Polyglots: Dual stats for creators and GTE/PTE

[JavierCasares]

When a user uses translate.wordpress.org to translate a plugin (or theme) and does so by translating a string from scratch (because it is new), when it reaches review by a GTE/PTE, several things can happen: that it is rejected, that it is approved, that it is sent for review or that the GTE/PTE has just reviewed some simple detail and is approved.

In the latter case, the feeling for a translator is that their string (and their work) has not served anything because that string is attributed to the GTE/PTE, which can generate some kind of frustration to the translator. As a note, for the record, I love the work that the GTE/PTE does, which is a lot.

There are probably many solutions, such as that the GTE/PTE use the "Changes Requested" function and that some type of notification or suggestion can be included that compares the previous string with the proposed one, so that the translator theirself makes the suggestion and the GTE/PTE simply approves it.

Alternatively, the GTE/PTE, when accepting the translation, indicates whether it is a major or minor change, and for a translation to have two users: the "major" creator or translator, and the GTE/PTE that makes the last "minor" translation. In this way, two statistics would be created and the main string would continue to be attributed to the one who has put more initial effort into that text.

Of course this needs a lot of discussion and a lot of opinion on the part of the GTE/PTE, but it would help translators to feel more involved with translations, since sometimes the feeling is "I have translated this, but I do not appear anywhere" or "someone has stolen my translation", in short, frustration.

This, in addition, could help a lot especially for new translators, who are the most likely to receive a correction from a GTE/PTE, and who would not feel the frustration of having made a production and not receiving the attribution.

Some ​discussion started in Slack.

[evarlese]

I agree that this can have a big impact, especially on new translators who want to be able to see their input on a project.

Are there any reasons (technical or otherwise) to not give credit to both the translator and the editor correcting the string? I also think using the Changes Requested status - to allow the translator to make the changes - will help in most circumstances.

[LidiaLAB]

This is a topic that we as GTE/PTE have been struggling from a while.
From the Editors' point of view, we also want the translations to be assigned to the person who had translated them.

Reaching out users has not always been so easy, and sometimes we need to accelerate the fix like near a new release of WP for the core projects.

Other times pinging a person for a "." seems too much to me.

I agree that now with the new Feedback tool I can change my workflow with little effort to encourage costant participation from new contributors in a new way and to get them recognized for all the efforts they do. And it's also a way to receive feedback from other editors on my own mistakes!

But I agree with who points out -in the Slack chat- the fact that "the original translator could dislike the new translation", so the change cannot be assigned without the original translator knows about it.

I disagree that "small" corrections are not relevant, because they do not conform to the guidelines and translators need to be aware of and also they can remember better if they do the correction themself.

I agree with who said -in the Slack chat- how is his team's workflow: "give feedback and then translate it yourself if no answer".

It could be after 1 or 2 weeks for standard project and based on urgency for the WP core and meta projects. We can use the Discussion dashboard to achieve this (there is a "Submitted on" column).

So the relevant point for me is how to notify the translators in a efficient way for them and for me as an editor.

Perhaps we can add more highlighted texts in the translate platform to inform users of this opportunity.

Like in the meta section of each string: "Do you want to be notified if translation changes are needed to be approved? Read this and opt-in for notifications." Maybe a shorter text...

It could also be an opportunity to clarify that translations have to be approved.

I agree with what was being contemplated -in the Slack chat-: we can reconsider the opt-in/out method. I don't think it is invasive for a person, who wants to see their translations approved, to receive initial notifications. Then in that emails we can make it clear how to opt-out of these notifications and the consequences of this (no more direct feedback for the translator).

[fierevere]

Sometimes corrections can be very small,
i.e.

• (missing) double space or (missing) space at beginning or end
• dot or comma
• correcting placehholder

I guess a checkbox "This is a minor change" (Just like in Wiki) will do, keeping string attribution to someone else than editing person

[Marc4]

revision contributors

[Marc4]

A couple of ideas:

1 - Hide the GTE/PTE from the translator list. So there will be no numbers competition.

2 - There could be one list on top of the other. In two different blocks. At the top, translators with their translated string numbers. Below that list a list of GTE/PTE, with their reviewed strings.

When a GTE/PTE edits a translator's translation and is approved in his name (GTE/PTE user), it is counted as a review, not as a translation. If the GTE translates a string not yet translated by anyone, it is attributed as a translation to the GTE/PTE, by placing it in the top list of translators.

2.1 - The GTE/PTE always remain in the lower block and have two values next to their username, the revised strings and the translated strings. This way they will not even affect the list of translators in the upper block.

Attached an example image representing idea 2. (The texts in the image are incorrect. They should be something like Reviewer or Editor Contributors in reference to the GTE/PTE block and list).

[Marc4]

editor contributors

[xavivars]

I really like Marc's proposal, and it would definitely solve some of the issues we've had in the past, by clearly differentiating translations from revisions.

It would still be up to the PTE/GTE to have something flagged as a translation or as a revision (with a checkbox), but having the revision flag checked by default every time a PTE/GTE reviews a string would probably be ideal.