Thu, 03 Jun 2010 01:22:31 +0000 |
Mark Doliner |
I noticed that this wasn't being used
|
Fri, 28 May 2010 23:30:20 +0000 |
Elliott Sales de Andrade |
It's probably best just to be explicit about the ordering here.
|
Fri, 28 May 2010 21:26:57 +0000 |
Elliott Sales de Andrade |
For some reason, this line creates the following warning:
|
Wed, 26 May 2010 23:20:18 +0000 |
Elliott Sales de Andrade |
I think it's more accurate to say that a DC is a P2P transfer, so if that
|
Sun, 23 May 2010 21:45:19 +0000 |
Elliott Sales de Andrade |
NULL-ify some reverse links so that there's no double-free on exit.
|
Fri, 21 May 2010 07:43:18 +0000 |
Elliott Sales de Andrade |
C comments only.
|
Fri, 21 May 2010 07:36:59 +0000 |
Elliott Sales de Andrade |
Make it more explicit that incoming and outgoing timeouts are different.
|
Thu, 20 May 2010 09:06:47 +0000 |
Elliott Sales de Andrade |
If there's stuff stuck in the DC queue, then try to send it over the SB if
|
Mon, 17 May 2010 08:45:46 +0000 |
Elliott Sales de Andrade |
Ref the slplink before destroying the DC, or we might lose our slpcall.
|
Mon, 17 May 2010 08:42:51 +0000 |
Elliott Sales de Andrade |
Use msn_dc_fallback_to_p2p where possible.
|
Mon, 17 May 2010 07:56:00 +0000 |
Elliott Sales de Andrade |
This is not a timeout, but an input handler.
|
Mon, 17 May 2010 07:34:52 +0000 |
Elliott Sales de Andrade |
Fallback to P2P if connecting to external IP didn't work immediately.
|
Sat, 15 May 2010 08:02:08 +0000 |
Elliott Sales de Andrade |
If removing a timeout and return FALSE in its handler isn't good, then
|
Sat, 08 May 2010 00:08:01 +0000 |
Elliott Sales de Andrade |
We can't both remove a timeout and return FALSE in its callback.
|
Thu, 06 May 2010 07:36:56 +0000 |
Elliott Sales de Andrade |
This should probably work a bit better. At least, we want to get rid of any
|
Wed, 05 May 2010 09:07:10 +0000 |
Elliott Sales de Andrade |
Valgrind pointed out that dest and src may overlap, so we need to use
|
Wed, 05 May 2010 08:18:35 +0000 |
Elliott Sales de Andrade |
I've seen packet sizes larger than our current maximum. Hopefully 8KiB is
|
Wed, 05 May 2010 08:14:40 +0000 |
Elliott Sales de Andrade |
Remote more seemingly unnecessary stuff.
|
Wed, 05 May 2010 08:11:05 +0000 |
Elliott Sales de Andrade |
I'm not sure what this stuff is, but it looks old and useless.
|
Wed, 05 May 2010 07:33:55 +0000 |
Elliott Sales de Andrade |
I'm pretty sure we need to actually use dc->msg_pos for something, instead
|
Wed, 05 May 2010 07:32:45 +0000 |
Elliott Sales de Andrade |
I think this finally works with both aMSN and the official client, for
|
Wed, 05 May 2010 07:29:24 +0000 |
Elliott Sales de Andrade |
Try and reduce some code duplication.
|
Wed, 05 May 2010 07:26:57 +0000 |
Elliott Sales de Andrade |
Make msn_dc_fallback_to_p2p non-static, and remove some old code.
|
Wed, 05 May 2010 07:02:05 +0000 |
Elliott Sales de Andrade |
The length of a DC packet does not include itself.
|
Sun, 25 Apr 2010 23:38:27 +0000 |
Elliott Sales de Andrade |
It seems like every DC packet requires a length, so why make every little
|
Sun, 25 Apr 2010 23:15:32 +0000 |
Elliott Sales de Andrade |
Obviously, if this comparison fails, I should be returning FALSE.
|
Sat, 24 Apr 2010 20:39:06 +0000 |
Elliott Sales de Andrade |
All these processing functions are called from places where dc != NULL,
|
Sat, 24 Apr 2010 20:37:49 +0000 |
Elliott Sales de Andrade |
Remove some duplicate code.
|
Sat, 24 Apr 2010 06:41:52 +0000 |
Elliott Sales de Andrade |
Support old-style Nonce, which should let us use direct connections to
|
Fri, 23 Apr 2010 22:38:13 +0000 |
Elliott Sales de Andrade |
Save remote nonce, and verify we get the same data when initiating the
|