I'm working on a REST interface for a system that doles out unique identifiers. For each request, the caller gives the system a namespace tag, and the system...
Applying the razor to this question leads me to ask: why can't you just use POST? For example: POST /numbers/FOO/next => 1 POST /numbers/FOO/next => 2 POST...
... Hash: SHA1 ... Eric> I'm working on a REST interface for a system that doles out Eric> unique identifiers. For each request, the caller gives the Eric>...
... That's not really a fetch then, is it? K. -- Blacknight Internet Solutions Ltd. <http://blacknight.ie/> Unit 12A Barrowside Business Park, Sleaty Road,...
... Nothing implicit about it. That's an update alright. -- Blacknight Internet Solutions Ltd. <http://blacknight.ie/> Unit 12A Barrowside Business Park,...
... Your problem is that what you are doing is not really fetch. Problem solved :) That said, if there was no issue with id's getting lost then I see no ...
... An example of a case where GET is arguably not appropriate is "page hit counters". Unfortunately page hit counters can only really be done by GET (unless ...
... It's only inappropriate if you assume that these things are accurate. After all, idempotency doesn't rule out side-effects completely, it just means that...
... The difference being that a GET can't cause those resources to change in any significant way, but that's not to say that they can't change due to some...
... Neither GET nor POST imply any "unique id" or "sequence number" internals. I'd believe something in a media type would inform the client that the server...
... Part of my concern about the POST-centric approach is that it is implicit in all the proposals in favour that the POST request has no body (please correct...
... You are correct, of course, but that wasn't really what I was driving at. An (explicit) empty POST can be regarded as simply shorthand for a non-empty...
My 2 cents, The resource you are GETting at /numbers/FOO/next can be described as "the next number in the sequence FOO". Thus the GET is idempotent as it ...
... This is similar to design issues around editing and (especially) mapping mom queues onto HTTP. The problem with GET here is multiple clients, and caches....
... But the notion of "next" implies state. I'm not sure whether you're proposing this state exists on the client or the server. If the latter, though, I think...
Elliotte Harold
elharo@...
Apr 8, 2007 7:02 pm
8205
... Here is the interface I finally settled on: POST /numbers/FOO/nextInc returns the next number in the sequence and increments it. GET /numbers/FOO/next...
... I can see your point; it would seem to indicate that what Eric really wants is a non-idempotent verb of his own devising, maybe `ADVANCE`. Pragmatically,...
... I'd briefly thought of that, but would much prefer not to create new verbs, primarily because the REST model involves a limited set of verbs. I'd think...
... I think it involves a uniform set of verbs; having a limited set is a means, not an end. ... Is verbing nouns as you did better than making new verbs? In ...
Over in Apache Jakarta land, it looks like they are wrapping up the commons projects on the grounds that most of the code was never maintained; its just...
Thanks for the heads up! I have a strong interest in both client and server components, but haven't followed at all the project organization. I do use the...
... I think so. ... heh, take a look at how much Java5 and 6 have adapted to support cookies and proxies out the box -and note that the automatic proxy stuff...
Some resource algebra... Suppose I have two resources X and Y where X is some arbitrary resource and Y is a resource that is a count of the number of GET ...
... As long as you don't depend on Y returning any particular value, sure it's RESTful, just like when you were using two separate resources above. The value...
Having problems with message search?
Fill out this form to ensure your group is one of the first to be migrated to the new message search system.
