On #805: Turn bigint IDs back into normal integers

[lentschi]

see https://stackoverflow.com/questions/44390225/how-to-turn-off-bigint-primary-keys-in-rails-5-1
and https://edgeguides.rubyonrails.org/5_1_release_notes.html#active-record-notable-changes

This fixes an issue introduced by me in 6b98fb2. I think this wouldn't be likely to cause any observable errors in the current master, but it breaks the PR #762 (should it ever be merged.)

[paroga]

should it ever be merged.

I still want to merge it, since it's very valuable, but I really want to test it with a bunch of real world databases first, which I did not find the time yet. 😞

[lentschi]

I still want to merge it, since it's very valuable, but I really want to test it with a bunch of real world databases first, which I did not find the time yet. disappointed

👍 Sure - better not hurry something like that! That's why I still left the PR in draft status - I myself am unsure if this might cause havoc.
(Also the tests in this PR currently fail. The reason for this is, that the shared_article does not work with the new constraint on article_category_id. To fix this I must first find a way to test article sharing locally.)