On Thu, 28 Jun 2007, Georg Hubik wrote:

>> >> User-Agent: curl/7.16.2 (i686-pc-linux-gnu) libcurl/7.16.2
>> OpenSSL/0.9.7a
>> > zlib/1.2.1.1
>> >> Host: 10.1.0.106
>> >> Accept: */*
>> >>
>> > <http>
>>
>> Here you should've shown us the server response headers as they would be
>> interesting for this case.
>
> There are no more headers. The listing is complete (the http content is
> stripped)

Really? All server respones are sent with a few headers first and then the
response body! If this response actually didn't respond with any headers, then
things are really weird...

> I figure you can use other clients fine with Digest on this URL?
>
> Exactly. I've posted in the first mail the headers which I get when using
> browser (I've tried Mozilla and IE - both works just fine). There is one
> more strange issue. When I login to the page via browser it somehow gets
> 'unlocked' and I can access it using curl for certain time, but then it gets
> 'locked' again. I've also tried to get all objects from the loggin page
> hoping that the authentication is dependent also on this, but no success.
> (there are no cookies)
>
> Is it possible to enable somehow more detailed logging? Any suggestions what
> to try?

--trace or --trace-ascii log *everything* that goes in and out.

But I'll have to admit I really don't know what's wrong here. AFAIK and as far
as the test cases go, libcurl's digest support works. You seem to have hit a
case where it doesn't, and it will be tricky for us to debug this...

-- 
  Commercial curl and libcurl Technical Support: http://haxx.se/curl.html