Idea: vector mark-up over screenshots

[Wohlstand]

Describe the problem

It's often to explain or point certain thing for the phrase it's needed to upload many variants of the same screenshot (just draw arrows, write down hint labels, etc.). It an inconvenience on the fact it's need to draw multiple variants of screenshot, and at the same time this leads a bloating of server storage usage by a lot of duplicated (with minor differences) screenshots.

Describe the solution you would like

I got an idea to allow making various vector-based hints (arrows, frames, text labels of different colours) that can be placed using WebLate itself and attached to every individual translation string while re-using the same screenshot for all of them. Additionally, when updating the same screenshont in a condition it has the same size, it's possible to just keep all old mark-ups (or clean-up all of them if really need to draw new one). Every of such hints might use very few memory, just few integers and short strings. And that data can be easily displayed and scaled over the screenshot itself.

Describe alternatives you have considered

No response

Screenshots

Some examples of such hints mark-up (currently they are just drawn on images themselves):

Additional context

No response

[github-actions]

This issue has been added to the backlog. It is not scheduled on the Weblate roadmap, but it eventually might be implemented.

In case you need this feature soon, please consider helping or push it by funding the development.