Updating language for POSTing to an LDPCv #239
Conversation
|
I don't understand the motivation for this paragraph:
it seems to say make a zero size memento which seems not very useful. But anyway, I think this is just a special case of I think the logic is really:
|
|
I agree with @zimeon that POSTing with a |
index.html
Outdated
| If an <a>LDPCv</a> does accept <code>POST</code> with a request body, it SHOULD respect a | ||
| <code>Memento-Datetime</code> request header for the created <a>LDPRm</a>. Absent this header, it MUST use | ||
| the current time. | ||
| If an <a>LDPCv</a> supports <code>POST</code>, a <code>POST</code> with both a request body and a |
There was a problem hiding this comment.
I think we should remove "both a request body and".
index.html
Outdated
| <p> | ||
| If an implementation does not support <code>POST</code> with a request body, the <code>Accept-Post</code> | ||
| header of any response from the <a>LDPCv</a> SHOULD indicate that no request body is accepted via the form | ||
| <code>Accept-Post: */*; p=0.0</code>, and that implementation MUST respond to any body-containing |
index.html
Outdated
| <code>Memento-Datetime</code> header SHOULD be understood to create a new <a>LDPRm</a> contained by the | ||
| <a>LDPCv</a>, reflecting the state of the <a>LDPRv</a> at the time of the <code>POST</code>. | ||
| Implementations MUST fail such requests that also contain a non-empty body by responding | ||
| with a 400 Bad Request. |
There was a problem hiding this comment.
I'm not sure I agree with the non-empty body. If it's a dumb implementation, that may be how the versions are created. If the impl doesn't support request entities on version creation via POST, that should be indicated somehow (as in the next 2 paragraphs).
There was a problem hiding this comment.
That is to say that a combination of Accept: from the server and a 400 response should tell the client that the operation was illegal - I don't think we should specify anything else here.
|
@zimeon @escowles @barmintor I've tried my best to incorporate your feedback. Seeing where we've landed now, it almost feels we could just as easily say that, if absent, the |
* Updating language for POSTing to an LDPCv Resolves: fcrepo#215
Resolves #215
I kept the SHOULDs to maintain the assertiveness of the original language, but could be compelled to upgrade to MUST, though I admit I'm not sure of the depth of the ramifications that would incur.