On 21/02/2014 8:59 p.m., Marko Cupać wrote: > Hi, > > in order to reduce cache.log spamming with info about non-existing > error dirs for variants of Serbian, I created some additional symlinks: > > sr-cyrl-rs -> sr-cyrl > sr-latn-rs -> sr-latn > sr-rs -> sr-latn > > Perhaps it can be done in official distribution. We provide a script to generate the symlinks as part of the langpack. If you download the latest Squid langpack you will find them there. > > BTW who creates these standards? Why are there 8 or more variants for a > language spoken by less than 10 million people? Moreover these variants > are all linked to 2 that should be enough: sr-latn and sr-cyrl. > You would think so. However these are human written languages, and like timezones politics plays a part. The ISO standards define a set of language codes, alphabet codes and country codes with a way to combine the three. The multiple variants in comes with being spoken as official language in a multiple countries or as in Serbians case having several alphabets. (Serbian with 3 only variants for each alphabet is one of the smaller languages. Spanish by comparision is the largest with over 20 official country variants.) Which ones are available for users to select is up to the UA (browser) authors. The ones available in Squid langpack are the ones we have been told by admin like yourself that users are actively requesting as their primary display language (that "spamming" in your log). Amos
