It seems we have this problem and a few others in the sipp project.
- issue management - at this moment there is something in sourceforge,
but nobody seems to take it into account
- I have seen a few patches on this list valid for various versions of
sipp and for the last year or so.
- on the list sipp-devel I have not seen any kind of activity since
november (when I joined the list).
- the most interesting developments have been through patches that
individual companies or developers used and shared on the list and none
has seen integration in the mainline (maybe I am wrong, but most of it
for sure).
>From my point of view we need:
- we need an issue management system
- a more active devel team (sorry guys)
- some new features in sipp 4.0 like:
-- integrating a logging/debugging framework - I am thinking
pantheios here (http://pantheios.sourceforge.net/) - maybe a more
experienced developer
-- integrating the nice dtmf patch from Dmitry Kunilov
-- some code restructuring especially for better mentainance
Maybe if we gather enough devels we can make a more dynamic community
arround sipp.
Regards
Radu Costin
Pe 24.05.2011 10:12, Paul D.Smith a scris:
> Can someone tell me where and I how I report bugs? I believe I have
> found two plus an enhancement:
>
> 1. get_host_and_port should not use strcpy
>
> This uses strcpy() to "copy" out the remote host. Unfortunately the
> source and destination passed to strcpy() overlap and this is
> specifically forbidden by strcpy(). The result in my case was a
> single nibble (4 bytes) being copied incorrectly.
>
> I've got a local fix which determines the length of the data to be
> copied using strlen() and then uses memmove() instead, which is
> overlap-safe.
>
> 2. IPv6 can fail
>
> I'm trying to send a REGISTER to a registrar via an outbound proxy.
> I'm using the --un option and this failed indicating that the protocol
> was not supported. After a lot of searching and adding printf() I
> discovered that the code which creates the "un" sockets is passed a
> value "is_ipv6" -- but this value has not been set-up at this point.
>
> I have a local fix which passes local_ip_is_ipv6 instead, which looks
> right to me.
>
> As a general comment, the IPv6 code is plain nasty with too many
> globals and parameters being passed around. SIPp should use a single
> global (there cannot be any use for more than one) or no globals and
> always pass parameters.
>
> Enhancement. There's no debugging/tracing!
>
> See #2. I spent most of a day narrowing this down by adding
> printf()s. SIPp should have a debug/trace class which is enabled in
> debug builds and perhaps with an additional command line parameter.
> This should trace at every key decision point so that bugs can be
> found quickly and rapidly.
>
>
>
> ------------------------------------------------------------------------------
> vRanger cuts backup time in half-while increasing security.
> With the market-leading solution for virtual backup and recovery,
> you get blazing-fast, flexible, and affordable data protection.
> Download your free trial now.
> http://p.sf.net/sfu/quest-d2dcopy1
>
>
> _______________________________________________
> Sipp-users mailing list
> Sipp-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/sipp-users
------------------------------------------------------------------------------
vRanger cuts backup time in half-while increasing security.
With the market-leading solution for virtual backup and recovery,
you get blazing-fast, flexible, and affordable data protection.
Download your free trial now.
http://p.sf.net/sfu/quest-d2dcopy1
_______________________________________________
Sipp-users mailing list
Sipp-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sipp-users
