The resource to my mind is "http://localhost/mysubresource". In other
words most of these problems go away if you avoid using parameters to
identify things. They're good for querying (of course), good for
services, and good for managing coordinate spaces. They're weak sauce
for identification. You can guess what I think of most web frameworks as
a result :)
I am currently developing a RESTful Web Service and will be using query strings extensively. There seems to be no guide line regarding how to form the query ...
This is strictly implementation level detail. Unless you are expecting clients to manufacture URIs by using HTML forms or URI templates or some other...
So I could be wrong here, but I thought what prashanthshenoyu describes is infact standard... in that.. http://localhost//MyResources/{Id}/State would be...
Kevin, ... I don't quite follow your point and I'm not sure which part you mean is standard, but there's certainly nothing about classes or instances that is...
This is just a convention adopted by an implementation. RoR has a similar convention, but there is nothing standard about it. It is all about convenience to...
... Agreed that it's just a convention, and that neither is "standard" in any sense. But there is (at least) one important consideration that might affect...
At Wed, 09 Jul 2008 15:32:17 -0700, ... Do you have an example of a cache which exhibits this behavior? The subtleties of web caching have often eluded me, but...
... Interesting ... I first noticed this behavior with Squid[1], when I was experimenting with trying to cache the results of Google-like search requests that ...
... Yep, it would. Treating them as non-cacheable is lesser evil than that :-). ... It denies on '?' characters as well, by default. ... Did a bit more...
... That's good point. They do not represent the same resource. If the sever thinks that the resource is indeed the same, it could very well use the...
... A cache is not allowed to do that: "A cache cannot assume that an entity with a Content-Location different from the URI used to retrieve it can be used to...
... The resource to my mind is "http://localhost/mysubresource". In other words most of these problems go away if you avoid using parameters to identify...
At Wed, 9 Jul 2008 20:28:13 -0700, ... In any case I think that there is some confusion in this thread, between a) treating the URIs: http://example.org/index ...
... Today, this directive from 2616 is mostly meaningless since rewrite rules can be used to turn query params into path segments and vice versa, and so...
... If that were true the cache in question would be broken (as others have pointed out). What you are probably thinking of is that many caches are configured...
Kevin, this is exactly what prompted me to ask the question in the first place. Â Semantically the two formats of the query string seem to be different to me ...
Hi Prashanth, ... actually, the URI carries no semantics at all, as far as clients are concerned. The server may parse the URI, but to clients it is nothing...
Ah, my bad then. I thought there was a general direction the REST community at large was taking with regards to the URI and how it breaks down, hence my use of...
... No. There are at most vague guidelines. I use this criterion: do I want to parametrise a computed resource, or is the thing I want an application level...