Wireshark-dev: Re: [Wireshark-dev] ADNS alternative
From: Jaap Keuter <[email protected]>
Date: Sun, 04 May 2008 01:02:02 +0200

True, but the DNS resolver is based on the event loop library it's part of. So this would have to be used as our event loop as well. Which will bring us in conflict with the GTK event loop I guess.

Sébastien Tandel wrote:

there is also libevent which is portable, maintained and seems well-documented :-)


On Fri, May 2, 2008 at 9:08 AM, Jaap Keuter <[email protected] <mailto:[email protected]>> wrote:

    Looking at the suggestions I see this:
    Poslib seems to have been developed until 2005, similar problem as
    ADNS. It's
    a C++ lib as well, causing a binding problem.
    UDNS looks interesting as well. As it's conceptual model is further
    from ADNS
    than c-ares is it might take more work to get it interfaced. Another
    thing not
    apparent is the cross platform support. There is mention of a Win32
    port, but
    that is stated to be out of date.


    Graham Bloice wrote:
     > Gerald Combs wrote:
     >> Jaap Keuter wrote:
     >>> Hi list,
     >>> I've stumbled upon c-ares
    (http://daniel.haxx.se/projects/c-ares/) as a
     >>> possible replacement for ADNS, which we currently have as async
    DNS resolver
     >>> library.
     >>> Pro's: Still being developed, IPv6 support, multi-platform, MIT
     >>> Con's: May suffer from same memory allocation issues on Win32
    as does ADNS.
     >>> Does anyone have further insights into the use of this library?
     >> It looks interesting. The fact that it's documented and actively
    developed is
     >> encouraging.
     >> _______________________________________________
     > It also has some Freexxx() functions that seem to handle freeing
     > allocations that the lib makes.
     >  From the web site, they also have a link to other similar libs,
     > (http://posadis.sourceforge.net/) and UDNS
     > (http://www.corpit.ru/mjt/udns.html) look interesting.
     > --
     > Regards,
     > Graham Bloice