- From: Alan Ruttenberg <alanruttenberg@gmail.com>
- Date: Thu, 31 Jul 2008 16:16:18 -0400
- To: "Booth, David (HP Software - Boston)" <dbooth@hp.com>
- Cc: Richard Cyganiak <richard@cyganiak.de>, T.V Raman <raman@google.com>, "seb@serialseb.com" <seb@serialseb.com>, "www-tag@w3.org" <www-tag@w3.org>, "kidehen@openlinksw.com" <kidehen@openlinksw.com>, "tthibodeau@openlinksw.com" <tthibodeau@openlinksw.com>
On Jul 31, 2008, at 1:23 PM, Booth, David (HP Software - Boston) wrote: >> > I think serving the JSON is the best option. Serving HTML from / > resource.json would defeat the purpose of having a JSON-specific > URI. It is quite likely that the user pasted the JSON URI into a > browser to test it, and *wants* to see the JSON that is returned. > Everyone knows how to paste a URI into a browser; few know how to > configure their browsers to specify their desired MIME types. I don't see how the best option is to ignore the accept header. If the accept header says to accept only html then you shouldn't respond with a different mime type as if that was an appropriate response. The 406 or 30x responses make more sense. It's like saying, in a negotiation, that it's a fine thing to ignore other negotiator and do what you want. Its not much of a negotiation in that case. -Alan
Received on Thursday, 31 July 2008 20:16:59 UTC