Mobicents weekly meeting notes, December 19, 2012
605 views
Skip to first unread message
Ivelin Ivanov
Dec 19, 2012, 2:36:39 PM12/19/12
to mobicents-public
Public IRC meeting on freenode.net #mobicents
Starts every Wednesday at 10am CST.
Attendees:
Bartek, Charles, George, Jean, Pavel, Sergey, Tom Q, Yulian, Ivelin
Summary:
#1 JSLEE: fixing bugs for Telscale edition; waiting on HTTP NIO RA feedback
#2 Diameter: aiming for Diameter 1.5.0.FINAL + Charging Server 1.0.0.ALPHA1 next week
#3 SS7: working on 2.0.0.BETA4 (36 issues / 21 fixed); Telscale SMSC Gateway 6.1.3.GA Released
#4 Restcomm: focus on handling out of band RTP events for DTMF
#5 SIP Servlets: Closed several issues on 2.x; merged NIO code in JSIP
#6 Media Server: road block with MGCP RI; working on clean room implementation of the MGCP API - JSR 23
#7 Management framework: proposed unified management framework; posted blog for community review http://goo.gl/skNFc
#8 Docs: helped Tom U with MSRP RA docs; still cleaning up JSLEE RA docs
#9 JSIP: discussing tools to cover base JSIP code to Android format
#10 Productization: Finalized Telscale-smsc-6.1.3.GA release; finishing Telscale-USSD-6.1.3.GA release
Log:
------
[10:08] <ivelin> #1 JSLEE
[10:09] <ivelin> hello?
[10:10] <@emmartins> hi there
[10:10] <@emmartins> sorry was on mail
[10:10] <ivelin> figured
[10:10] <ivelin> is the mail 50 proof or higher?
[10:10] <@emmartins> I'm working out the current SLEE issues which will be included on Telscale, and reviewing the nice contribution
[10:11] <@emmartins> by Tom, wrt MSRP RA
[10:11] <@emmartins> I have somehting like 15k unread mails
[10:11] == charles_r [c9bfc34a@gateway/web/freenode/ip.201.191.195.74] has joined #mobicents
[10:12] <@emmartins> spread over 3 gmail accounts, red hat, icloud, etc...
[10:12] <@emmartins> it's a nightmare
[10:12] <@emmartins> most are commits noise that are not related with my work
[10:12] <ivelin> what are some of the big issues?
[10:13] <@emmartins> we should work out that in mobicents btw
[10:13] <@emmartins> we have sepaarted projects, but a common big output google group notifying about commits
[10:13] <ivelin> filters baby, they do magic
[10:14] <@emmartins> there are no big issues, just small features such as more stats on event routing
[10:14] <@emmartins> and small bugs
[10:14] <@emmartins> like working MGCP with latest MMS
[10:14] <jean> yeah filter are working out pretty good
[10:14] <jean> for me
[10:14] <ivelin> small things add up
[10:15] <@emmartins> I was not able to start on updated SMPP RA, and that is a fail
[10:15] <@emmartins> ivelin: did Alex said anything wrt common admin or any other slee stuff?
[10:16] <@emmartins> baranowb: hello, any update on snmp?
[10:16] <ivelin> yes, he blogged about his management findings
[10:16] <jean> alex published a blog post wrt common admin http://ammendonca.blogspot.pt/2012/12/build-your-own-mobicents-dashboard-in.html
[10:16] <ivelin> I think he has a good action plan in place
[10:17] <ivelin> essentially each project should expose management instrumentation via JMX
[10:17] <ivelin> then Alex will take care of CLI, REST, JavaScript, Charting etc. on top
[10:17] == fcr has changed nick to fran
[10:17] <@emmartins> that we already have
[10:17] <@baranowb> emmartins: only played with SLEE a little, I hope I can get it going on xmas week, shouldnt take long though
[10:18] <@emmartins> and I don't really like another tech beyond GWT at this point
[10:18] <ivelin> its very important for the JMX instrumentation to be meaningful to sysadmins. As long as there are good descriptions of the attributes and methods in English, most of the tooling will work nicely out of he box
[10:18] <@emmartins> no way we are going to redo the console again
[10:19] <@emmartins> what is gwt biggest issue, static compilation?
[10:19] <@emmartins> even as7 is ok with it atm
[10:19] <ivelin> GWT, JQuery, whatever Alex decides is fine for me. He is good with UI design.
[10:20] <ivelin> more important to me is that we don't have to reinvent CLI, SNMP, etc. frameworks for each project
[10:20] <@emmartins> sure, but that is something else really
[10:21] <@emmartins> what I was looking for
[10:21] <@emmartins> was a common definition of what to use on each management/monitoring client
[10:21] <@emmartins> from what we already use
[10:21] <@emmartins> not new directions
[10:22] <@emmartins> anyway, something to discuss better and sync offline
[10:22] <ivelin> yes, bring it up, I don't quite understand what you mean
[10:22] <ivelin> #2 Diameter
[10:22] <@emmartins> wait
[10:22] <ivelin> oops
[10:22] <@emmartins> do you have Amit's update on SLEE
[10:23] <@emmartins> or related projects that can impact SLEE dev?
[10:23] <ivelin> he said he hasn't looked at the HTTP RA yet
[10:23] <ivelin> Sergey, do you have other updates related to JSLEE?
[10:23] <@baranowb> ss7 stuff has been released, next release is in 2013, no new changes iirc
[10:23] <@emmartins> baranowb: thanks, please share with us a weekly dev progress
[10:23] <@emmartins> on snmp
[10:24] <@emmartins> we need at least a bit of progress every week
[10:24] <@emmartins> and essentially we are already late for our plans of telscale release this month
[10:24] <svetyutnev> sorry, no update for SLEE from SS7
[10:24] <svetyutnev> from me
[10:25] <@emmartins> svetyutnev: the SS7 repo was updated?
[10:25] <@emmartins> on SLEe
[10:25] <@emmartins> (wrt compatibility with the new SS7 release)
[10:25] <svetyutnev> as far as I remember - no
[10:26] <@emmartins> ok
[10:26] <@emmartins> ivelin: please add a note on the chat notes to work this matter too
[10:26] <@emmartins> sync ss7 slee repo with ss7 before actual slee release happens
[10:27] <@emmartins> well, never mind, I need to sync with amit on plenty of stuff, better to contact him for a broader chat
[10:28] <ivelin> Eduardo, which matter?
[10:28] <svetyutnev> sirry , last update for CAP RA was 4 days ago
[10:28] <@emmartins> most important, update plans for telscale release
[10:29] <@emmartins> I will mail all involved in SLEE so we find an hour or so to chat here
[10:30] <ivelin> so start an offline topic on the telscale jslee roadmap. Let's iron that out.
[10:30] <ivelin> ok, that works too.
[10:30] <@emmartins> n, mail exchanges are too slow
[10:31] <@emmartins> ok, you can move on the chat if there is no questions
[10:31] <slegrik1> emmartins, include me as well in that chat, pls
[10:31] <@emmartins> sure, np
[10:31] <ivelin> #2 Diameter
[10:32] <ivelin> From Alex: "There's no major update for Diameter. Will try to release Diameter 1.5.0.FINAL + Charging Server 1.0.0.ALPHA1 during next week. "
[10:32] <ivelin> Bartek, do you want to rebuff Alex? Now is the time. He can't respond.
[10:34] <ivelin> chicken
[10:34] <ivelin> #3 SS7
[10:34] <@baranowb> nothing from me on diameter fornt unfortunetly
[10:34] <svetyutnev> SS7:
[10:34] <svetyutnev> 2.0.0.BETA4 Release : More issues registered. Total 36. Open 21. Release date is 10th Jan 2013
[10:34] <svetyutnev> For jss7 stacks major issues are:
[10:34] <svetyutnev> linkedId operations in CAP (and MAP) was processed badly and it is fixed now
[10:34] <svetyutnev> some missed management events are added for sctp, m3ua, sccp stacks
[10:34] <svetyutnev> progress in implementing insertSubscriberData MAP operation
[10:35] <svetyutnev> now I am in fixing some abnormal situations processing in TCAP/MAP/CAP
[10:35] <svetyutnev> 2.1.0.BETA1 : New branch will be created after 2.0.0.BETA4 release.
[10:35] <svetyutnev> 2.1.0.BETA1 will be focused on CAP V3 operations for GPRS
[10:35] <svetyutnev> Follow the announcement https://telestax.zendesk.com/entries/22636323-telscale-6-1-3-ga-smsc-is-now-available.
[10:35] <svetyutnev> Next release will include store-and-forward feature
[10:35] <ivelin> lots of community activity on jss7
[10:35] <svetyutnev> It will include USSD Push/Notify functionality. Completing docs and release script.
[10:35] <svetyutnev> EIR : Adding the release job @ Jenkins.
[10:35] <svetyutnev> Fixing Cassandra integration issue with Jenkins and pending Docs.
[10:38] <ivelin> lots of SS7 activity before the holidays.
[10:38] <ivelin> let me check if Dmitri has updates on the card
[10:39] <@baranowb> there is release planned early next year iirc
[10:39] <ivelin> yes, its a packed roadmap for 2013
[10:39] <ivelin> ok, he's not responding atm. Will ask him for offline update.
[10:39] <ivelin> #4 Restcomm
[10:41] <tommyq> Okay so not much has changed here... I am still working on OOB patch for MMS. I should have working code by the end of the week.
[10:42] <ivelin> ok, that's a big change. Let's see how it works out.
[10:42] <yulian_o> what is oob?
[10:42] <ivelin> out of band
[10:43] <ivelin> rtp events for dtmf
[10:43] <gvag> from my side, i finished the performance tests, there was some issues with the mgcp-client which Yulain is working
[10:44] <ivelin> Tom is working on a separate git clone for this change. Please keep an eye on it.
[10:46] <yulian_o> if you ask me he is completely out of direction
[10:46] <ivelin> lol
[10:46] <ivelin> how so?
[10:47] <jean> yulian_o, no way to fix it in current MMS from your POV ?
[10:47] <yulian_o> first of all small comment : you changed the dtmf package classes names to dmtf , the problem was in folder name so it had to be opposite
[10:47] <yulian_o> about out of band : he is trying to detect it in input
[10:47] <yulian_o> which is not good
[10:48] <ivelin> yes, I noticed the package name, but it was wrong before Tom looked at it. Agree that the package has to be dtmf
[10:48] <ivelin> yes, and is that wrong?
[10:48] <ivelin> why?
[10:48] <yulian_o> yes i know , and he wanted to change what is good , but it should be dtmf everywhere not dmtf everywhere
[10:48] <yulian_o> why is not good?
[10:49] <ivelin> agree on the package name
[10:49] <ivelin> why the oob direction is not good?
[10:50] <yulian_o> because this way it will work only on first point. so in case of restcomm if he will request to detect tones on ivr it will never happen since tones will not go any further then packet relay ( if i remember correctly its first endpoint for ua )
[10:51] == mode/#mobicents [+o baranowb] by ChanServ
[10:51] <tommyq> still not done :) but everyone is entitled to an opinion :)
[10:51] == dmisol [25be3422@gateway/web/freenode/ip.37.190.52.34] has joined #mobicents
[10:51] <tommyq> continue
[10:52] <ivelin> Yulian, I think that was an earlier design. Tom decided to take a longer route and do it right.
[10:52] <ivelin> The events will be flowing through the core all the way to the IVR
[10:53] <yulian_o> ok
[10:53] <yulian_o> i only see what he has committed
[10:53] <tommyq> which is what I'm saying :)
[10:53] <tommyq> Yulian you will be the first to get pinged to validate the work
[10:54] <tommyq> still need your blessing to make sure I get it right
[10:54] <ivelin> ok, let's give Tom a little breathing room to finish off and request review
[10:54] <tommyq> after all it is your baby
[10:54] <tommyq> :)
[10:54] <ivelin> #6 Media Server
[10:54] <ivelin> talking of which :)
[10:55] <yulian_o> working on mgcp client , have some problem with mgcp ri , but got solved it by creating parsing classes separately from parameters and requests
[10:55] <jean> don't forget #5 after #6 just saying
[10:55] <jean> :-)
[10:56] <yulian_o> also next release will be parsing byte arrays and not strings so it should optimize perfomance
[10:56] <ivelin> argh. I spend too much time counting from 1 to 10 with my 3 year old pooper.
[10:56] <yulian_o> this is final change for mgcp client excluding bugs if will be any
[10:57] <tommyq> :)
[10:57] <yulian_o> thats it for mms
[10:57] <ivelin> ok, are you essentially rewriting the RI but keeping the APIs?
[10:57] <yulian_o> about dtmfs - out of band
[10:57] <yulian_o> finally got trace from tom
[10:57] <yulian_o> for the problematic ones
[10:58] <yulian_o> all of them ( google , medical ) are behaving same way
[10:58] <tommyq> yes sorry about the delay the customer had a series of mishaps that delayed feedback
[10:58] == charles_r [c9bfc34a@gateway/web/freenode/ip.201.191.195.74] has quit [Ping timeout: 245 seconds]
[10:58] <yulian_o> while sending audio each 20ms , out of band is send on separate channel each 10ms , and it comes together with silence on audio channel ( 3 frames per each 20 ms )
[10:58] <ivelin> Yulian do you understand better what the problem is now?
[10:59] <yulian_o> i think it can be resolved without going deeply into core
[10:59] <yulian_o> just need to change out of band to inband translation
[10:59] <ivelin> so when you mix inbound there is a sequence of DTMF and silence which is detected as multple tones instead of one?
[10:59] <yulian_o> it should generate 5 frames of tone when first packet for tone is received and then simply block other packets that belongs to same tone
[11:00] <yulian_o> no silence are complete zeroes ( when decoded ) so it does nothing
[11:00] <yulian_o> the problem is the way out of band packets are arrived
[11:01] <yulian_o> mms receives more then it moves to core ( it moves one packet per each 20ms ) so from 200ms tone you get 400ms packets which are send much after tone detected to core
[11:01] == charles_r [c9bfc357@gateway/web/freenode/ip.201.191.195.87] has joined #mobicents
[11:02] <yulian_o> so after mgcp i will change the code for outofband generation
[11:02] <yulian_o> and have one more small issue reported by client
[11:02] <ivelin> I see
[11:02] <yulian_o> hope to complete it in the end of week
[11:02] <ivelin> that may help a little with the specific issues at hand
[11:02] <ivelin> but it won't solve all out of band issues
[11:03] <ivelin> like compression problems
[11:03] <yulian_o> what do you mean by compression?
[11:03] <ivelin> and ability to send out of band to remote RTP end points
[11:03] <ivelin> g729, ilbc, etc
[11:04] <yulian_o> out of band have nothing in common with codecs since it generates raw pcm
[11:04] <yulian_o> which is sent to core
[11:04] <yulian_o> out of band tones sending is different question
[11:04] <ivelin> that's true
[11:05] <ivelin> ok, no extra distractions from me
[11:05] <ivelin> #5 SIP Servlets
[11:06] <jean> #1 WebRTC/SIP over WebSockets - Roadmap to be defined by Orange Labs first week of January
[11:06] <jean> #2 MSS in Arquillian - No news here
[11:06] <jean> #3 JSR 359 - George accepted as EG Member. Going to JSR 359 F2F Meeting with Jean second week of January
[11:06] <jean> #4 MSS on ClickStack CloudBees - Still in discussion with CloudBees on enabling SIP in their infrastructure.
[11:06] <jean> #5 RestComm - Working on Out of Band Patch to MMS for fixing DTMF Issues seen at customer. Finished performance test, MGCP Client not good enough being reworked.
[11:06] <jean> #6 MSS 2.X - Closed a couple of issues added domain integration to MSS on AS7. TLS NIO has good performance reported by customer. NIO Branch merged to JSIP Trunk
[11:09] <ivelin> what's the latest from CB. Its been over a week.
[11:09] <jean> cloudbees needs to bainstorm internally
[11:09] <jean> on how they want to go about supporting VoIP
[11:09] <gvag> and then we will have a meeting with them to see how we move on
[11:09] <jean> will have an answer by January
[11:10] <ivelin> ok, good
[11:12] <ivelin> #7 Management framework
[11:13] <ivelin> Alex is on PTO, but as it was mentioned earlier in this chat
[11:13] <ivelin> he has an action plan
[11:13] <ivelin> JMX at the app level and on top of that bridges for CLI, REST, SNMP, JavaScript
[11:15] <ivelin> #8 Docs
[11:15] <charles_r> worked with tom on msrp
[11:15] <ivelin> Tom U , right?
[11:16] <charles_r> i suggested he added some use case
[11:16] <yulian_o> sorry for interrupting i would like to talk a bit on jain sip , if its possible can you add it in the end of meeting?
[11:16] <charles_r> yes ivelin tom u
[11:17] <ivelin> Yulian, yes
[11:17] <ivelin> ok
[11:17] <ivelin> what else Charles?
[11:18] <charles_r> all RA have the same issues as they are created using the same doc
[11:19] <charles_r> we need to rework the base doc so that new ones are up to date
[11:20] <charles_r> do we have a template for RA docs?
[11:21] <charles_r> @emmartins could confirm this
[11:21] <jean> charles_r, you can look at the SIP RA
[11:21] <jean> it can act as Template
[11:21] <jean> since this is the most widely used RA probably
[11:21] == charles_r [c9bfc357@gateway/web/freenode/ip.201.191.195.87] has quit [Quit: Page closed]
[11:21] <@emmartins> charles_r: there is a template in svn
[11:21] <jean> http://code.google.com/p/jain-slee/source/browse/?repo=sip#git%2Fresources%2Fsip11%2Fdocs
[11:21] <@emmartins> but essentially copy/paste is more or less same work
[11:23] == charles_r [c9bfc34a@gateway/web/freenode/ip.201.191.195.74] has joined #mobicents
[11:23] <charles_r> ok jean
[11:25] <charles_r> will continue on more RA slee doc review.
[11:27] <charles_r> ping
[11:27] <yulian_o> pong
[11:28] <jean> pong
[11:28] <tommyq> pong
[11:28] <jean> lol
[11:28] <jean> shall wemove to JIAN SIP
[11:28] <tommyq> :)
[11:28] <tommyq> yep
[11:29] <charles_r> all post from old site are now on the new telestax site
[11:29] <jean> thx for that charles_r
[11:30] <jean> ok #9 JAIN SIP
[11:30] <yulian_o> ok ,
[11:30] <yulian_o> i am working currently on android application
[11:30] <yulian_o> since sip is supported only started from 2.3.6
[11:31] <yulian_o> and also android os loves to remove features sometimes
[11:31] <yulian_o> its not reliable to use built in sip stack
[11:31] <yulian_o> so i wanted to use jain sip
[11:31] <jean> is there still a lot of handsets below 2.3.6 ?
[11:31] <yulian_o> there is enough with 2.3.2 and 3
[11:31] <yulian_o> 2.2 also
[11:31] <jean> ok
[11:32] <yulian_o> there are several problem while trying to use jain sip on android
[11:32] <yulian_o> 1) android does not allows java.* and javax.* classes
[11:32] <jean> do they expose the javax.sip ?
[11:32] <yulian_o> 2) logging - file system is read only so it does not works by default
[11:33] <yulian_o> i am not sure , i think there are both javax,sip and gov.nist packages in cellphone but i can not use them
[11:33] <yulian_o> so in the end of all i have started a stack , but i had to rename classes ( did refactoring with eclipse ) and modified logging classes
[11:34] <jean> hum... It is going to be a hell to maintain
[11:34] <yulian_o> the problem starts when i will want to upgrade the version
[11:34] <jean> between main javax and renaming
[11:34] <jean> what modifications of logging is needed
[11:34] <jean> ?
[11:34] <jean> you can set the prop to off by default
[11:34] <jean> to avoid logging
[11:35] <yulian_o> it does not allows to write for files
[11:35] <yulian_o> even when you disable logging it tries to create a file when stack is started and therefore gets exception
[11:36] <yulian_o> but logging is less important since i think i can change it through settings ( modify classes which may be usefull for android for example )
[11:36] <yulian_o> the main problem is javax.
[11:36] <jean> right we can even include a fix in JAIN SIP for logging to avoid that
[11:36] <jean> issue
[11:36] <jean> right javax is very problematic
[11:37] <yulian_o> anyone have some thoughts on how to maintain this ( i mean both old structure and new structure )
[11:37] <jean> I'll need to think about that
[11:37] <jean> but I would think of a tool rather
[11:37] <yulian_o> i have one option that may be possible
[11:37] <jean> that would do the conversion
[11:37] <jean> instead of maintaining a fork
[11:37] <yulian_o> we can create org.mobicents or anything else instead of javax.
[11:37] <jean> take jain sip run the tool
[11:37] <jean> gives you a jar
[11:38] <jean> with no javax
[11:38] <yulian_o> and then we can create javax. interfaces ( i think there are only interfaaces ) and extend from org.mobicents
[11:38] <jean> no fork to maintain
[11:38] <ivelin> #9 was actually Productiozation
[11:38] <jean> like post process tool
[11:38] <ivelin> you stole it from Pavel :)
[11:38] <yulian_o> sorry :)
[11:38] <jean> just like #6 and #5 sorry about that :-)
[11:39] <jean> yulian
[11:39] <slegrik1> nevermind, I can be #10
[11:39] <jean> something like
[11:39] <yulian_o> 10x
[11:39] <jean> ant android-convert
[11:39] <yulian_o> we can avoid any conversion
[11:39] <yulian_o> if we will extend interfaces
[11:39] <jean> that will turn javax.sip to org.mobicents.sip
[11:39] <gvag> or maybe maven plugin can do the work so we have a profile for android
[11:39] <yulian_o> for example org.mobicents.SipStack
[11:40] <yulian_o> and javax.sip.SipStack extends org.mobicents.SipStack
[11:40] <yulian_o> then ri will use javax
[11:40] <jean> no we can't
[11:40] <yulian_o> why?
[11:40] <jean> it means we need to change the API classes
[11:40] <jean> and those are licensed by Oracle
[11:40] <yulian_o> no
[11:40] <yulian_o> they will remain a same
[11:40] <yulian_o> you mean you can not change a code?
[11:41] <jean> right
[11:41] <jean> all the javax classes cannot be changed without going through a new rev of the spec
[11:41] <jean> can you give a gist example ?
[11:41] <jean> of what you have in mind ?
[11:41] <yulian_o> what do you mean gist?
[11:42] <jean> sample of what you have in mind
[11:42] <yulian_o> you mean file structure?
[11:42] <jean> yes
[11:42] <jean> what you want to modify
[11:43] <yulian_o> org.mobicents.sip package , 100% same structure as current javax. just package name changed
[11:43] <jean> impl is using javax.sip
[11:43] <jean> those will have to be modified
[11:43] <jean> to use org.mobicents.sip
[11:43] <jean> is that what you have in mind ?
[11:43] <yulian_o> then javax. package will have all classes it have now but instead of functions defined it will simply extends interfnaces from org.mobicents.sip
[11:43] <yulian_o> no
[11:44] <yulian_o> there will be 2 classes for each , same everything just different package
[11:44] <yulian_o> this way you can generate ri with mobicents classes and also generate ri with javax classes
[11:44] <jean> duplicate code ?
[11:44] <yulian_o> no
[11:44] <yulian_o> lets take example
[11:45] <yulian_o> there is javax.sip.SipStack right?
[11:45] <yulian_o> so we will copy it to org.mobicents.sip.SipStack
[11:45] <yulian_o> and then instead of current javax.sip.SipStack there will be
[11:45] <yulian_o> public interface javax.sip.SipStack extends org.mobicents.sip.SipStack {}
[11:45] <yulian_o> thats it
[11:46] <yulian_o> so you have one file with code , second that extends it
[11:46] <yulian_o> this way you keep old naming working
[11:46] <yulian_o> while allowing to use it with new naming
[11:46] <jean> yeah but my point is that
[11:46] <jean> javax.sip.SipStack cannot be modified
[11:46] <jean> to extend org.mobicents
[11:46] <jean> not allowed by the license
[11:48] <yulian_o> as i see there are several releases for each file
[11:48] <yulian_o> 1.0 , 1.1 and 1.2 if i am not mistaken
[11:48] <jean> this was on maintenance release of the spec
[11:48] <jean> back in 2006
[11:49] <yulian_o> problem :0
[11:49] <yulian_o> :)
[11:49] <jean> 1.2 is corresponding to JSR 32
[11:49] <jean> let's just have a convert tool
[11:49] <jean> should be easy to do
[11:50] <yulian_o> ok , another question. can we add classes to jain-sip-ri ( not javax ) ?
[11:50] <jean> and JAIN SIP Community will not like having org.mobicens classes :-)
[11:50] <jean> yes
[11:50] <jean> jain sip ri is fine
[11:50] <jean> all gov.nist can be modified
[11:50] <yulian_o> so i can create logging clases for android
[11:50] <jean> yes
[11:50] <jean> the logging system is already pluggable
[11:51] <yulian_o> yes i know that
[11:51] <jean> you can contribtue those classes
[11:51] <jean> to JAIN SIP
[11:51] <yulian_o> i will
[11:51] <yulian_o> ok i will dig it further to see how and what can be doen
[11:51] <yulian_o> done
[11:51] <jean> what we can do
[11:51] <jean> is havce a maven profile
[11:52] <jean> as with each release
[11:52] == dmisol [25be3422@gateway/web/freenode/ip.37.190.52.34] has quit [Ping timeout: 245 seconds]
[11:52] <jean> snapshot
[11:52] <jean> or full release
[11:52] <jean> we can publish both the regular jar
[11:52] <jean> and the android jar
[11:52] <jean> and include the svn revision number
[11:52] <jean> for tracking if any issues
[11:52] <yulian_o> ok , in any case i will try to find the way to access javax.sip from android since if its there there should be some way to work with it
[11:53] <jean> we can just do a simple grep replace script
[11:53] <jean> yulian_o, you will need to differentiate in this case
[11:53] <jean> for prior to 2.3
[11:53] <jean> you will need to include javax.sip
[11:53] <jean> which may break
[11:53] <jean> for devices 2.3+
[11:53] <jean> or ship 2 versions of the app
[11:53] <jean> one for prior 2.3 or post 2.3
[11:54] <jean> a convert tool might be easier
[11:54] <jean> to have one stack that works for all devices without any checks
[11:54] <yulian_o> yes thats the optimal
[11:55] <yulian_o> 10x for help
[11:55] <jean> np
[11:56] <gvag> I think that you need to consider the sip stack of android
[11:56] <gvag> if its already running then there will be a problem
[11:56] <gvag> for ports for example
[11:56] <jean> I don't think the SIP Stack runs by default does it ?
[11:56] <yulian_o> it does after 3.x
[11:57] <yulian_o> or 2.6
[11:57] <jean> ok pick a differnt port then 5080 by example
[11:57] <gvag> maybe check if it is running and use it or start your stack
[11:57] <yulian_o> But when trying in a 2.3 emulator, it appears that the same classes (but probably a different version) already exist in the framework, and that causes problems.
[11:57] == dmisol [25be3422@gateway/web/freenode/ip.37.190.52.34] has joined #mobicents
[11:58] <gvag> :(
[11:58] <yulian_o> it looks like android took the javax. ( api jar ) modified it and put it to OS
[11:58] <yulian_o> same with gov.nist.javax.sip
[11:58] <yulian_o> so when you try to access those files even when you add jar OS gets you its local copy
[11:59] <jean> yeah Google is using the JAIN SIP Stack
[11:59] <jean> old version
[11:59] <jean> IIRC
[12:00] <jean> #10 Productization
[12:00] <slegrik1> ok
[12:00] <slegrik1> and I jumped immediatly on Telscale-ussd-6.1.3.GA release, preparing release scripts. I am nearly done
[12:00] <slegrik1> so I expect to have a first binary for testing tommorrow. The ultimate plan is to have release before Xmass... to get one more gift
[12:00] <slegrik1> Did not find time for anything else, still having licensing in the buffer - once USSD GW Telscale release is out I will resume on this topic
[12:01] <slegrik1> that's it
[12:02] <jean> Great
[12:02] <jean> nice productization rooll
[12:02] <jean> rollouts
[12:02] <jean> thx pavel
[12:04] <jean> I guess that's all for today
[12:05] <jean> thx everyone !
[12:05] <ivelin> any other topics for today?
[12:05] <ivelin> next week is holiday for most of us
[12:05] <ivelin> so the next weekly meeting will be on January 2nd
[12:05] <tommyq> Happy Holidays You Guys!
[12:06] <gvag> happy holidays to everyone
[12:06] <ivelin> Happy Holidays everyone! Have fun, be safe and be ready for a great 2013! :)
[12:06] <jean> happy holdiays !
[12:06] <slegrik1> happy holidays to all of you !
[12:06] <charles_r> what no meeting foe 2weeks?
[12:07] <jean> 1 week charles_r
[12:08] <charles_r> ah ok
Reply all
Reply to author
Forward
0 new messages