libcurl to build without Digest support. (I figure it should also explicitly
disable Negotiate and NTLM.)

file that was already completely downloaded caused an error, while it
doesn't if you don't use --fail! I added test case 194 to verify the fix.
Grrr. CURLOPT_FAILONERROR is now added to the list stuff to remove in
libcurl v8 due to all the kludges needed to support it.

  replacement, curl only replaced the Host: header on the initial request
  and didn't replace it on the following ones. This resulted in requests with
  two Host: headers.

  Now, curl checks if the location is on the same host as the initial request
  and then continues to replace the Host: header. And when it moves to another
  host, it doesn't replace the Host: header but it also doesn't make the
  second Host: header get used in the request.

  This change is verified by the two new test cases 184 and 185.

PUTs even when told to use HTTP 1.0, which is not correct.

that gets set with CURLOPT_ENCODING

connections.

<300 to be good.

POSTing over HTTPS:// with NTLM works fine now. There was a general problem
with multi-pass authentication with non-GET operations with CONNECT.

server doesn't require any auth at all and then we just continue nicely. We
now have an extra bit in the connection struct named 'authprobe' that is TRUE
when doing pure "HTTP authentication probing".

might be used uninitialized

if a connection is tunneled through a proxy. A tunnel is done with CONNECT,
either when using HTTPS or FTPS, or if explicitly enabled by the app.

not be used both in one request. Fixed it and added test case 172 to verify.

take a CURLcode as a second argument, that is non-zero when Curl_done()
is called after an error was returned from Curl_do() (or similar).

memory.h is included everywhere for this.

that list in case something goes wrong in the function and we must bail out.
Courtesy of the torture testing.

happens

work when using proxy by converting the IDN-name to the ACE-encoded version
before the request-URL is passed to the proxy.

all things up to work with encoded host names internally, as well as keeping
'display names' to show in debug messages. IDN resolves work for me now using
ipv6, ipv4 and ares resolving. Even cookies on IDN sites seem to do right.

for it. It saves one extra copy of the header.

I also added comments for several functions in formdata.c

  stuff added a few weeks ago.  Turns out that if you specify --proxy-ntlm and
  communicate with a proxy that requires basic authentication, the proxy
  properly returns a 407, but the failure detection code doesn't realize it
  should give up, so curl returns with exit code 0. Test case 162 verifies
  this.

length limit of the hostname part of the URL.

instead to make it easier to find/read