1 files changed, 74 insertions, 0 deletions
diff --git a/lib/README.httpauth b/lib/README.httpauth
new file mode 100644
index 000000000..960504510
--- /dev/null
+++ b/lib/README.httpauth
@@ -0,0 +1,74 @@
+
+1. PUT/POST without a known auth to use (possibly no auth required):
+
+   (When explicitly set to use a multi-pass auth when doing a POST/PUT,
+   libcurl should immediately go the Content-Length: 0 bytes route to avoid
+   the first send all data phase, step 2. If told to use a single-pass auth,
+   goto step 3.)
+
+   Issue the proper PUT/POST request immediately, with the correct
+   Content-Length and Expect: headers.
+
+   If a 100 response is received or the wait for one times out, start sending
+   the request-body.
+
+   If a 401 (or 407 when talking through a proxy) is received, then:
+
+   If we have "more than just a little" data left to send, close the
+   connection. Exactly what "more than just a little" means will have to be
+   determined. Possibly the current transfer speed should be taken into
+   account as well.
+
+   NOTE: if the size of the POST data is less than MAX_INITIAL_POST_SIZE (when
+   CURLOPT_POSTFIELDS is used), libcurl will send everything in one single
+   write() (all request-headers and request-body) and thus it will
+   unconditionally send the full post data here.
+
+2. PUT/POST with multi-pass auth but not yet completely negotiated:
+
+   Send a PUT/POST request, we know that it will be rejected and thus we claim
+   Content-Length zero to avoid having to send the request-body. (This seems
+   to be what IE does.)
+
+3. PUT/POST as the last step in the auth negotiation, that is when we have
+   what we believe is a completed negotiation:
+
+   Send a full and proper PUT/POST request (again) with the proper
+   Content-Length and a following request-body.
+
+   NOTE: this may very well be the second (or even third) time the whole or at
+   least parts of the request body is sent to the server. Since the data may
+   be provided to libcurl with a callback, we need a way to tell the app that
+   the upload is to be restarted so that the callback will provide data from
+   the start again.  This requires an API method/mechanism that libcurl
+   doesn't have today. See below.
+
+Data Rewind
+
+   It will be troublesome for some apps to deal with a rewind like this in all
+   circumstances. I'm thinking for example when using 'curl' to upload data
+   from stdin. If libcurl ends up having to rewind the reading for a request
+   to succeed, of course a lack of this callback or if it returns failure, will
+   cause the request to fail completely.
+
+   The new callback is set with CURLOPT_IOCTLFUNCTION (in an attempt to add a
+   more generic function that might be used for other IO-related controls in
+   the future):
+
+     curlioerr curl_ioctl(CURL *handle, curliocmd cmd, void *clientp);
+
+   And in the case where the read is to be rewinded, it would be called with a
+   cmd named CURLIOCMD_RESTARTREAD. The callback would then return CURLIOE_OK,
+   if things are fine, or CURLIOE_FAILRESTART if not.
+
+Backwards Compatibility
+
+   The approach used until now, that issues a HEAD on the given URL to trigger
+   the auth negotiation could still be supported and encouraged, but it would
+   be up to the app to first fetch a URL with GET/HEAD to negotiate on, since
+   then a following PUT/POST wouldn't need to negotiate authentication and
+   thus avoid double-sending data.
+
+   Optionally, we keep the current approach if some option is set
+   (CURLOPT_HEADBEFOREAUTH or similar), since it seems to work fairly well for
+   POST on most servers.