freetags 4.1: Base related entries on amount of shared tags

[onli]

Instead of taking all entries with common tags and to order them by date, now it orders all those entries first by the amount of common tags. That way entries that are topically related will feature more prominently in the list of related entries, if tagged accordingly

[mattsches]

I like the idea 👍

However, I'm not sure if this works with all supported DBMS, some tend to be restrictive concerning GROUP BY clauses (there is the ONLY_FULL_GROUP_BY sql_mode in MySQL, and PostgreSQL might complain about missing columns in the GROUP BY clause, etc…).

[onli]

Good point, I tested only on sqlite so far. I'll give this a testrun on my dev server with mysql and postres as configured by uberspace.

[onli]

But of course, if you can test as well or have a solution for one of these DBs, that would be great :)

[onli]

@mattsches, okay, got it I think! You were completely right, postgres did not like the missing columns. I added them, expecting the count to fail then, but no, everything seems to work fine. Tests still welcome ;) But less necessary now.

Thanks for catching this!