[This is in response to "(COUCHDB-269) _list doesn't allow POST requests", but I'm posting to the list as the issue seems to be broader than just the _list URI pattern.]

Using POST to fetch documents seems a bit weird — by REST principles that should be a GET, shouldn't it? Using a POST obscures the fact that the request is read-only and potentially cacheable. Any reason this couldn't be handled by a GET, passing in the document IDs? Is it the potential size of the resulting URI? (Is it kosher to pass a request body in a GET?)

—Jens

Attachment: smime.p7s
Description: S/MIME cryptographic signature

Reply via email to