- From: Bryce Nesbitt <bnesbitt@bepress.com>
- Date: Tue, 23 Dec 2008 08:30:35 -0800
- To: ietf-http-wg@w3.org
- Cc: "Adrien de Croy" <adrien@qbik.com>
Received on Tuesday, 23 December 2008 16:31:10 UTC
But, browser support is not pertenant to this particular need. This is a need 100% in the world of robots. And in this case I'm dealing with robots that are already required to properly handle the 503 response. For an example, see http://www.openarchives.org/OAI/openarchivesprotocol.html So what's the procedure for proposing a spec change like this, aimed at robot efficiency, and understanding that browsers don't care now and won't care in the future? Keep in mind that if a 20X Retry-After header is ignored, the robot will simply get a 503 later. So the compatible fallback is perfect. And if browser authors really wanted to get into the game, they could play also, as long as they fix their 503 implementation first. On Tue, Dec 23, 2008 at 3:52 AM, Adrien de Croy <adrien@qbik.com> wrote: > > > Hi Bryce > > I couldn't find a browser that honoured Retry-After even on a 3xx response. > So I wouldn't hold out much hope about widespread support for honouring it > on a 2xx. It looks like it hasn't really been picked up by the browser > vendors as far as I can tell. > >
Received on Tuesday, 23 December 2008 16:31:10 UTC