Bogdan-Andrei Iancu writes:

 > I hesitate to do it because you can get into BRANCH_ROUTE from a 
 > FAILURE_ROUTE - and add_rr_param() does not work inside a FAILURE_ROUTE  
 > (mainly due the static variable I mentioned before) - I need to work a 
 > bit on it.

ok, i didn't know that.  it may then be better to wait until
add_rr_param can also be called from FAILURE_ROUTE.

 > you mean flags per contact/branch after location lookup? we already have 
 > them.....

so if i test registrar/nat_flag in BRANCH_ROUTE, i correctly get the nat
status of that branch?

 > I see no problems (conflicts or duplicity) with the patch, so I guess is 
 > ok to commit it.

it looks nicer in the code if i don't need to first call add_rr_param,
but i don't think it is necessary to apply the patch if there is no
performance advantage over first calling add_rr_param.

-- juha

_______________________________________________
Devel mailing list
[email protected]
http://openser.org/cgi-bin/mailman/listinfo/devel

Reply via email to