Tryton - Issues

 

Issue3672

Title Generalize error message
Priority feature Status in-progress
Superseder Nosy List ced, pokoli, resteve, reviewbot, xcodinas
Type behavior Components
Assigned To ced Keywords review
Reviews 35291002,64481002
View: 35291002, 64481002

Created on 2014-02-12.13:01:10 by ced, last changed by reviewbot.

Messages
review64481002 updated at https://codereview.tryton.org/64481002/#ps120001
review64481002 updated at https://codereview.tryton.org/64481002/#ps100001
review35291002 updated at https://codereview.tryton.org/35291002/#ps380001
review64481002 updated at https://codereview.tryton.org/64481002/#ps80001
review64481002 updated at https://codereview.tryton.org/64481002/#ps60001
review35291002 updated at https://codereview.tryton.org/35291002/#ps360001
review35291002 updated at https://codereview.tryton.org/35291002/#ps340071
review64481002 updated at https://codereview.tryton.org/64481002/#ps40001
review64481002 updated at https://codereview.tryton.org/64481002/#ps20001
review64481002 updated at https://codereview.tryton.org/64481002/#ps1
msg44984 (view) Author: [hidden] (ced) (Tryton committer) (Tryton translator) Date: 2018-11-23.09:08:07
Here is WIP review64481002 for modules
review35291002 updated at https://codereview.tryton.org/35291002/#ps310001
review35291002 updated at https://codereview.tryton.org/35291002/#ps290001
review35291002 updated at https://codereview.tryton.org/35291002/#ps270001
msg44952 (view) Author: [hidden] (ced) (Tryton committer) (Tryton translator) Date: 2018-11-21.01:05:31
I uploaded a WIP, it is already functional as the trytond tests pass. I still have to improve some exception testing in tests. But it can already be discussed about the derived exceptions I created etc.
Next step will be to make reviews for each needed module.
msg44928 (view) Author: [hidden] (ced) (Tryton committer) (Tryton translator) Date: 2018-11-19.18:36:58
I will have to remove the _error_messages because _sql_constraint will become ambiguous if we keep it.
msg44924 (view) Author: [hidden] (pokoli) (Tryton committer) (Tryton translator) Date: 2018-11-19.09:59:37
You've been added as collaborator on the review, feel free to pick the issue.
msg44920 (view) Author: [hidden] (ced) (Tryton committer) (Tryton translator) Date: 2018-11-18.18:39:31
I would like to resume this review and finish it for inclusion.
review35291002 updated at https://codereview.tryton.org/35291002/#ps260001
review35291002 updated at https://codereview.tryton.org/35291002/#ps220001
review35291002 updated at https://codereview.tryton.org/35291002/#ps210001
review35291002 updated at https://codereview.tryton.org/35291002/#ps170001
review35291002 updated at https://codereview.tryton.org/35291002/#ps150001
review35291002 updated at https://codereview.tryton.org/35291002/#ps130001
review35291002 updated at https://codereview.tryton.org/35291002/#ps100001
review35291002 updated at https://codereview.tryton.org/35291002/#ps60001
msg35198 (view) Author: [hidden] (xcodinas) (Tryton committer) Date: 2017-08-17.11:43:30
>Yes the goal is to remove _error_messages but we could do it in two steps, 
>first support ir.message and after progressively replace _error_messages.
>ir.message should use the translation mechanism.

I've updated the review and I've added the design to follow to replace error 
messages, the idea is to directly remove _error_messages.
msg35012 (view) Author: [hidden] (ced) (Tryton committer) (Tryton translator) Date: 2017-08-03.22:14:08
Yes the goal is to remove _error_messages but we could do it in two steps, first support ir.message and after progressively replace _error_messages.
ir.message should use the translation mechanism.
review35291002 updated at https://codereview.tryton.org/35291002/#ps40001
review35291002 updated at https://codereview.tryton.org/35291002/#ps20002
msg32360 (view) Author: [hidden] (pokoli) (Tryton committer) (Tryton translator) Date: 2017-03-08.14:18:29
I have a few concerns about the migration:

- Should we remove cls._error_messages? Or deprecate them and register the  messages to the new ir.translation table. 
- Should we keep the current translations values? Or we should introduce the ir.message records as new strings to translate? The latter have the benefit that they can be reviewed but I'm wondering if this will be too much work.
New review35291002 at https://codereview.tryton.org/35291002/#ps1
msg32358 (view) Author: [hidden] (xcodinas) (Tryton committer) Date: 2017-03-08.13:46:03
I'm going to make a script that will migrate the actual error data to this design.
msg32357 (view) Author: [hidden] (xcodinas) (Tryton committer) Date: 2017-03-08.13:41:37
I add review35291002 that adds ir.message and uses it for the error messages.
msg32036 (view) Author: [hidden] (pokoli) (Tryton committer) (Tryton translator) Date: 2017-02-17.14:37:25
Just droping a note to take in account issue6277 when implementing this.
msg15589 (view) Author: [hidden] (ced) (Tryton committer) (Tryton translator) Date: 2014-02-12.13:01:08
As the wizard always skip those error translation, if there are changes on it,
it keep for the language the old one and the new.
It is not a big issue because the get_source will search for matching origin.
This could be fixed by changing the way error messages are defined.
We should create an Model for such message (ir.message) where the «name» will be
translatable and replace all the error message defined in the models by record
in this table. This way, we will have unambigus links: «xml id» - «db id» -
«model» «res_id».
History
Date User Action Args
2018-12-05 23:37:14reviewbotsetmessages: + msg45175
2018-12-02 22:07:26reviewbotsetmessages: + msg45100
2018-11-30 11:42:25reviewbotsetmessages: + msg45094
2018-11-30 01:16:44reviewbotsetmessages: + msg45092
2018-11-30 00:44:14reviewbotsetmessages: + msg45091
2018-11-29 12:16:47reviewbotsetmessages: + msg45074
2018-11-29 11:35:51reviewbotsetmessages: + msg45070
2018-11-29 00:36:11reviewbotsetmessages: + msg45069
2018-11-27 18:45:39reviewbotsetmessages: + msg45046
2018-11-23 18:41:57restevesetnosy: + resteve

Showing 10 items. Show all history (warning: this could be VERY long)