| irc.libera.chat #tryton log beginning Fri Jan 12 12:10:01 AM CET 2024 | ||
| -!- cedk(~ced@gentoo/developer/cedk) has joined #tryton | 23:27 | |
| -!- ChanServ changed mode/#tryton -> +o cedk | 23:27 | |
| -!- Pegasus_RPG(~Thunderbi@wsip-184-184-229-74.lv.lv.cox.net) has joined #tryton | 00:27 | |
| -!- cedk(~ced@gentoo/developer/cedk) has joined #tryton | 00:32 | |
| -!- ChanServ changed mode/#tryton -> +o cedk | 00:32 | |
| -!- springwurm(~springwur@2a01:76c0:104:c700:54a4:1957:115d:73c9) has joined #tryton | 06:23 | |
| -!- LordVan(~LordVan@gentoo/developer/lordvan) has joined #tryton | 06:41 | |
| -!- mrichez(~Maxime@mail.saluc.com) has joined #tryton | 07:28 | |
| -!- cedk(~ced@gentoo/developer/cedk) has joined #tryton | 07:34 | |
| -!- ChanServ changed mode/#tryton -> +o cedk | 07:34 | |
| -!- tbruyere(~Thunderbi@mail.saluc.com) has joined #tryton | 07:37 | |
| -!- yangoon(~yangoon@pd9e4a932.dip0.t-ipconnect.de) has joined #tryton | 08:10 | |
| -!- LordVan(~LordVan@gentoo/developer/lordvan) has joined #tryton | 08:21 | |
| LordVan | hi | 08:21 |
|---|---|---|
| LordVan | cedk, quick question .. had a user send me a strange error after 6.8.8 client update (from 6.8.2 so i don'T know for sure which caused it but i looked at the history) | 08:21 |
| LordVan | could this https://foss.heptapod.net/tryton/tryton/-/commit/c320a786923f19e138459d9a128c7d8cb097f1b1#9b2c55c4ba56b60c54b6cf53defbf79fea4040d2 | 08:21 |
| LordVan | cause an error on the account_invoice page during selecting date ? | 08:22 |
| LordVan | unfortuantely i do not have a trace due to user fail and taking a screenshot only of part of it | 08:22 |
| LordVan | [will try to validate just wnat to know if i am on the right path] | 08:22 |
| pokoli | LordVan: could you share the screenshot? | 08:22 |
| LordVan | the exception was a type error about 'int' not being subscriptable | 08:23 |
| LordVan | pokoli, was jsut going to see where i could paste it | 08:23 |
| pokoli | unless you do not give us any clue about the error we can not perform any investigation | 08:23 |
| LordVan | i know iw as just wondering if i should type the stuff in dpaste or post the screenshot but unsure where to upload .. lemme stick it in my cloud | 08:23 |
| pokoli | LordVan: the commit about you linked is avoud domain inversion in xxx2many and multiselection value. I'm not sure if this is the case of account invoice | 08:24 |
| pokoli | probably its easier to ask the user to reproduce the problem, note the steps and post the full traceback | 08:24 |
| LordVan | yes i know but there is a reason. i will post it while trying to investigate | 08:24 |
| LordVan | i had to rollback first cuz it was urgent | 08:24 |
| LordVan | i do have a db backup to test later | 08:25 |
| LordVan | https://cloud.lordvan.com/index.php/s/XacBxNTK982YDsA | 08:25 |
| LordVan | but yeah due to time constraints i did rollback to 6.8.2 client and all worked fine | 08:31 |
| -!- acaubet(~Thunderbi@194.224.31.235) has joined #tryton | 08:31 | |
| pokoli | LordVan: yes, it seems your error is related to the linked commit | 08:35 |
| LordVan | but yeah i intend to test more when i get a chance and report it properly. just thought i see if i am on the right path first. thanks | 08:36 |
| LordVan | pokoli: https://dpaste.org/zAhfM | 09:00 |
| LordVan | that happens when i try to open the generated invoice | 09:00 |
| pokoli | LordVan: it seems you have a domain which have a 'in' operator with a int value | 09:01 |
| pokoli | LordVan: where it should be a list of integers | 09:02 |
| LordVan | pokoli, yes but the trce is not helping me find out what the issue is | 09:02 |
| LordVan | and any clue? | 09:02 |
| pokoli | LordVan: just run the client on linux and update the code to print the values and you will get more inforamtion | 09:03 |
| LordVan | pokoli, i will do that but cannot right now as i don'T have access from here to the linux box | 09:03 |
| LordVan | my code bte: https://github.com/LordVan/tryton-modules/tree/master/rm_extra_data | 09:03 |
| LordVan | pokoli, it might be useful to print that by default on the trace tbh | 09:04 |
| LordVan | https://dpaste.org/pzoEj same with different widget triggering it after that | 09:05 |
| LordVan | but all seems to go back to that code | 09:06 |
| pokoli | LordVan: validate validates all field, and there is a wrong domain on some field | 09:07 |
| pokoli | you just need to know which field and fix the domain | 09:07 |
| LordVan | pokoli, yes but i need more info for that so i need to experiment on my dev box later | 09:07 |
| LordVan | unfortuantely i am on windows @ work due to CAD software,.. | 09:07 |
| pokoli | LordVan: it seems related to some custom code because I canset invoice date without issues on latest 6.8 client using demo database | 09:08 |
| LordVan | pokoli, do you think there is a (generic) way to print out more useful information from that traceback though? -- i mean for the upstream client in general not just me debugging now | 09:08 |
| pokoli | I normally use the demo database to test if this is upstream related or related to custom code | 09:08 |
| LordVan | good point | 09:08 |
| LordVan | i could/should have done that too . will do it next time | 09:08 |
| pokoli | LordVan: just check your code and you will probably find it | 09:08 |
| pokoli | It is a custom domain on a field which returns the wrong type for the operator | 09:09 |
| LordVan | yes looking now .. thing is i barely did any changes on the invoice except add 2 char fields and some invoice grouping stuff (But that happens before this, and is currently disabled on production so ..) | 09:10 |
| LordVan | anyway I do not know too muich how the whole domain/validation in tryton works internally but i am checking my code to see if i notice anything. thansk | 09:10 |
| LordVan | so the best place wuld probabyl be putting the logging / .. in record.py validate right? | 09:14 |
| pokoli | LordVan: yes, knowing the field which is causing the error will help | 09:15 |
| LordVan | that's for sure | 09:15 |
| pokoli | LordVan: it seems the error depends on the modules installed. There is amodule which you use and it is not used on the demo databse which raises the error | 11:18 |
| pokoli | LordVan: I got the same error on another database | 11:18 |
| LordVan | pokoli, ah ok good cuz i did not find it yet - i will be able to run it on linux a bit later | 11:59 |
| LordVan | pokoli, do you want to have a list of modules, or have you found it already? | 12:32 |
| -!- cedk(~ced@gentoo/developer/cedk) has joined #tryton | 12:47 | |
| -!- ChanServ changed mode/#tryton -> +o cedk | 12:47 | |
| -!- springwurm(~springwur@2a01:76c0:104:c700:bcdd:71f0:7b50:72e3) has joined #tryton | 13:05 | |
| -!- LordVan(~lordvan@gentoo/developer/lordvan) has joined #tryton | 14:37 | |
| LordVan | pokoli, cedk https://dpaste.org/NP6Bs | 14:46 |
| LordVan | the client 6.8.8 with some trace | 14:46 |
| LordVan | it (for some reason) tries to access the first element of value .. but that is an integer.. | 14:47 |
| LordVan | in fact a lot of values at this point are integers so I think the single_value is set wrongly in this case | 14:48 |
| LordVan | also it seems to only happen when I access an invoice for the very first time after it was generated | 14:49 |
| LordVan | I will see if i can find more info and then report it as a bug (hopefully knowing more about it then) | 14:49 |
| LordVan | https://dpaste.org/FpmEx for some reason the single_value parameter is set to False for the sale line id | 14:56 |
| LordVan | I am unsure how to debug/find the cause of this right now as I am not that familiar with these parts of tryton, so if anyone got suggestions where to look/how to trace this back to where it came from ;) | 14:56 |
| cedk | LordVan: looks like https://bugs.tryton.org/issue12843 | 15:15 |
| LordVan | ah yes | 15:16 |
| LordVan | cedk, I had just now created the bug https://foss.heptapod.net/tryton/tryton/-/issues/12930 | 15:17 |
| LordVan | if you want to close it as duplicate feel free | 15:17 |
| LordVan | unless you see anything useful in it | 15:17 |
| LordVan | cedk, about that bug .. would it be an option to just add an extra check there if value has the correct type(is subscriptable) and if not just return value instead of value[0] | 15:29 |
| LordVan | or would that cause some issues otherwise? | 15:29 |
| cedk | LordVan: no it should not have non correct type | 15:35 |
| LordVan | indeed | 17:20 |
| LordVan | anything i can help with | 17:20 |
| LordVan | let me know .. gtg now though | 17:24 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!