- From: Julian Reschke <julian.reschke@gmx.de>
- Date: Wed, 5 Feb 2025 10:18:42 +0100
- To: ietf-http-wg@w3.org
On 05.02.2025 09:30, Kevin Marks wrote: > ... > What I'm getting at is that, for a http GET, the Postelian answer is to > fall back to 200 and send the resource, rather than a draconian reading of > the spec. > ... So assuming you GOT (:-) a resource with a Last-Modified date of "x". In the meantime, the resource has changed and now has a Last-Modified date of "y". If the client now submits a PUT request to that resource with "If-Unmodified-Since: x" with the intent not to overwrite changes, and the server chooses to ignore that request header field, the changes will (or could be) lost. So: ignoring a conditional request header field can have bad consequences. And yes: - Last-Modified timestamps only have 1 sec resolution, - a server that supports PUT really should understand that conditional header field The point being: it's not always safe to ignore conditional request fields. How does WebDAV -- indeed defining a new conditional field -- handle this for "If" (<https://greenbytes.de/tech/webdav/rfc4918.html#HEADER_If>)? It supports negotiation of features (see <https://greenbytes.de/tech/webdav/rfc4918.html#HEADER_DAV>). That may not be the best way to do this, but note that this was conceived in RFC 2518, as of Feb 1999, predating RFC 2616. Best regards, Julian
Received on Wednesday, 5 February 2025 09:18:48 UTC