Scott1234

I will give OPUS an experiment I have not spent any time with it yet. Like I said the network is good, should be no packet loss, and PCAP's look clean. I usually find the issue is at the start at call setup the first 2 seconds then its fine. Both sides normally hear glitchy bad audio then goes good. I some times hold off saying hello for a second as I know its better that way or I get the comment 'this sounds like a bad line' ... lol. Maybe play more ring back locally to allow better establishment ? This also happens dialling into a attendant playing an audio book, so eliminating any carrier issue, 3 hops to pbx sub 5ms, zero loss. Maybe its a TLS thing

Any one know if the app has a jitter buffer like a desk phone? I don't recall the app being so bad, but I am trying to use it more and notice its very susceptible to getting audio glitches all most like its running with 0 buffer, and I cant find any obvious network problems. I don't get the same when using Bria Enterprise. What else could I look at to tune it more? Generally the PCAP's look clean but the experience heard (inbound) at the app says otherwise. It sounds like I should see bad PCAPS but I don't and feel its happening at the app level. My network is rock solid and wifi on or off still the same. Bria is fine, using the user portal is fine, just the mobile app. ¯\_(ツ)_/¯ I have a group setup that just plays MOH which is an an audio book and sit in there to hear the book, I do have external DID's on it as well for external testing but its on all levels. 1.22 (65) 68.0.30

I mean the logic of the setup. When set to user must hit pound to trigger its not obeying. it will still preform the search with out hitting pound, via the time out? Maybe its meant to be like so? I was just trying to disable people from using extension dialling in the Auto Attendant by making them need to hit pound but not advertise it.

FYI not sure if related to this, but the setting, Extension input - User Must Hit Pound It does nothing, it still operates the same as when the extension matches. Pound is not used anywhere else in the given Auto Attendant

Sharing is caring, This is a far better Yealink dial now pattern for dialplans.xml, 3 digits extensions 2-9 xx for AU. "^(+?61[0-9]{9})|^([2-9]{1}[0-9]{7})|^(0[2|3|7|8]{1}[0-9]{8})|^([04|05]{1}[0-9]{9})|^((1300|1800)[0-9]{6})|^(13[0-9]{4})|^(?:000|106|112)|^[2-9]\d{2}" Summary, <?xml version="1.0" encoding="UTF-8"?> <dialnow> <!-- Matches Australian + or 61 numbers with or without country code --> <Data DialNowRule="^(+?61[0-9]{9})" LineID="0"/> <!-- Matches Australian local area numbers without area code --> <Data DialNowRule="^([2-9]{1}[0-9]{7})" LineID="0"/> <!-- Matches Australian numbers with area code 02, 03, 07, or 08 --> <Data DialNowRule="^(0[2|3|7|8]{1}[0-9]{8})" LineID="0"/> <!-- Matches Australian mobile numbers starting with 04 or 05 (05 proposed) --> <Data DialNowRule="^([04|05]{1}[0-9]{9})" LineID="0"/> <!-- Matches Australian 1300 or 1800 numbers --> <Data DialNowRule="^((1300|1800)[0-9]{6})" LineID="0"/> <!-- Matches Australian 13 numbers --> <Data DialNowRule="^(13[0-9]{4})" LineID="0"/> <!-- Matches Australian special numbers emergency 000, 106, or 112 --> <Data DialNowRule="^(?:000|106|112)" LineID="0"/> <!-- Matches Australian special access numbers (dial before you dig etc. 11 or 12) --> <Data DialNowRule="^((11|12)[0-9]{2})" LineID="0"/> <!-- Matches 3 digits extensions 2 to 9 xx --> <Data DialNowRule="^[2-9]\d{2}" LineID="0"/> </dialnow>

Thanks, out of curiosity what NAT types does this make it more complex? I run non standard SIP ports to to prevent SIP ALG issues for customers who don't make sure their router implementation of SIP ALG is disabled, as I am over having to have customers make sure sip ALG is disabled, nothing but a nightmare. I notice MS Teams typically does not suffer this fait by using TLS. Only. and 5061 does not seem to be inspected like SIP 5060.

It means the web app needs a lot of work Honestly this is the future bread and butter, it needs to be A-Class. no matter what they tell you. I say desk phones are dead, yeah there might be the occasional desk phone deployment, but the base level of tech experience of users increases over time, the physical desk phone solution is out of date to these modern users. If your web/apps are not up to scratch, you won't make it, long term. Just saying. best get onboard now or lose your market position, hint hint nudge nudge wink wink... ... lets go V69.1 ? Edit, Not only does it need to work well, it also needs to look good, visually. no half baked solution's. I trust the team can pull it off. considering how light weight the pbx is running on any hardware, its in a good position.

Does anyone know if this is a bad idea or not? For now, focusing only on Yealink phones and only hosted environments. By default, STUN is disabled in the PBX base Yealink config file, and when a phone connects, you see the local LAN IP in the contact, along with the device address being the WAN IP. When wire sharking, you see the SBC attempts to send audio to the local LAN IP along with the WAN IP. With STUN appropriately enabled, the phone will register with the wan IP in both the contact and device address and shows a clean result as it's not wasting time & resources trying to send to a LAN IP that will never work. The only benefit of keeping the LAN IP is quickly finding a phone when remote supporting a customer you need to connect to. Is there any other reason? As it all appears to work including BLF's etc. Any unforeseen danger?

Great so I wasn't going crazy ha ha, thanks.

Happy days! thanks. We are scoring great now on .0.30

In this instance where I say they were treated the same, they were configured with different lengths, but were not respecting the hearing ring back vs waiting with hold music when tested. But specifically the Condition logic area. This setup does not work, pic included. When the caller is waiting because single agent is busy on call with music (tested with DND function, to busy the phone, maybe this is my downfall?) they do not get to wait 20 seconds before extra agents, it follows the 10 second of hearing ring back instead even though the caller is holding with music. ¯\_(ツ)_/¯

Hey ya's I think the All Agents Busy option from Redirection should also be able to be used as Condition logic. Is this something that could be done? I like trying to minimise the amount of call Q's needed so that reporting metrics stay in the one Q, but also get why you would have the separate Q and report on that on its own, but having the option would be nice. For example, four agents, all four are busy, I would have to exit the group via Redirection to another Q that has backup agents in it that all need to ring right away etc. Kind of like using Number of waiting calls exceeds but not having to wait for at least 1 call to trigger. Also Handy at this point would also be Individual controls for each Condition on how many agents can ring at once, and Algorithm logic. separate to the base Q setup. Also, it seems Waited Longer Than and After Hearing Ring Back, appear to be treated as the same, if I have both set on a group it seems to just default to Waited time even during ring back, this one caught me off guard today and I burnt a bunch of time messing around at a new install, trying to figure out if I was missing something. thoughts? and one else experienced it?

Interested to see where you go, I had this on my list next to see what I could build. Was also wondering if we could leverage the chartjs/graph.js that gets used to build some static graphs to embed into an e-mail, like a daily management summary info that you would typically find on a wall board. that's not the domain midnight report, don't really want to see all the calls summary just nice chart overview metrics. I had a look on GitHub the other week to see if any one had been publicly working on one and found this, vodia-dashboard-api/dashboard at master · guptahitesh1304/vodia-dashboard-api (github.com) 3 year's old but might help you get going, its on my list to look over.

Thanks ! I'll check it out when its out. Play back of some of the uploaded ring back files also didn't work either, maybe related to what you found.

Can't seem to delete custom Ring Melody's once they are uploaded though the "Upload Audio" (base level) I think its been doing this for a number of versions. Just now I want to actually change the slots they are in. Currently 68.0.28. It just ignores the delete operation. If you upload a ring back file you can remove that. I see it's storing the uploaded Ring melody's to domain.com/img/ringer2.wav but not related to pbxwebai dir. Mabye something getting confused when you are running with pbxwebai folder?

Nice I look forward to testing.

A test from 69 still has the same issue with the Message-ID included, with no other additions added to the header, so it must be the space after the : that is required. I am sure a quick one for you to test and check.

Ha ha, I asked chat GPT and it gave me some directions to go down, the sources it was referencing are no longer published. Q, What would a Misspaced header look like? A, Here's an example of a header with a misspaced header that would trigger the "Misspaced headers" rule in SpamAssassin: From : john.doe@example.com To : jane.doe@example.com Subject : Important message regarding your account Date : Mon, 3 Apr 2023 08:30:00 -0400 In this example, there are extra spaces between the header name and the colon separator in each header field. For instance, there is a space between "From" and the colon separator, which should be "From:". This would trigger the "Misspaced headers" rule in SpamAssassin, and add points to the overall spam score of the message. Q, What should a correctly formatted one look like ? A, Here's an example of a correctly formatted header in an email message: From: john.doe@example.com To: jane.doe@example.com Subject: Important message regarding your account Date: Mon, 3 Apr 2023 08:30:00 -0400 In a properly formatted header, each header field name (such as "From", "To", "Subject", and "Date") should be followed immediately by a colon separator, with no spaces or tabs in between. The field name and value should be separated by a single space. Headers in email messages typically follow the format shown above, and adherence to this standard is important for ensuring that email messages are properly interpreted and processed by email clients and servers. Q, What about the spacing after the : in the header? A, According to the RFC 5322 standard for Internet Message Format, the recommended format for email headers is as follows: field-name: field-value Here, "field-name" refers to the name of the header field, and "field-value" is the value associated with the header field. The colon separator should be followed by a single space. For example, a properly formatted header for an email might look like this: From: john.doe@example.com To: jane.doe@example.com Subject: Example Email Date: Tue, 4 Apr 2023 11:30:00 -0400 Note there is also a single space following the colon separator in each header field. It's worth noting that some email clients or servers may not strictly adhere to the recommended format in RFC 5322, and may allow for variations in header formatting. However, sticking to the recommended format can help ensure that emails are properly interpreted and processed by all email clients and servers. Here is the raw PBX, Content-Type:text/plain Subject:Voice: Test From:<phones@mydomain.com> To:<test-xxxx@srv1.mail-tester.com> Date:Mon, 03 Apr 2023 00:45:59 -0000 MIME-Version:1.0 Message-Id: <20230403004559.CA021E1B60@mydomain.io> This is a test email. Could it simply be the spacing after the : ? The message ID part is added by my server processing the request, formatted with the space. Here is one sent from outlook/o365 that scores 10/10. Date: Fri, 31 Mar 2023 08:52:26 +0000 From: ABC <my@domain.com> Reply-To: ABC <my@domain.com> To: test-123@srv1.mail-tester.com Message-ID: <0101018736de3a32-7dedff49-63aa-49d3-99c2-8ce5307dc9f1-000000@domain.com> Subject: TEST - 123 123 Mime-Version: 1.0 Content-Type: multipart/alternative; boundary="--==_mimepart_64269f465fd_4a1804e18610dc"; charset=UTF-8

yeah I was not 100% sure it was the order, but I also could not find any thing else odd looking to trigger it, unless that part has been stripped out by the time its processed by the remote side and re-rapped. https://www.mail-tester.com/ you can do a test your self there are many like this but I find this one is good. I try to aim for a score of at least 8+ There was as not a whole lot of insightful info on exactly what HDRS_MISSP meant or any examples I could find.

Yeah I mean in that section its setting the display name in the quotes part, but not sure if that would get stripped in your setup.

You can also just input the line under reg_email like so. "YOUR NAME HERE" <abc@abc.com>

I think its the placement of headers but I could be wrong, It should be, From: To: Subject: Date: Message-ID: MIME-Version: Content-Type: PBX is, Content-Type: Subject: From: To: Content-Transfer-Encoding:base64 Date: MIME-Version:

Hey, was doing some work recently on moving all mailing to AWS SES and whilst checking how the PBX shapes up I am all ways losing 2.5 points on spam assassin due to HDRS_MISSP "Misspaced headers" I only get this on the pbx. Can you guys look at the header structure and make it complaint ? just so we don't all lose 2.5 points out of the gate, also some of the base e-mail templates are missing alt text's. <img src="https://portal.vodia.com/img/googleplay.png" width="135px" height="40px" /> <img src="https://portal.vodia.com/img/appstore.png" width="135px" height="40px" /> <img src="https://portal.vodia.com/img/msstore.png" width="135px" height="40px" /> <img style="border-top-style: none; border-right-style: none; border-left-style: none; border-bottom-style: none; max-width: 200px" src="data:image/png;base64,

Just came across this with a customer with power outage who had recently added the app usage as an option. I normally use the "No agent has phone registered" logic to control what to do in a office power outage, where in the past this would divert off to a human answering service for this customer. I assumed as they all had apps registered it would not follow that logic, but as the app is not treated as an active registration when backgrounding it treats them as not registered, even though technically its capable of pushing though to the app. Is there any options I am missing to allow for this? for now I have just said to use "all primary agents are logged out"

Maybe we can use these new AI chatGPT type tools to build a UI Galileo AI · Copilot for interface design (usegalileo.ai)