Last modified: 2010-04-02 15:47:18 UTC
Created attachment 6267 [details] Use $wgLang for date formatting, not $wgContLang The „revision as of“ message shown above the diff in feeds (page history feeds, RC feed) displays the date in the content-language format, but the text of the message (and many other items in the feed) are in the user’s preferred language. See the URL for an example. Note that this might have been an attempt to make the feed language-independent, but that fails in many ways. See also bug #19391.
Hmm, tough one. The above is not a fix I think. It's killing one symptom. First question is: should a feed always be in content language? If so, all text elemenents should be checked and rendered in the content language. Is anyone able to provide clarity on the expected behaviour of the feed with regards to language?
I don’t think this is a tough bug... The tough thing is in bug #19391. This is just a trivial fix, so that the feed is at least _consistent_. Note that currently, _everything except the date_ is in the user’s language, not content language. (Heck, the date itself is just one of parameters to wfMsg, not wfMsgForContent!) The issue whether the feed should be changed to use the content language always (and how to achieve that nontrivial problem – think about the formatting of log entries in the feed!) would be better debated in bug #19391, IMHO. (I have also provided a link to a relevant bug #17868 there.)
(In reply to comment #2) > I don’t think this is a tough bug... The tough thing is in bug #19391. This > is just a trivial fix, so that the feed is at least _consistent_. Note that > currently, _everything except the date_ is in the user’s language, not > content language. (Heck, the date itself is just one of parameters to wfMsg, > not wfMsgForContent!) > > The issue whether the feed should be changed to use the content language always > (and how to achieve that nontrivial problem – think about the formatting of > log entries in the feed!) would be better debated in bug #19391, IMHO. (I have > also provided a link to a relevant bug #17868 there.) > Looks like bug #19391, bug #17868 have been resolved. Does the issue reported in this bug still occur?
(In reply to comment #3) > Looks like bug #19391, bug #17868 have been resolved. > Does the issue reported in this bug still occur? Sure, it does. Check the linked URL, it contains items like “مراجعة أقدممراجعة 14:10, 19 August 2009” – interspersing English date with Arabic text. Those two bugs are not duplicates of this, or anything like that.
Done in r64521.