Re: Problem with url-retrieve when HTTP status is 304

[Stefan Monnier]

>>> With CVS Emacs, url-retrieve triggers an error when it gets an HTTP status
>>> of 304 *and* the requested page is not in the cache.  To reproduce, try
>>> this:
>> 
>>> (setq url-request-extra-headers
>>> '(("If-Modified-Since" . "Sun, 18 Sep 2005 17:34:28 GMT")
>>> ("If-None-Match" . "\"91ad777b5809a56e2b2b9d7c4833fceb\"")))
>>> (url-retrieve-synchronously "http://tumbleblindly.org/feed/";)
>> 
>> Just so I understand the problem better: where did you get info such as
>> "Sun, 18 Sep 2005 17:34:28 GMT" ?

> From the HTTP headers of the URL in question.  I fetched the headers
> outside of the URL library.

The problem here is that it's completely separate from the cache itself.
Who knows, maybe the cache was filled 10 years ago, so if the server replies
with "unchanged since Sun, 18 Sep 2005 17:34:28 GMT" we still shouldn't use
the file that's in the cache.


        Stefan