Note, this doesn't depend on fa related bugs (but the more general collation bugs)

Unfortunately ckb is not on the list of supported collations by the icu library (which is the program we're using to sort things). So basically this bug is in the same boat as bug 46235, that is the collation needs to be first accepted to the cldr project: http://cldr.unicode.org/index/cldr-spec/collation-guidelines

Basically, if the language is not listed at http://www.unicode.org/repos/cldr/trunk/common/collation/ then its unlikely that we support the proper sorting for it.

OK. Now what are we suppose to do?

Can you set wgCategoryCollation to 'uca-fa'? Beacause Persian collation is very similar to our language.

(In reply to comment #3)

Can you set wgCategoryCollation to 'uca-fa'? Beacause Persian collation is
very
similar to our language.

This bit is easy enough.

Change 76242 had a related patch set uploaded by TTO:
(bug 52015) set uca-fa category collation for ckbwiki

https://gerrit.wikimedia.org/r/76242

(In reply to comment #5)

Change 76242 had a related patch set uploaded by TTO:
(bug 52015) set uca-fa category collation for ckbwiki

https://gerrit.wikimedia.org/r/76242

Given that this is for a different language, we should be explicit what the difference is, and make sure there is concensus

I tested Persian language for our language, it is answerable. Community is aware about this change.

Change 76242 merged by jenkins-bot:
(bug 52015) set uca-fa category collation for ckbwiki

https://gerrit.wikimedia.org/r/76242

reedy@tin:/a/common$ mwscript maintenance/updateCollation.php --wiki=ckbwiki --previous-collation=uppercase
Fixing collation for 36522 rows.
Selecting next 10000 rows... processing...10000 done.
Selecting next 10000 rows... processing...20000 done.
Selecting next 10000 rows... processing...30000 done.
Selecting next 10000 rows... processing...36522 done.
36522 rows processed

Thanks, now all in order!