Maybe improve or fork commons-app/apps-android-commons instead?

[matkoniecz]

See commons-app/apps-android-commons#4371 and commons-app/apps-android-commons#883

Maybe improving existing app would be better idea? More people who can help in maintaining and creation, Commons part is already coded, existing userbase.

Even if they would reject contribution - maybe forking it would be better than work from scratch.

[Discostu36]

I get the idea, but I fear that I have too many ideas (or too strict an idea of how I want it to work) to put all of that into the Wikimedia Commons app. When forking, I would throw away 90% of the code. But what I will certainly do is to look at the app code when it comes to communication with the Commons API.

[matkoniecz]

I would at least consider it - it would greatly increase chance that code in 5 or 10 or 15 years is still being used.

But obviously, it is your project (and when I was in similar situation I was much luckier with StreetComplete matching exactly what I wanted to achieve)

[mnalis]

In my experience, apps-android-commons is quite buggy (across several versions, it still randomly loses pictures, fails to upload, retry upload not working, such critical bugs still open after several years without being any closer to solution etc.)

It is what it is, and I can see devs are trying their best to cope and hopefully those will be fixed eventually, but still I wouldn't really recommend using that codebase for new projects - writing new code from scratch has its own benefits too (much easier writing and maintenance than trying to learn the functioning of existing code - especially one that is buggy in subtle and hard to reproduce / debug ways).

There seems to be https://github.com/CommonsLab/CommonsAPI if one wants to avoid writing that part of the code from scratch.