Opened 3 years ago
Closed 3 years ago
#5844 closed enhancement (wontfix)
Cleaning up translations which lack an author.
Reported by: | NekoJonez | Owned by: | |
---|---|---|---|
Milestone: | Priority: | normal | |
Component: | Translate Site & Plugins | Keywords: | |
Cc: |
Description
To follow up from the meeting we had today in the #polygots channel, I think it would be a great idea if somehow we can easier clean up the old translations that lack an author.
The discussion started from here: https://wordpress.slack.com/archives/C02RP50LK/p1627417555092500
We could do this in various ways, which each have their upsides and downsides.
1) Attach them to the project author (plugin/theme) -> Then they have an author and it's of the developer BUT it's still going to be a witch hunt to find and edit them.
2) Put them all back into waiting, which is going to show them all for GTE's & PTE's but it's might ruin a lot of already done core translations.
3) Attach them all to a generic account? This well make less of a witch hunt but the question is then... how far do we go with this? Since, in which projects are these strings? One per type of project, so one for Core/Meta, one for theme's and one for plugins? And do we do it per locale...? Or per "main locale" ? (EG: One for nl_NL, nl_BE and nl_NL-formal instead of 3)
As @dd32 already pointed out:
dd32 10:19 AM
Right, it’s not, I could probably find a way.. but it’s not as simple.. given there are <counting… counting…> nearly 1 million projects+locales matches with unknown authorship strings.
Currently, a changeset have been made already to help somewhat in this:
https://meta.trac.wordpress.org/changeset/11154
Which is a great help already. But, it's only project based for now.
Personally, I'm a fan of the 3rd option. But, I think that we need to work this out and discuss this further and look with the developers to see what's possible since if we are talking about over 1 million strings... If we do this wrong, this might get extremely messy.
So, I created this ticket just to keep track of this idea/thing and post possible idea's and thoughts. :)
IMHO, these are correctly not authored by any specific user. I don't think we should assign them to any user.
As of [11154] you can find these translations in a specific project, but, there isn't a listing of the other 850k projects/locales that have imported strings in this stage.
Unfortunately we cannot really filter to -just- translations that were auto-imported from a plugin, although we can mostly assume it. Future strings will go into waiting per [11155]. Previous strings are reviewed by the plugin author, as it's assumed that they knew what they were doing, similarly to strings that plugin authors do not mark as translateable.
They should only be in plugins.
IMHO; This is a waste of time, The strings can be corrected when/if someone finds one that is out of place, or that doesn't meet the locales current glossary. With the sheer number of strings that are in this state (12.5 million strings over 850k translation sets (project/locale)) it's not likely that there's going to be anything significant to find.