All, In an effort to reduce the amount of Wiki spam on the RESTwiki, it now requires a login in order to edit. Eugene tells me this will eradicate the bot...
I'm wondering if it would be okay to use a extension header to indicate acknowledgement of receipt of a resource in order to trigger new state. Take some queue...
This seems to be more about getting data to the server reliably. I was actually thinking more about how to get something like ActiveMQ's HTTP JMS interface to...
I am new to these concepts, so links to old discussions are welcome; I have done some research but it might be better to just ask a possibly naive question. ...
It violates the safety of GET, since the intent of the GET message is to update state on the server. It's idempotent, but I don't know why that's important. ...
... I think the core of this confusion is: what is the resource? in an e-commerce app it seems like the resource is being decorated for each user which is a...
... Of course. I have the bad habit of conflating the two concepts. ... I want to make sure that the client can re-establish context at anytime but GETting...
... I seems to me that the server needs to receive some acknowledgement that the client has successfully processed the message in order to clear the head of ...
Okay, here's another crack at this using GET and DELETE. 1. Start with a queue at /queue. A GET at that queue hangs until a message occurs. GET /queue The...
... Maybe its possible to do something analogous to TCP acknowledgment with HTTP headers, but I would have the server signal that its looking for the...
Nice article Paul. For many producers and one consumer it suggests a protocol more like this: Producer ======== request POST /queue/next response HTTP 200 OK ...
Sorry, if I had read the article properly I might have copied the first part of the protocol correctly. Producer ======== request POST /queue response HTTP 201...
... IME, you want to avoid that if you can, but if you can't then you're likely best off making the ack explicit by using a POST request from the client to the...
... I [obviously] misunderstood your original intent: you're trying to do the actual `pop` operation on a queue data-type ... I'd do so in a hybrid of what...
... Josh, that's very nice indeed. It is a perfect example of how to put 'API' semantics into the message instead of implicitly extending the semantics of the...
Hi again, My original posting stated that I assumed many producers and one consumer. I thought that was the usual setup for queues, ie. to regulate input into...
... [Various race conditions described here. Other exist.] ... The problem is the GET followed by the DELETE are intended to be atomic. A side effect is needed...
... be atomic. [...] Indeed, I'll note that there is nothing in the HTTP spec which prohibits returning a body in the response to a DELETE request. So, ...
... Since GET and DELETE are both idempotent, there is not need to come up with any other mechanism for atomicity. GET and DELETE can be called repeatedly in...
Man, this has gone on for a while. I don't remember what the actual application was -- or did anyone bother to ask or mention it? How does this queue thing fit...
I've yet to see a workable and glitch-free description of a classical "message queue" using HTTP. Even a variant that uses DAV would be useful to have...
... One of us must be confused - the issue of whether the methods are idempotent or atomic are somewhat orthogonal. The point that Patrick was trying to make,...
Hi Group, I'm very new to REST but am interested in moving my current development effort into the model. That said, I must confess I need some clarification...
... POST is more general than create, and in fact even PUT can create a resource. "create", AFAICT, emerged from the desire for some to fit the new "uniform...
... It's hard to say generally, but I think you're seeing two different things: * an inherent similarity in the two methods, including a bit of a semantic...
... I haven't found a reason to use PUT rather than POST for creating a new resource. I'm still waiting for the killer example where it makes sense. Using PUT...
... Josh, while that may be your theory (and you are welcome to it), the notion that there are two levels of REST is completely false. There are different...
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.
