In #59 (resp. #97) several new properties of categories have been added to the database: multi-complete (and the dual), multi-terminal object (and the dual), locally copresentable, accessible (and the dual), ℵ₁-accessible (NOT the dual), finitely accessible (NOT the dual), generalized variety, multi-algebraic, locally multi-presentable, locally finitely multi-presentable, locally poly-presentable. (Also other properties which however are OK.)
The PR did not decide these properties for several categories. But this should be done, in particular for those categories where - before that PR - all properties had been decided. (Let us ignore difficult categories like the category of schemes for now.)
Namely, for these categories we should decide the mentioned properties. A first PR can handle only the easy cases. A later PR can handle the difficult cases. Or we can have one PR per property. In either case: PRs should be small, something like #59 (resp. #97) should not happen again.
Ignore the properties that are unrelated to the PR (for example if Met is regular).
In #59 (resp. #97) several new properties of categories have been added to the database: multi-complete (and the dual), multi-terminal object (and the dual), locally copresentable, accessible (and the dual), ℵ₁-accessible (NOT the dual), finitely accessible (NOT the dual), generalized variety, multi-algebraic, locally multi-presentable, locally finitely multi-presentable, locally poly-presentable. (Also other properties which however are OK.)
The PR did not decide these properties for several categories. But this should be done, in particular for those categories where - before that PR - all properties had been decided. (Let us ignore difficult categories like the category of schemes for now.)
Namely, for these categories we should decide the mentioned properties. A first PR can handle only the easy cases. A later PR can handle the difficult cases. Or we can have one PR per property. In either case: PRs should be small, something like #59 (resp. #97) should not happen again.
Ignore the properties that are unrelated to the PR (for example if Met is regular).