Vodia PBX

Why are you using port 5061? I see that port 5060 should be used. ibm# host -t NAPTR sipgate.co.uk sipgate.co.uk has no NAPTR record ibm# host -t SRV _sips._tcp.sipgate.co.uk Host _sips._tcp.sipgate.co.uk not found: 3(NXDOMAIN) ibm# host -t SRV _sip._tcp.sipgate.co.uk Host _sip._tcp.sipgate.co.uk not found: 3(NXDOMAIN) ibm# host -t SRV _sip._udp.sipgate.co.uk _sip._udp.sipgate.co.uk has SRV record 0 0 5060 sipgate.co.uk. ibm# host sipgate.co.uk sipgate.co.uk has address 217.10.79.23 sipgate.co.uk mail is handled by 10 mail-in.netzquadrat.de.

Is there a specific reason why you are using the 1.5 version? Email has improved a lot in the 2.0 version. 2.1 will even have a spooling feature.

Oh maybe you have a race condition. If they both attempt to redirect after 10 seconds, sometimes the phones "win" and sometimes the PBX "wins"...

Well, that tell you that the PBX was able to determine the destination, but there was no valid response coming back... Check out the logging: http://wiki.pbxnsip.com/index.php/Log_Setup. There you can see how to turn on the logging and see what SIP messages the PBX sends and receives.

Okay okay...

Well, extensions starting with '1' are generally a problem, as it makes it more difficult to come up with a dial plan that uses the 1 to indicate 10-digit NANPA numbers. In general, we recommend to use extensions in the [4-7]xx range, so that you can have 8xx for mailbox access and possibly 9xxxx for external calls. The numbers 0-3 are useful for the auto attendant direct destinations. Your behavior is quite strange - maybe a license problem?

Well, if you use the email feature, you can e.g. send the voicemail to a mailing list. What about seeing it from the user's perspective? I think the application here is that someone goes on a vacation or gets sick and wants a smart way of having someone else taking over. In an email server (e.g. Outlook), you can set the "Absence Message" and then the email server will then act accordingly. If we redesign that subsystem, lets not just copy from existing legacy systems, lets take it one step further...

Well, the "tone" language is for the ringback tone and the busy tone. This might differ from the "spoken" language, e.g. when you use US English for voice and UK for ringback. That's why...

Hmm good point. I don't even remember using DTMF after realizing it has voice recognition. DTMF should be a piece of cake compared to voice recognition...

Probably the DNS address can not be resolved. Keep in mind that SIP also used NAPTR and SRV records. On log level 8 you can see how the PBX step-by-step resolves addresses.

There is a new version coming out that will address this problem and emulate a key system much better than before.

Do they have different domains? Maybe they have an IP address conflict? I sounds weired indeed.

Can you make an example? I guess the call to the paging account will just trigger the page, no matter how long the page text is?

Okay, there is a SuSE10 version available at http://www.pbxnsip.com/download/pbxctrl-suse10-1.5.2.11.

Well, you need to perform the following steps: Translate the prompts Record the voice prompts Edit the prompts For the translation, you'll probably need the original texts in English (contact sales at pbxnsip.com offline for this). The recording should be done in a "good" recording studio, and the editing must use exactly the names that the PBX also uses for other languages. If you want to translate also the web pages, be prepared to several thousand lines of text segments that need to be translated. Email is not so much. In any case, you need a UTF-8 compinant editor that accepts XML.

Yes, that is the way it was supposed to work. Essentially, the method is used to reach gateways that are behind NAT - because these devices must REGISTER in order to be accessible from the PBX.

It is really weired. I assume you have only one domain called "localhost"? The trunk sends the call directly to a known extension? There must be something stupid. We'll keep an eye on this.

Push2Talk does not change the behavior on the PBX, just on the phone. It is just more "natural" to hold down the "talk" button while making a page annoucement. Version 7 is available already. But I think it is difficult to get back to a 6.x version (probably tftp update), so you may do this on one or two phones and see if you like it. If you have to manually downgrade those phones, it is not too much work.

At the moment I would not know how to set this up. But it is a good feature request. We will put this into the ticket system, so that it does not get forgotten.

Well, you should use a 2-letter code (like us fr se) to identify the language. The language must be also lised in the dict so that it can be shown on the web interface pull down menu. If you tell us what letter code you want to see we can put it into the next release. And of course it would be great if we can include the files for distribution... <?xml version="1.0" encoding="utf-8"?> <language name="en"> <file> <item id="en">English</item> <item id="de">German</item> <item id="sp">Spanish</item> <item id="uk">English (UK)</item> <item id="fr">French</item> <item id="ru">Russian</item> <item id="it">Italian</item> <item id="cn">Chinese</item> <item id="dk">Danish</item> <item id="jp">Japanese</item> <item id="se">Swedish</item> ... and so on. See http://wiki.pbxnsip.com/index.php/Localization for more information.

Okay, we were able to reproduce the problem. The problem was actually a stupid problem in determining what SRTP patch level we should use, for some strange reason were were checking if the version was bigger then 6.10, not 6.5.10! It would be great if you can check the version at http://www.pbxnsip.com/download/pbxctrl-1.5.2.10a.exe.

Are the numbers exactly the same? Maybe there is a difference with the leading "1" in the beginning. If you are usiong the "Default PnP Dialplan Scheme" in the domain, the PBX might help you by normalizing the numbers before comparing them.

You need to change the volume on the source. For the WAV files: you need to edit those files with a audio editor. Should not be a big thing, this way we save the CPU time for scaling the signal.

That is a very good idea! I think this way, we can keep a lot of messages and even after a restart the PBX will be able to deliver emails.

Well, the PBX includes a line parameter to identify which trunk it is. If the service provider (or where ever the PBX registers) is RFC compliant, that guarantees a correct trunk identification. Gateway trunks are a little bit more tricky, they must be identified by the IP address - the PBX uses the outbound proxy for this. If the from header contains a known user name, the PBX will "override" the trunk identification and then assume the call comes from that user (after authentication).