Proposed (works for me) fix of versioning detection

[paulgear]

This is my proposed fix for the incompatibility between scripts/import_tasks_from_local.sh and the versioning decisions in versioning.py and backend_localfile.py.

Fixes #920 (hopefully).

[diegogangl]

Hey, thanks for sticking around and submitting a fix for this! I haven't tested yet, but the code looks good (hinted, documented and all too). There's only caveat: the local backend is going away once the gtk4/newcore port lands (this pr: #894).

We could put these changes in the versioning file, or wait until the branch lands and port them over in the new datastore (which fills the roles of datastore+requester+local backend).

[paulgear]

Hey, thanks for sticking around and submitting a fix for this! I haven't tested yet, but the code looks good (hinted, documented and all too). There's only caveat: the local backend is going away once the gtk4/newcore port lands (this pr: #894).

That PR says in the opening statements that it's on fire - is that still the case? I'd personally rather not track the moving target of a future version, because I want this to work for the currently-available stable version. But I have a working data file now, so my biggest barrier to getting my life back is solved. 😄

We could put these changes in the versioning file, or wait until the branch lands and port them over in the new datastore (which fills the roles of datastore+requester+local backend).

I really don't think versioning is either well-named, or an appropriate place for file operations - it's really about XML structure conversions.

I'll leave it up to you what to do with this. Once the new backend structure arrives I'm happy to help work out how to make this fit in with that.