[JDEV] Announce: new ICQv7 transport

[alex_]

>>>>> "Filip" == Filip Sielimowicz <sielim at wp-sa.pl> writes:

 Filip> I have too less documentation about protocol, but this is, what we have
 Filip> implemented:

 Filip> - Presences - sending statuses of users.
 Filip> - Presences - adding new users, subscription and "UserAddedU" messages.
 Filip> - this is not finished.
 Filip> - Messages - Normal, URL.
 Filip> - Offline messages request, receiving and ack.
 Filip> - Short and Advanced Search (!)
 Filip> - the protocol is fully implemented, but not all fields are
 Filip> used in responces.

 Filip> What we are going to do soon:
 Filip> - multimessages - no documentation ...
 Filip> - Auth Request/Given/Denied - protocol is implemented, but documentation
 Filip> ... We need some more dokumentation about when/why/how to use it.

 Filip> The source code is based on ickle-0.1.2 ICQ client classes (modified and
 Filip> expanded, some bugs fixed, it compiles under standard RH 6.2, no <sstream>
 Filip> or "sigc" headers ... - this was horrible ...).

 Hello!

 We need some consistency here, I suggest not to mix ICQ-protocol specific
 functionality and Jabber-protocol one. It would be cool you submit your
 implemented ICQ-protocol research to Barnaby Grey (th author of ickle) to
 let him improve libicq2000 functionality, which lacks search, multimessages,
 auth-stuff...

 Filip> We are very interested in cooperation, mainly the documentation and Your
 Filip> source code.

 Yeah, we need to cooperate, let me just to test the transport more carefully..

 Filip> In a moment we have also some problem with stability - bad use of pth
 Filip> causes segmentation faults from time to time... 
 
 I experienced that too, and these segfaults force me not to mix C and C++ code.
 IMHO, C++ is not pth-safe, as soon throwing exceptions 'cause to segfault.

-- 
 Regards, Alex_