- From: Fred Covely <fcovely@bcftech.com>
- Date: Tue, 20 Aug 2002 13:43:29 -0400 (EDT)
- To: <www-lib@w3.org>
Libbwww'ers (maybe it should be Libbwwwites?): Anyway I've posted before on a variety of issues (answers and questions), and recently ran into a wierd one. I'm open for suggestions/opionions and I'll run them down. I am trying to robotically drive a web site that appears to be: Server: Microsoft-IIS/5.0 My software is routinely driving all kinds of other sites that use IIS 5.0. This particular site is secure (HTTPS). Every page on the site does a POST, followed by a redirect to the same location, except that the redirect sends the POST data back over in a GET. Still no great shakes as many other sites do something like that. On this particular site though I get intermittent: HTTP/1.1 403 Access Forbidden errors on the POST (not the subsequent GET mind you). I have a good tracing utility and it appears that the HTTP traffic sent by my robot and a running version of IE are exactly the same. Two other interesting pieces of data. 1) My robot has a setting wherein I can drive POST redirects with POST's rather than GETS. If I turn that switch on so that all POST's redirect with POST rather than emulating IE (which redirects with GETS), then everything works great, eg. no intermittent 403's. 2) If I set the LIBWWW short/long writepostdelay to large values (10sec/90sec), at certain critical pages, I can make it work with the POST/GET redirect scheme. Any ideas? thx fhc Fred Covely BCF Technology fcovely@bcftech.com
Received on Wednesday, 21 August 2002 03:21:04 UTC