1

According to AJAX redirect dilemma, how to get redirect URL OR how to set properties for redirect request it doesn't seem to be possible to find out when an HTTP redirect occurs nor the target URL the browser gets redirected to.

As such, how can one compare two URLs in a reliable manner? For example, my code compares two URLs:

http://example.com/users/joe@home.com/ and http://example.com/users/53/

The two actually refer to the same user (the first request redirects to the second using HTTP 303) but there doesn't seem to be a way for the browser to discover this so it mistakenly believes the two users differ.

Technically speaking, the resource representation could contain its canonical URI, but I'd like to avoid such workarounds/hacks if possible. Any ideas?

Community
  • 1
  • 1
Gili
  • 86,244
  • 97
  • 390
  • 689
  • To do what eventually? Of course, if the path of the URI differs, they're two different resources. So what are you trying to achieve anyway? – DanMan Dec 27 '12 at 13:46
  • @DanMan, you're wrong. The URIs might be different but they refer to the same resource. That's the whole point of URI canonicalization. URI identity is vital for caching purposes as well as some business logic. For example, the client would like to invoke an operation on all users aside from itself. If it can't compare users, it cannot exclude itself from processing. – Gili Dec 27 '12 at 16:02
  • If everything had it's own, unique address like it should, you wouldn't have to think about this. If the 2nd URI is the preferred one, you should redirect with status 301. – DanMan Dec 27 '12 at 16:10
  • @DanMan, there are good reasons for URI canonicalization and the use of status 303. I can't go into the details here, but if you read [RESTful Web Services](http://shop.oreilly.com/product/9780596529260.do) they are explained at length. – Gili Dec 28 '12 at 03:19

1 Answers1

1

I ended up returning the canonical URI as part of the response body. It's an ugly hack, but no one seems to have any other idea.

Contributions welcome! Feel free to post new answers and I'll consider them.

UPDATE: W3C is discussing fixing this issue at https://www.w3.org/Bugs/Public/show_bug.cgi?id=15417

Gili
  • 86,244
  • 97
  • 390
  • 689