You can check here:

http://feeds.archive.org/validator/check?url=http%3A%2F%2Fsocietecivile....

or here:

http://feeds.archive.org/validator/check?url=http%3A%2F%2Finso.ath.cx%2F...

I submitted a patch few times ago on the list, but it was not received... You should only remove call to htmlentities in

format_rss_channel and format_rss_item (common.inc)

I can resubmit the patch if you want

Comments

dries’s picture

dries commented
Priority: Normal » Critical

But special HTML characters will not be encoded then?

dries’s picture

dries commented

I think we are best of using UTF-8 (Unicode) encoding instead of ISO-8859-1. It would eliminate all problems with HTML entities. Furthermore, the RSS specification(s) assume(s) that the document is encoded as UTF-8 unless otherwise specified (thus UTF-8 being the default encoding).

See PHP's utf8_encode() function and note that you can specify the charset htmlentities() should use.

al’s picture

al commented

I think utf-8 has issues with pound signs. As in, GBP. # is a hash, £ is a pound. Damned Americans... Only local images are allowed.

moshe weitzman’s picture

moshe weitzman commented
Priority: Critical » Normal
dries’s picture

dries commented
Component: Base system » base system
Assigned: Unassigned » dries

I have tested this extensively using French, German, Japanese and Korean characters and I am confident that this bug has been fixed. Marking it as such.

Anonymous’s picture

(not verified) commented