- From: Sergiu Dumitriu <sergiu.dumitriu@gmail.com>
- Date: Sat, 9 Sep 2006 03:23:09 +0300
- To: public-webapi@w3.org, "Tim Berners-Lee" <timbl@w3.org>, tabulator@csail.mit.edu
Received on Saturday, 9 September 2006 17:51:13 UTC
In section 1.7 ("Not in this Specification") of the current draft [1], one of the items listed is: * "Property to disable following redirects" >From the original URI of the request to the actual data location there can be almost any number of redirects. In some applications maybe it is of no interest where exactly is the data, as long as it finally gets to the user, so stopping for each 3xx would be a waste. On the other hand, there could be applications where the data is of no interest unless it comes from the specified location. In my opinion, making a global decision whether or not to transparently follow redirects is not the best solution. getResponseHeader returns only the last response header [2], so in the case of transparent redirection the client code cannot access the intermediate 3xx responses. If the XMLHttpRequest object is to transparently follow redirects, then there should at least be a way to log and access all these intermediate steps. [1] http://www.w3.org/TR/XMLHttpRequest/#notcovered [2] http://www.w3.org/TR/XMLHttpRequest/#dfn-getresponseheader
Received on Saturday, 9 September 2006 17:51:13 UTC