client does not report a missing field in the console either12:21
cedkvincentvdl: I think you don't have the right view12:23
udonocedk: sorry not linear approximation, least square the german wp mentioned:
vincentvdlcedk: we added a dummy field (next to sequence) to the tree and that field is visible on the screen...12:26
vincentvdlcedk: we will create a small demo12:27
-!- JoePass(~gasbakid@ has joined #tryton12:41
-!- GasbaKid(~GasbaKid@ has joined #tryton12:41
cedkudono: this one
cedkudono: I already almost implemented for a customer module to load sales (tax incl.) from cash register14:05
cedkbut if we implement thing like that, the amount of invoice line could change depending of the other lines of the same invoice14:07
udonocedk: AFAIK it is not allowed to change the net amounts, because they are usually used for official tax calculations. But I see no problem to tweak the brut amounts including the taxes.14:21
udonocedk: and it shout be possible to tweak the tax amounts, too.14:22
cedkudono: I mean when editing14:23
cedkudono: but never after openning the invoice14:24
udonocedk: yes, only on editing.14:24
cedkudono: it will be great to have scenario for what we want to have14:28
udonocedk: yes14:30
udonoyangoon: May you provide a prose text about the steps needed to show the rounding problem? If you like I can provide the proteus/doctest part for the test.14:35
-!- gavinf( has joined #tryton14:39
-!- gremly(~gremly@ has joined #tryton15:08
Timitoscedk: i know the problem but i need to dive into it much more as it is a complex problem. i think it is the right way to start a documentation about it and once the documentation is clear we should implement it.15:08
-!- zodman(~andres-va@foresight/developer/zodman) has joined #tryton15:59
-!- pepeu(~manuel@ has joined #tryton16:51
-!- trifon( has joined #tryton16:55
-!- bechamel( has joined #tryton17:24
-!- jbunting( has joined #tryton17:51
-!- JoePass(~gasbakid@ has joined #tryton17:53
-!- GasbaKid(~GasbaKid@ has joined #tryton17:54
-!- chrue( has joined #tryton18:21
-!- cedk(~ced@gentoo/developer/cedk) has joined #tryton19:12
-!- paepke( has joined #tryton19:17
-!- plantian( has joined #tryton19:17
-!- vincentvdl(~vincent@ has left #tryton19:26
-!- jcm( has joined #tryton19:32
-!- trifon( has joined #tryton19:49
-!- GasbaKid(~GasbaKid@ has joined #tryton20:16
-!- hoRn( has joined #tryton20:32
hoRnlittle question: with a wizard I create an instance of a model with a workflow - this action is always rolled back.20:54
hoRnhow can i force the commit?20:55
hoRnwithout the workflow the creation of the model works fine20:56
hoRni'm going nuts21:20
-!- plantian( has joined #tryton22:06
plantianHi, I'm running trytond 1.4, is it best to update from 1.4 to 1.6 and then wait for 1.8 or 1.4 straight to 1.8 or what is the best strategy?  Thanks.22:07
cedkhoRn: could you be more precise?22:14
cedkplantian: it is like you want22:14
hoRncedk: I can't be more precise, because I don't know whats happening ;(22:16
cedkhoRn: how do you create the instance of the model?22:16
hoRncedk: now i have the same problem in copy in the treeview22:17
plantiancedk: To migrate I will have to go from 1.4 to 1.6 to get to 1.8 anyways right?  Do I need to upgrade the database manually or do most modules and trytond have ways of detected an upgrade?22:17
cedkplantian: you don't need to pass through every version (normally)22:18
cedkplantian: you just have to run --update on the database22:18
plantianI have some modules I wrote so I think I will need to manually update those though to use new trytond conventions.22:19
cedkhoRn: treeview == list view?22:19
hoRncedk: yes22:19
cedkplantian: yes of course22:19
cedkplantian: and before any update process22:19
plantianBest approach for that is combination of reading CHANGELOG in trytond and the mailing list.22:19
plantianI guess?22:19
cedkplantian: yes but normally your module will make trytond stop before anything22:20
plantianAlthough tryton project has many more interesting modules that might satisfy my original needs.22:20
cedkhoRn: for the same model?22:20
hoRncedk: the error occurs if i add a workflow - i think i have a error in the workflow of the model22:20
plantiancedk: Okay, I will experiment.  Thanks.22:20
cedkhoRn: could you share the code?22:20
cedkplantian: for sure, you should perhaps re-think the design22:21
cedkplantian: but we try to avoid new stuff to existing module22:21
hoRncedk: one moment22:21
hoRncedk: - that is the model22:23
plantiancedk: Yeah, I wrote a very cludgy point of sale interface, product listing module, price list module and some stock module extensions.22:23
cedkplantian: of course before any thing: make backup ;-)22:24
hoRncedk: -that is the workflow22:24
cedkhoRn: you must put ModelWorkflow as first parent22:24
hoRnoh - thats all - back in seconds ;)22:25
plantiancedk: hahaha yeah, I will22:26
cedkhoRn: I did not check the workflow definition but you can see it from client22:26
cedkhoRn: by the way, you don't need to put method of Function field above22:26
hoRncedk: Nooooooooooooooooooooo - that was the cause of the error - i was searching over 1 hour (cry)22:27
cedkhoRn: and no need to prepend _ to the name, they will not be callable from rpc22:27
cedkhoRn: I tried to implement exception in this case but it is not an easy task22:28
cedkbut I think it will be solved when we will merge ModelSQL and ModelStorage22:29
hoRncedk: ok, a big red <blink> in the documentation would be nice :)22:29
cedkhoRn: patch is welcome :-)22:30
hoRncedk: i have ported the quality modul form openErp to tryton - i will contact the author and you if it can be a module in tryton22:31
cedkhoRn: have you link?22:32
hoRncedk: seems good - i have simplyfied the stuff a bit - becaus i was in hurry22:34
cedkhoRn: I'm not sure to understand the usage22:34
-!- GasbaKid(~GasbaKid@ has joined #tryton22:35
cedkhoRn: do you have direct link to bzr repo22:36
hoRncedk: i will conect it to the module project - with this module you can define qualitychecks22:36
cedkhoRn: bazaar is a mess to find stuff22:36
hoRncedk: waht are you looking for?22:37
cedkhoRn: I find it was the code22:39
hoRncedk: i can send you my version - so you can test it in tryton ;)22:40
cedkhoRn: I suggest you to post on tryton-dev and upload on codereview22:40
cedkhoRn: this is the best way to be included in Tryton22:41
hoRncedk: should i contact the original author before?22:41
cedkhoRn: it is as you want, the code is GPL2+ so you can put yours under GPL3+22:42
cedkhoRn: as soon as you keep the copyright notice of NaN22:42
hoRncedk: OK22:42
cedkhoRn: but you can inform them to be nice22:42
hoRncedk: i will do so22:42
-!- gavinf( has joined #tryton22:46
-!- yangoon_rm( has joined #tryton23:06
plantiancedk: Did someone start the production module I saw a comment on MRP and its in the goals but I'm not sure the production module in the goals are the same as MRP?23:29
cedkplantian: it is the same23:41
-!- jbunting( has left #tryton23:49
-!- blast_hardcheese( has joined #tryton23:52

Generated by 2.11.0 by Marius Gedminas - find it at!