On 12/05/08 18:05, Juha Heinanen wrote:
Henning Westerholt writes:
Already finished features and the roadmap can be found as usual in our wiki at [2, 3].
in my opinion the roadmap should include normal failure route processing of also tcp connect failures. now it is very bad that script writer needs to handle failures in two places in the script depending if udp or tcp is used for a particular contact.
It is in my todo lists, but I was pretty busy with other projects.
also, i would like to see a conclusion regarding local route implementation, which was left out by default from 1.4.
Work is undertaken to fix what is possible, the new $T_req(pv) and $T_rpl(pv) got a similar solution (however, for local route is needed more than what was done here so far). As we are going to have first the SIP Router core and then go for full asynchronous mode and contexts (sip router has already kind of asynchronous processing support already), for 1.5.0 will be a jailed protection against the issues caused by local route.
There is another problem here, with the processes started by MI transport modules - local route is executed by these processes but they are not properly initialized, as a sip message worker. Probably another temporary hack to fix this one as well. Let's see how it gets until we freeze the code...
Cheers, Daniel