- From: Julian Reschke <julian.reschke@gmx.de>
- Date: Wed, 08 Apr 2009 14:13:06 +0200
- To: Anne van Kesteren <annevk@opera.com>
- CC: Mark Nottingham <mnot@mnot.net>, HTTP Working Group <ietf-http-wg@w3.org>
Anne van Kesteren wrote: > On Wed, 08 Apr 2009 13:58:58 +0200, Julian Reschke > <julian.reschke@gmx.de> wrote: >> No, implementations do not have to, according to the spec. > > The spec does not matter for this. If the spec doesn't matter, why are you even interested in changing it? >> I believe you that browsers currently do, but I'm pretty certain that >> there's lots of code out there that doesn't accept relative URIs (or >> malformed URIs), which would become non-compliant by this change. > > Pointers? None right now. >> So, again, this would be an incompatible change. > > To make HTTP compatible with deployed Web content and actually a useful > spec to implement against. Again, we can't step out of our charter. What's needed is *proof* that most implementations (not only browsers!) already do this today because they were forced by broken content to support it. BR, Julian
Received on Wednesday, 8 April 2009 12:13:55 UTC