Re: [reSIProcate] testing ice4j against reTurn

[Scott Godin <sgodin@xxxxxxxxxxxxxxx>]

Hi Daniel,

Glad to hear it is working good for you.  You are right about the state of the config framework - I'd like to see a framework that reads settings from a text based name/value pair configuration file, similar to what is used in the ichat-gw project under the apps directory.  

I'm not working on any reTurn changes at the moment and I'm open to anyone stepping in that can improve things.  : )

I'm pretty bad a checking or doing anything with Bugzilla at this point - but I agree that it's a good tool to move forward with.  Another option is to markup and edit the wiki page.   

Regards,

Scott

On Sun, Jan 8, 2012 at 4:42 PM, Daniel Pocock <daniel@xxxxxxxxxxxxx> wrote:

On 08/01/12 20:12, Scott Godin wrote:
> Good point.  Right now reTurn requires you to use an actual IP address
> and not 0.0.0.0.  We should modify the program to fail to start if
> in_addrany is specified on the command line.

The implementation seems quite good and it is working, but the config
stuff is a little unfinished - I was curious about whether you are
already working on some of those final touches, or hoping other people
will step in?

I may have a use for running it in a couple of places and I don't mind
tweaking some of these things as I go

Also, I noticed the other outstanding issues listed in the README, I
think it would be good to capture stuff (whether it is bugs or just
things that are unfinished) in Bugzilla, then people can annotate things
or even note the things they are working on

>
> Scott
>
> On Sun, Jan 8, 2012 at 2:07 PM, Daniel Pocock <daniel@xxxxxxxxxxxxx

> <mailto:daniel@xxxxxxxxxxxxx>> wrote:
>
>
>
>     I've just run the ice4j samples:
>
>       ant run-sample
>
>     against reTurn from resiprocate
>
>     It appears to run fine, except when I initially started up reTurn with
>     no command line arguments
>
>     - It bound on 0.0.0.0
>     - The host is multi-homed, (multiple public IPs)
>     - it would receive on one IP and the source address of every response
>     packet was the other IP
>     - consequently, none of the response packets went through the NAT to
>     the UA
>     - there is no clue what is going wrong until you look at it with a
>     packet sniffer
>
>     When I bind explicitly to one of the addresses, it works fine
>
>     Should it detect when it is running on a multi-homed host and exercise
>     some control over the source IP of response packets?
>
>
>     _______________________________________________
>     resiprocate-devel mailing list
>     resiprocate-devel@xxxxxxxxxxxxxxx

>     <mailto:resiprocate-devel@xxxxxxxxxxxxxxx>
>     https://list.resiprocate.org/mailman/listinfo/resiprocate-devel
>
>