cURL / Mailing Lists / curl-library / Single Mail


RE: [PATCH v2] Add connection delay to Happy Eyeballs.

From: Steve Holme <>
Date: Sat, 2 Nov 2013 22:05:00 +0000

On Thu, 30 Oct 2013, Björn Stenberg wrote:

> Thank you for testing. My refactoring of the synchronous error
> handling code was a smidge too aggressive. Here is an updated
> patch.

I briefly mentioned this in my base64 decode thread earlier today but I just
wanted to post the more finer detail here:

If I took my Proxy Server down whilst I was testing my base64 work I
received the following output from curl:

* Rebuilt URL to:
* About to connect() to proxy port 8080 (#0)
* Trying
* Adding handle: conn: 0x139acb0
* Adding handle: send: 0
* Adding handle: recv: 0
* Curl_addHandleToPipeline: length: 1
* - Conn 0 (0x139acb0) send_pipe: 1, recv_pipe: 0
* Whut?
* Whut?
* Whut?
* Whut?

This patch appears to address this and returns the standard failed connect
output - thank you.

However, I would like to ask (and I appreciate you work closely with Daniel
so may not be too relevant) if you wouldn't mind:

a) Splitting your patches up a little - this one appears to address 2 or 3
separate issues (the add of the delay, IP address ordering, failed connect
fix and tidy up) which would help with others reviewing them as well as
enabling the application of individual parts if necessary

b) Use our standard format for commit comments as taken from

---- start ----
      [area]: [short line describing the main effect]

      [separate the above single line from the rest with an empty line]

      [full description, no wider than 72 columns that describe as much as
      possible as to why this change is made, and possibly what things
      it fixes and everything else that is related]
---- stop ----

So for example:


happy-eyeball: Add connection delay to Happy Eyeballs

Added a 200ms delay between the first and second address family socket
connection attempts.


connect: Post Happy Eyeball tidy up

Removed some old code that handled synchronous connects. Since all
sockets are now non-blocking the logic has been made a lot simpler.

Many thanks and I hope I haven't trodden on any toes with mentioning this



List admin:
Received on 2013-11-02