logs archiveBotHelp.net / Freenode / #2600hz / 2015 / July / 27 / 2
voxter
a couple times, it didnt blink on the 'first time'
but then did the second try
neurosys
voxter: did you get a chance to review the change log?
voxter
neurosys: i did, but nothing stood out
neurosys
hmm
voxter
just doing a couple more tests here.
here are the release notes
http://downloads.polycom.com/voice/voip/uc/UC_Software_5_4_0_Release_Notes.pdf
neurosys
voxter: it actullay drops the call you placed on pack when you attempt the pickup.
voxter
heh interesting - when i place a call to a user
then do a directed call pickup for them,
the directed call pickup works, but then their voicemail starts to play
callflow execution kept going it seems!
presumably something needs to be sent to the originally executed callflow to tell it to stop after a pickup has occurred
neurosys
voxter: off-topic: new 5.4 supports OpUS :)
only in the 500 and 600s
voxter
neurosys: i saw that yeah! I wonder how it compreas
compares*
lazedo: so the directed pickup uses the sip:kfp stuff as well - does this mean that the kazoo_query code is now obsolete entirely? I suppose I could remove it, but for the instances in which a replaces: is present, but sip:kfp isnt being used, it looks like that would still facilitate a fast pickup, just another way - except that if statement is failing in its current form
neurosys
voxter: seems to me it's just connecting the wrong leg
voxter
neurosys: so, i wasnt seeing what you're now experiencing
neurosys: although, unless you find the way to turn dialing *31 off, and instead sending the invite to the sip:kfp+ contact thats in the dialog header, i'd suggest downgrading to 5.3 anyways
neurosys
voxter: like when i call my cell, park and then retrieve, the retrive begins to play the MoH where the dropped call that was park - left off
voxter: true
voxter: but then i lose my web up.osdIncomingCall.Enabled="0" :(
voxter
what is that feature? and its only in 5.4?
neurosys
oh wait that was 5.3
5.3 was out in may right?
voxter
yeah
there's also a 5.3.1
im running that now, it was released this month
neurosys
voxter: nice
voxter: I can tell you fast pickup on the yealinks do not work either
They do the exact same thing
when you enable fast pickup via dialog info
lazedo
neurosys: voxter i tell you that cisco, polycom, yealink, grandstream, bria, &. ALL work with fast pickup. not sure what issues youre finding, really.
neurosys
lazedo: im downgrading the VVX to 5.3 to see
voxter
lazedo: im unsure what neurosys is experiencing, hes saying its like its bridging him to the wrong leg - once i downgraded my vvx to 5.3, its sending the sip:kfp+ stuff properly again - I'm sure there must be a way to modify this behavior in 5.4, but I cant find i tyet.
lazedo: the only "bug" i have (and this may be fixed in > 3.20 already) is when i do a directed pickup on someone, after their ringing timeout (in their callflow) I am handed off to the next callflow item, which happens to be voicemail
so I pick up the call, and then 20-30 seconds later, depending on their ring timeout, both parties hear that persons voicemail
lazedo
freeswitch version ?
voxter
lazedo: FreeSWITCH Version 1.4.7~64bit ( 64bit)
lazedo
that seems like a not handled kazoo-pickup , ie, wrong leg
voxter
yeah... Like i said, im running 3.20, so i dont know if its already been fixed upstream or not
lazedo
voxter: freeswitch 1.4.7 should be oout build with the proper patches
voxter: the code for pickup on a-leg is merged into freeswitch master, it will be available in 1.6
voxter
can try upgrading to master on a dev box and re-testing. we might be able to produce a patch for it too.
lazedo: oh so is that just something we need to wait for 1.6 on?
this 1.4.7 is from the 2600hz repo
kazoo-freeswitch-R15B-1.4.7-10.el6.x86_64
lazedo
voxter: no, you dont have to wait for 1.6
voxter: 2600hz repo stable is 1.4.7-11
neurosys
lazedo: ah so it's you? :Pp
voxter
oh weird
we're attached to the staging repo
neurosys
1.4.15 is stable no?
voxter
lazedo: I can try fixing that and updating to -11 to see if its been handled there.
neurosys
1.4.15-2*
voxter
lazedo: the only other "issue" - and i dont know if you want to call it that, is when a replaces header is sent to kamailio WITHOUT sip:kfp+ in the To:, the kazoo_query seems to fail for that older method. Is there something that needs to be modified to make that 'fallback' method, without sip:kfp+ work?
i dont know enough about how that older method should work to understand why its failing
lazedo
voxter: the fallback still works
voxter
lazedo: well, eventually yes, but only after the kazoo_query mechanism retries and fails and gives up - the fallback method is to just send it through as dialplan, and ignore the replaces: header
lazedo
voxter: maybe your problem is that the kfp is int replaces header
voxter: *in the
voxter
lazedo: instead of the call-id? that sounds likely, because each time ive traced this query, the call-id that it passes in the replaces header, is not a call ID in freeswitch
it looks as though its getting the "call id" from the dialog
the dialog of the kfp, that is
and then if i query freeswitch for the parked call, with that call-id, it doesnt exist.. the call id is different
lazedo
voxter: create a pastie with a sip trace of what the ua sends to kamailio
voxter
lazedo: sure. I'll include parking and picking up.
Will take a minute - i just need to upgrade the firmware again to 5.4 to change the behavior
lazedo
voxter: capture also the NOTIFY sent to the UA after park
voxter
lazedo: will do. I'll capture everything to the UA
« prev 1 2 next »