Cloudflare just introduced MASQUE to Warp this week. You can enable it with the command warp-cli tunnel protocol set MASQUE from terminal. If you ask Bard you will be provided the wrong command.

If you want to know why you should use MASQUE instead of wireguard I don’t have a good answer for you. From a user point of view I think it doesn’t really matter. MASQUE does have it’s advantages because it uses QUIC but for day to day I don’t think you will see any difference. I just turned it on for myself today.

I was using an intel NUC as a mini server for several years. It (or the power supply) died in T.S./hurricane Ernesto. I purchased a NUC 13 Pro and went about setting it up with Proxmox. At first I had issues installing anything. It turns out that my new Kexin USB sticks are garbage. That disappoints me because I use them for my camera sd cards.

After I got that figured out we went right back into a multi hour power outage. That stalled everything. Once I could get back to it I then proceeded to install Ubuntu 24.04 as my VM image. I decided against using the LXC containers. I read it is easier to backup a full image. We shall see. I have not setup a backup yet and I do have to play with LXC containers.

My old SSD still worked. That was good. I really only had a Unifi Controller, Unifi Video and rustdesk. The other stuff like TTS was just some fun learning. I know I did it the wrong way but I did eventually get everything moved over. Unifi Controller and Rust Desk where in docker containers. I had to copy a few directories over and then mess with them. Unifi Video was not in a docker image but it is now. The instructions all said to create tar files of the containers and then copy them over. I couldn’t get docker to see any of these containers on the SSD. It probably was user error but it does not matter now because I got it to work.

For a backup I haven’t settled on what exactly I want to use for the backups and where to keep them. I have to keep my onsite data to a minimum due to the potential of a hurricane wiping out the place.

For now I am happy. If the NUC would have died a natural death this is the path I wanted to take. I had an older version of Ubuntu on it. I had installed the desktop version (it was a failed attempt at some other testing and I never used it) and it was getting full of clutter of my other testing.

I could have gone with a cheaper mini computer and installed proxmox but I wanted to have that peace of mind of having a quality build. 32GB of RAM is also a lot more than the 8 GB I had previously.

Umidigi is a Chinese phone company that I have had great success with for several years. I would buy around a dozen of their phones each year on Aliexpress. Unfortunately I am now done buying their phones. I have had several with batteries that didn’t last a year and one who’s screen stopped working. The last one is the most important one because it was given to my retired father in law who was then left without a phone for a week and no way to contact his family who lives no where near him.

I don’t know what happened to them but it makes me feel like a fool. My last batch of 6 phones has exactly 1 still working without issues. I guess the era of affordable phones are over.

I started to move my 3CX customers to FusionPBX. I only have a few customers but the first one is complete. I had used fusion pbx before but it has been 4 years since I had it in production. I was really concerned that It was not going to go well. The problem I always seem to run into is that changes don’t always take until you reboot the server. Still, it took me several hours to work out the bugs. I found this morning that my time conditions were not exact. Technically that would mean that I am still working out the bugs.

When I mentioned earlier about changes not taking effect you can save a lot of time by looking at the XML. In another time condition issue the XML did not update when I changed the ring group extension. I have ran into this issue on a few occasions now.

Some of the previous issues that I encountered were phone template rated, opus not working ,gateways not registering and lastly plenty of lockouts from fail2ban. I have worked through all these issues and on top of this I even figured out how to use Bulkvs for caller ID lookups.

This is a small pbx with only 10. extensions. No one uses a softphone but if they did I would have them use Groundwire.

I was updating the firmware on a Yealink T46U when I noticed that Opus disappeared. The firmware I was upgrading to was 108.86.0.90. I upgraded another phone to check and that one also was missing Opus. I thought it was strange so I went and opened a ticket with Yealink. They provided me with the info below which did fix my problem.

At first, I forgot to replace the ‘x’ with the account number, but after I did that, it worked. Basically, you are uploading a text file that ends in ‘.cfg’. This text file has only one line, as shown below. When the instructions say to change the ‘x’ to the account number, they are asking which account on the phone should have the Opus codec enabled. Since I only have one account on my phone, this wasn’t an issue.

I am using FusionPBX (freeswitch) and I did have to change the Opus sample rate to WB(16khz) in order to get it to work. I was successful with a test call with the T46U and my softphone, groundwire.

Due to external requirement, Opus is now disabled by default in new firmware. If you need to use the Opus codec, please import the parameter or add to Provision template below:

account.x.codec.opus.enable = 1

Change the x to the account number. For example: 1,2,3…

If you don’t know how to import, please try to import cfg in attachment. Remember to edit it and then import at WebUI->Setting->Configuration-> Import Cfg.

Almost lost my sd card in my Raspberry PI. I assume it became corrupt from the latest power outage but I am not sure. I was able to log into it today and when I went to reboot it never came back up. I hooked up a monitor where it then told me there was some corruption.

I had a hard time from there. My sd card reader seems to be shot. I ordered a new one while brainstorming what to do. I thought I had a cloned sd card but that did not have the right OS on it. Eventually I remembered that I have a sd card reader in my NUC. That was perfect as that runs Ubuntu. I ran FSCK using that and voila I am back in business!

Its probably time for me to swap out the sd card anyways but it is the biggest gripe I have. I know there are work arounds but I wanted to keep this particular PI stock and run it off my POE switch.

I wanted to bring more visibility to this ridiculousness. This article on Techdirt isn’t very long. In fact, it’s not a new phenomenon because it’s been happening for a while. However, it is maddening that this is even allowed. Republicans from across state and federal government keep taking credit for sending money to their constituents. This is a good thing but it becomes a bad thing when they voted against these very same bills. It’s awful and most likely a lot of people don’t even realize it. Hopefully, the word spreads a little more with posts like these.

I am hoping to help draw attention to this excellent article about the failure of government to protect our privacy. I saw this article on Techdirt this morning. It’s actually written by the same person that would write articles for dslreports aka broadbandreports.com.

Basically, the problem lies in the US’s lack of privacy rules. Instead of banning tiktok they need to pass privacy rules that make it illegal for tiktok to use their customer’s data in nefarious ways. So many other companies are buying and selling your data and manipulating their users that it is literally affecting the entire country. It seems the government has already forgotten one of the most famous privacy incidents ever with Cambridge Analytica.

It’s a good read and I hope it gains some traction.

I wanted to screenshot these settings and share them for the next time I have to do this. Ofcourse, it was super easy to register after I got rid of the special characters in the password.

Here you will see that it is a standard gateway. The username and password are created in BulkVS under Interconnection -> Trunk Group Manage. The FROM USER field is the same as the username. Proxy will be sip.bulkvs.com:5060. Lastly, I might have it wrong but the retry seconds should be dropped to under 30. Step 2 on the Bulkvs website says this:

Step 2: Configure your system with the BulkVS SIP Registration credentials from step 1. Your system should register with the BulkVS SBCs every 25 seconds or less

FusionPBX does not like special characters in passwords and I wasted far too much time finding that out the hard way. To be fair I should have installed Wireshark and ran a capture but I wanted to be lazy.

Over the past two days I had been trying to register a Bulkvs trunk. I kept getting this message in the logs: Failed Registration with status Operation has no matching challenge [904]

Using Gemini or just regular google and searching for Freeswitch and fusionpbx led me to much of the same stuff and nothing was helpful. Eventually, I did find the answer in the fusionpbx forums.

WIth that out of the way I definitely will be taking screenshots on how to setup a Bulkvs SIP trunk. It wasn’t anything special after I fixed the password.