On Thu, 7 Jun 2001, Pierre Z. wrote:

> > Well, this is next to impossible to understand if you don't show more
>
> somehow I expected that too, and in fact I needed to understand from
> you exactly what information you'd need to better understand where the
> problem lies - I guess I got that now: everything. What I'm unsure of is
> how exactly I can gather all that information.

Well, most often it is enough that you show us curl's command lines and the
headers the remote server sent back. Possibly also how curl's (HTTP) requests
looked like.

In almost every case like this (and I've worked a few during the last couple
of years), it is a matter of understanding curl, the protocol and how the
remote servers/scripts generally behave. It is also about guessing, as we
don't know for sure how the remote program works or what features in curl's
request that it checks/rejects...

Did you try to fake a "known browser" user-agent string?

> I guess I can somehow save a log of all headers going back and forth to a
> file and mail it all to you.

It would of course help to understand.

> > Did curl follow the Location: headers correct?
> how and where can I see this?

When you use -i you get to see the headers together with the file's contents.
The headers include a "Location: [URL]" that tells curl where to get the next
page.

-- 
     Daniel Stenberg -- curl dude -- http://curl.haxx.se/