First alpha release about plone.restapi

[tisto]

@buchi @lukasgraf 1.0a1 released. Sorry it took so long...

[tisto]

@lukasgraf I think you listed everything. We can work on the actions right after the first alpha release...

[buchi]

@tisto @lukasgraf I would suggest to make a 1.0a1 release now.
The Framing and File/Image download changes are compatibility breaking and I would like to have 'em in a changelog. Thus I propose to do this in a later release.

[lukasgraf]

@buchi Framing has already been deferred.

Regarding breaking changes, we've already had plenty of those (members -> items, items_count -> items_total (Search), the introduction of batching, search -> @search, just to name a few).

But I'm fine with deferring File/Image download as well until after the release.

[tisto]

@buchi @lukasgraf I would actually like to wait for the file/image changes. I don't think implementing @lukasgraf's proposals is that much work.

[buchi]

@tisto @lukasgraf I'm fine with waiting for the file/image changes, but then I really would like to have a changelog to document changes.

[lukasgraf]

@buchi file/image changes are ready for a review, just didn't want to have it merged if I you felt strongly against having those changes in the first release already. Will open a PR right now: #115

[tisto]

All open issues for a first alpha release have been closed / merged. Expect a first release soon!

[lukasgraf]

@tisto excellent, thank you! 🎉

I noticed that the RTD docs build is currently failing again - docs for some features that already have been merged into master (like JWT auth and batching) haven't been published yet.

I briefly tried triggering a build manually that failed because RTD was missing sphinxcontrib.httpdomain. We removed that dependency during the sprint, because it wasn't / isn't being used, but it seems RTD tries to reuse build environments to some extent. Maybe we should just re-add it?

The other failure (TypeError: isinstance() arg 2 must be a class, type, or tuple of classes and types) is, IIRC, a version conflict that's flaky / not exactly deterministic - during the sprint I was able to work around that one by just manually triggering the build again (didn't have the nerve to look into it any deeper).

[tisto]

Ok. Will look into that right away...