- freeswitch-dev@lists.freeswitch.org - FreeSWITCH Mailing Lists

[Freeswitch-dev] Any prompts y’all want Callie to record?
by Chad Phillips 07 Dec '17

07 Dec '17

My company is scheduling some time with Callie from GM Voices to record a few voiceover pieces. Since the setup cost is always the most expensive part of the whole deal, I always like to reach out to the FreeSWITCH community and see if there are any new phrases we’d like to add to the sound files. Anybody have ideas? As long as it’s not too crazy many words, I’ll just eat the cost :) Let me know by Friday December 1st — we’ll probably submit everything the following Monday. Chad

2 2

[Freeswitch-dev] Fwd: FreeSWITCH Video Conferencing Blackout issue
by Richard Hsu 06 Nov '17

06 Nov '17

Hi all, [Issue] I am experimenting with the FreeSWITCH version 1.6.2 video calling capability. In particular, we are seeing an issue where sometimes the video element in HTML will start streaming videos just fine, but most of the time, one side of the video call, or both, will have a screen being pitch black. The audio is totally fine though. [Self-diagnosis] Looking into chrome://webrtc-internal, it indicates that there is no frame being sent, and looking into the FreeSWITCH log, it appears this line: "[ERR] switch_rtp.c:8098 Error: video SRTP protection failed with code 10". is always present. I looked into the source code, which corresponds to the switch_rtp.c file, it seems like the "srtp_protect()" function, which transforms/encrypts an RTP packet into SRTP packet, returns an error code. I tried by patching it with the changes here: https://freeswitch.org/jira/br owse/FS-10667, since it seemed like a similar issue, but it didn't work either. I understand there are several ways to go about this: 1) stick a debugger or 2) try out the latest version. [Question] Without other major changes, what else would you recommend me to look at? More importantly, has anyone seen this strange behavior of FreeSWITCH before? I appreciate your effort with FreeSWITCH so far and would love to hear your thoughts! Thanks, Richard -- Richard Hsu Massachusetts Institute of Technology Department of Electrical Engineering and Computer Science Class of 2016

2 1

[Freeswitch-dev] ciphers on verto and profile wss
by alexis 31 Oct '17

31 Oct '17

Hello, im facing a problem with webrtc and i would like to share the situation and see how wrong i am :) As you know, google (well, not only google) (mostly of our webrtc users uses chrome) is pushing about to move to ECC signed X509 certificates, ECC keys are smaller that RSA, and it benefits in the calculation of encryption, less data in the cable, etc etc etc. Not the point of this mail actually. I have an X509 cert signed with an ECC key (working on nginx, tomcat, etc). Now i want to use this same cert in freeswitch, i dont use mod_verto (but im trying with it too in this problem). usually, enabling wss on SIP internal profile is enough for us, we have a phone developed by us (using sipjs) that is enough for our needs and works perfect. Thing is, if i build a wss.pem tls.pem dtls-srtp.pem with this cert (the ecc signed one) websocket does not work. and here's the the detail of that . websocket starts? yes (profile, verto), all start, ports are up, ssl is up, you can connect with openssl s_client and it works . what's the problem then? that wss on sip profile and/or wss on verto does not accept any ECDHE-ECDSA cipher at all (chrome, firefox receives a server handshake failure right after the client hello) . Could be the cert wrong? yes, but if i enable sip-tls on the internal profile, i'm able to connect to port 5061 and ECDHE-ECDSA is accepted (sip tls works perfect with this same certificate) Im not a good C programmer (i work with java, javascript and python), but i've been working with mod_verto/mod_verto.c and mod_verto/ws.c trying to load all ciphers in the ssl_init methods and whenever was at my reach without success. FS version is 1.6.19 from source, openssl is 1.0.2g, by now i'm testing on ubuntu (production is debian 8.6). if you can lead me where to check or change or any clue to get it working will be extremely appreciated. thanks in advance, best regards

2 1

[Freeswitch-dev] Delay Bye Message
by Alexander Perkins 10 Oct '17

10 Oct '17

Hi. I have what appears to be an unusual request. One of my clients has a clause in his contract with the local telco/sip provider that requires the majority of calls to be at or greater than 12 seconds. On a former platform, he has an option called 'Delayed Bye Message.' This option, if checked, allows him to specify how many seconds to delay hanging up the call once leg B initiates the request. So, all calls will last at least 12 seconds. Does anyone know how to do that with Freeswitch? Any help is greatly appreciated. Thank you, Alex Perkins

4 3

[Freeswitch-dev] "rtp-timer-name=none" resolve bug FS-10522
by Alexander Haugg 06 Oct '17

06 Oct '17

Hi, if i have the „rtp-timer-name“ set to „soft“, the rtp stream have some bugs (see https://freeswitch.org/jira/browse/FS-10522 ). The problem exists only, for webrtc calls client --> freeswitch. Now i try the solution, to set the value to „none“ and it works fine. What for problems could arise with this configuration? Thanks a lot.

1 0

Re: [Freeswitch-dev] [Freeswitch-docs] TR: [ mod_voicemail ] uuid parameter
by Giovanni Maruzzelli 30 Sep '17

30 Sep '17

Hello Enrico, I would write a simple lua script that will read the database, and then play the desired message to caller. Then, transfer to voicemail application. Best wishes for your thesis, let us (or me) know how it goes. -giovanni sent from mobile cell: +39 347 266 56 18 Giovanni Maruzzelli OpenTelecom.IT On Sep 28, 2017 23:28, "jungle Boogie" <jungleboogie0(a)gmail.com> wrote: On 26 May 2017 at 01:37, Enrico Vergata <enrico.vergata(a)sewan.fr> wrote: > Hello community, > > PS: Please find attached the mod_voicemail.c which includes my modifications > if this can help you to better understand what I mean. > You should make a pull request with these changes. > > > > Enrico Vergata > > > > > > > > > _______________________________________________ > Freeswitch-docs mailing list > Freeswitch-docs(a)lists.freeswitch.org > http://lists.freeswitch.org/mailman/listinfo/freeswitch-docs > -- ------- inum: 883510009027723 sip: jungleboogie(a)sip2sip.info _______________________________________________ Freeswitch-docs mailing list Freeswitch-docs(a)lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-docs

1 0

[Freeswitch-dev] HTML5 Verto speed test computation question
by Philippe Sultan 09 Sep '17

09 Sep '17

Hi, The bandwidth computation in the HTML5 JavaScript client in file html5/verto/js/src/jquery.jsonrpcclient.js, is : var up_kps = (((this.speedBytes * 8) / (this.up_dur / 1000)) / 1024).toFixed(0); var down_kps = (((this.speedBytes * 8) / (this.down_dur / 1000)) / 1024).toFixed(0); this.speedBytes being the total number of transferred bytes, converted to bits after *8, I was wondering if the total shouldn't be divided by 1000 rather than 1024 in order to get a a result in kbits/s, that is : var up_kps = (((this.speedBytes * 8) / (this.up_dur / 1000)) / 1000).toFixed(0); var down_kps = (((this.speedBytes * 8) / (this.down_dur / 1000)) / 1000).toFixed(0); Am I missing something here ? Thanks ! Philippe

2 1

[Freeswitch-dev] Compiling mod_shout and mod_amqp
by Rory Toma 07 Sep '17

07 Sep '17

I can't figure out how to get mod_shout and mod_amqp to compile. I've added them to the spec file, have even taken out the conditionals so building them is on by default. They don't end up in modules.conf (I can't figure out how this is generated, either). The 2 or 3 places where this looks like it is being generated aren't showing my changes. The funny thing is that a correct Makefile is generated, and if I cd there and type "make" it builds correctly. But the output rpm isn't being made. What am I missing here? This is on a CentOS-7.3 system. What is the normal way to enable building these?

1 1

[Freeswitch-dev] ENUM+EMER=ENUMER
by Oleg Khovayko 04 Aug '17

04 Aug '17

Dear Colleagues, 0. As we know, there is no exist ENUM service for private person or small PBX owners. e164.org is dead. And before death, it applied a "dirty tricks" to requests - for instance, ignored toll-free record in their ENUM database, and routed TF-calls to their specific terminator. We think, any centralized ENUM system vas these risks: - Non-transparent routing rules, controlled by some centralized organization - Risk of ban or mis-route some specific call, according "wishes" of ENUM-operator - Risk of business termination of ENUM operator. 1. We, Emercoin Team ( https://emercoin.com/team ), are designed decentralized blockchain-based ENUM. General idea: ENUM-records are stored in the cryptocurrency blockchain, and by this way - records are transparent, available everywhere, any anyone can run his own ENUM-server, which process these records. Phone owner is owner of his ENUM record. So, system is decenrtralized, and does not have any centralized control point. To preserve malicious ENUM-allocation, there is proposed verifiers - special agents, who verify number owner, and provide signature to the owner of ENUM record. 2. System is already works! For instance, see logs from my own FreeSWITCH and Emer node: a) Call to TollFree with FreeSWITCH: fc66e67d-1d79-e711-ad65-ac1f6b02117c Dialplan: sofia/internal/1001(a)olegh.ftp.sh Regex (PASS) [Lookup ENUM] destination_number(18668886080) =~ /^(\d{7,})$/ break=on-true fc66e67d-1d79-e711-ad65-ac1f6b02117c Dialplan: sofia/internal/1001(a)olegh.ftp.sh Action ring_ready() fc66e67d-1d79-e711-ad65-ac1f6b02117c Dialplan: sofia/internal/1001(a)olegh.ftp.sh Action set(hangup_after_bridge=true) fc66e67d-1d79-e711-ad65-ac1f6b02117c Dialplan: sofia/internal/1001(a)olegh.ftp.sh Action set(continue_on_fail=true) fc66e67d-1d79-e711-ad65-ac1f6b02117c Dialplan: sofia/internal/1001(a)olegh.ftp.sh Action enum(+18668886080 enum) fc66e67d-1d79-e711-ad65-ac1f6b02117c Dialplan: sofia/internal/1001(a)olegh.ftp.sh Action bridge(${enum_auto_route}) . . . 57e3c17e-1d79-e711-ad65-ac1f6b02117c 2017-08-04 10:02:04.648767 [DEBUG] switch_core_state_machine.c:40 sofia/external/%2B18668886080(a)tollfree.alcazarnetworks.com Standard INIT b) Debug log from the local Emer node: 2017-08-04 14:02:04 EmcDns::HandlePacket: Handle packet_len=55 2017-08-04 14:02:04 EmcDns::HandlePacket: msgID : 41179 2017-08-04 14:02:04 EmcDns::HandlePacket: Bits : 0100 2017-08-04 14:02:04 EmcDns::HandlePacket: QDCount: 1 2017-08-04 14:02:04 EmcDns::HandlePacket: ANCount: 0 2017-08-04 14:02:04 EmcDns::HandlePacket: NSCount: 0 2017-08-04 14:02:04 EmcDns::HandlePacket: ARCount: 1 2017-08-04 14:02:04 EmcDns::HandlePacket: qno=0 m_hdr->QDCount=1 2017-08-04 14:02:04 EmcDns::HandleQuery: Translated domain name: [0.8.0.6.8.8.8.6.6.8.1.enum]; DomainsQty=12 2017-08-04 14:02:04 EmcDns::HandleQuery: Key=0.8.0.6.8.8.8.6.6.8.1.enum QType=23 QClass=1 2017-08-04 14:02:04 EmcDns::HandleQuery: After TLD-suffix cut: [0.8.0.6.8.8.8.6.6.8.1.enum] 2017-08-04 14:02:04 EmcDns::SpfunENUM: Domain=[0.8.0.6.8.8.8.6.6.8.1.enum] N=12 TLD=[enum] Len=64 2017-08-04 14:02:04 EmcDns::SpfunENUM: ITU-T num=[18668886080] 2017-08-04 14:02:04 EmcDns::SpfunENUM Search(enum:18668886080:0) 2017-08-04 14:02:04 EmcDns::SpfunENUM TF-match N=[18668886080] RE=[^1(800|888|877|866|855|844)[2-9]\d{6}$] -> 1 2017-08-04 14:02:04 EmcDns::HandleE2U: Parsed: 100 10 E2U+sip !^(.*)$!sip:\1@tollfree.alcazarnetworks.com! As you see, FreeSWITCH sends request to the local domain "enum", and node provided rule to route to @tollfree.alcazarnetworks.com, and FS followed this rule. c) You can test system with our public gateway "enumer.org", for example: $ dig -t naptr +short 53995390081.enum.enumer.org 3. What we looking for a) We looking for collaboration to create and maintain verifier WEB-site and calling engine. By our understanding, this site must to do: - Show form, when someone pays for verification (to preserve malicious calls to innocents). Payment can be collected in EMCs. - Make verification call, and speak 6-digit number. - User enter these numbers into the form, and get signature to deploy into his ENUM-record. b) Customers. We would like invite customers to deploy their ENUM-records into ENUMER. ---- If you would like to collaborate, please contact me. Thanks, Oleg

1 0

[Freeswitch-dev] mod_sofia hold/unhold issue correction problems
by Dmitriy Borisov 27 Jul '17

27 Jul '17

Hello! There are some issue (or not???) in mod_sofia module. You can reproduce this issue by holding call via 0.0.0.0 SDP: v=0 o=OneXS 99765105 2 IN IP4 87.237.24.14 s=- c=IN IP4 0.0.0.0 t=0 0 m=audio 49760 RTP/AVP 8 101 a=fmtp:101 0-15 a=rtpmap:101 telephone-event/8000 a=inactive and unholding via invite without SDP. In version 7a77e0b (branch v1.6) place of errorneus (as I think) code is near: 7641 if (switch_channel_var_true(channel, "sip_unhold_nosdp")) { 7642 switch_core_media_gen_local_sdp(session, SDP_TYPE_RESPONSE, NULL, 0, "sendrecv", 7643 zstr(tech_pvt->mparams.local_sdp_str) || !switch_channel_test_flag(channel, CF_PROXY_MODE)); 7644 } else { 7645 switch_core_media_gen_local_sdp(session, SDP_TYPE_RESPONSE, NULL, 0, NULL, 7646 zstr(tech_pvt->mparams.local_sdp_str) || !switch_channel_test_flag(channel, CF_PROXY_MODE)); 7647 } 7648 7649 if (zstr(tech_pvt->mparams.local_sdp_str)) { 7650 switch_log_printf(SWITCH_CHANNEL_SESSION_LOG(session), SWITCH_LOG_WARNING, "Cannot find a SDP\n"); 7651 switch_channel_hangup(channel, SWITCH_CAUSE_DESTINATION_OUT_OF_ORDER); 7652 } else { 7653 if (sofia_use_soa(tech_pvt)) { 7654 *switch_log_printf(SWITCH_CHANNEL_SESSION_LOG(session), SWITCH_LOG_WARNING, tech_pvt->mparams.local_sdp_str); // This is my add* 7655 nua_respond(tech_pvt->nh, SIP_200_OK, 7656 SIPTAG_CONTACT_STR(tech_pvt->reply_contact), 7657 SOATAG_USER_SDP_STR(tech_pvt->mparams.local_sdp_str), 7658 SOATAG_REUSE_REJECTED(1), 7659 SOATAG_AUDIO_AUX("cn telephone-event"), 7660 TAG_IF(sofia_test_pflag(profile, PFLAG_DISABLE_100REL), NUTAG_INCLUDE_EXTRA_SDP(1)), TAG_END()); 7661 } else { 7662 nua_respond(tech_pvt->nh, SIP_200_OK, 7663 NUTAG_MEDIA_ENABLE(0), 7664 SIPTAG_CONTACT_STR(tech_pvt->reply_contact), 7665 SIPTAG_CONTENT_TYPE_STR("application/sdp"), SIPTAG_PAYLOAD_STR(tech_pvt->mparams.local_sdp_str), TAG_END()); 7666 } 7667 } In described situation code flow go throw line 7654 and post to nua_respond correct SDP (with correct c=IN IP4 re.al.ip.addr and a=sendrecv): 2017-07-24 11:00:54.252668 [WARNING] sofia.c:7654 v=0 o=Sipean 1500875192 1500875195 IN IP4 x.x.x.x: s=Sipean c=IN IP4 x.x.x.x t=0 0 m=audio 11642 RTP/AVP 8 101 a=rtpmap:8 PCMA/8000 a=rtpmap:101 telephone-event/8000 a=fmtp:101 0-16 a=ptime:20 a=sendrecv sofia library sends SDP: v=0 o=Sipean 1500875192 1500875195 IN IP4 x.x.x.x s=Sipean c=IN IP4 x.x.x.x t=0 0 m=audio 11642 RTP/AVP 8 101 a=rtpmap:8 PCMA/8000 a=rtpmap:101 telephone-event/8000 a=fmtp:101 0-16 a=ptime:20 On SIP client side call in unholded state, but at FreeSWITCH side it is still in HOLDED state I have experimented with nua_respond, but maximum result is a correct call state on SIP client side (holded), but not unholding the line. Where should I search for errors? Or this scenario can't work correct?

2 3