On Mon, 26 May 2008 13:15:49 +0200, Julian Reschke <julian.reschke@gmx.de> wrote: > Anne van Kesteren wrote: >> On Mon, 26 May 2008 12:18:34 +0200, Ian Hickson <ian@hixie.ch> wrote: >>> I do think that setting the header to the empty string (as if "" had >>> been >>> passed instead) is better behaviour, from an author's point of view. If >>> that would resolve this issue then that seems like a good choice. >> As far as I can tell three out of four browsers stringify to "null" / >> "undefined" for arbitrary headers and since authors should pass a >> string anyway I really don't think it's worth changing this. > > That behavior simply is not useful, and likely will hide bugs. Since it is the behavior of many JavaScript APIs I don't think authors will be surprised. > FF3 (removing the header) demonstrates that this case (null value) > either should be left unspecified, or specified with a more useful > behavior. Firefox demonstrates that it is not interoperable with the three other browsers. It does not demonstrate anything about the need to do this differently than currently specified. -- Anne van Kesteren <http://annevankesteren.nl/> <http://www.opera.com/>Received on Tuesday, 27 May 2008 13:07:00 GMT
This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 27 May 2008 13:07:01 GMT