Petition Details and Chat

Petition ID 35: Character-set encodings bug on the Wiki
Submitted Thu, 20 Oct 11 18:30:57 +0000

Junebug writes: The Wiki of Lies has long been known for being unable to display '...' and various other character combinations. Here, finally, is someone who seems to understand what's going wrong.

Naesten writes: There's something going wrong with encodings and the Wiki of Lies. For example, see http://enquirer.improbableisland.com/dokuwiki/doku.php?id=reality>. Notice that weird junk at the end? It looks suspiciously like what happens when a UTF-8 encoded character is interpreted as ISO 8859-1 (or the related codepage 1252)... which, in fact, is exactly what's going on. Presumably, this is because of the line "Content-Type:text/html; charset=iso-8859-1" in the HTTP headers sent by the server. You should probably try and find out where that's coming from so you can either get rid of it, or correct it to refer to UTF-8. (DokuWiki now only supports output in UTF-8, so don't bother trying to get it to do something else: see . Presumably, this header is either generated by Apache (presumably because that's what you've configured it to do) or by your dokuwiki templates. (Perhaps your templates are starting to output HTML before DokuWiki gets a chance to set the header?) Anyway, this happens *whenever* someone types "..." (among other things; see for a few more details) on a wiki page.

Junebug again: I hope this, or discussions with Naesten, can lead to the removal of this longstanding irritant.

Back to the List of Petitions

You have to be logged in to chat.