Hello Kurt.
I have thumbed this out on my mobile device so I apologize for any errors, some part of it was done as voice to text, so let's hope Google does the job correctly.
German is already supported in a default installation. However, the translation was entirely done by AI, so I would love to hear your feedback and any changes you'd like to submit.
I have a lot of ideas for the communications interfaces for voice within tickets CAD, but in an effort to get the software released, I've decided to table those until a future version.
Most likely before the end of the year, but it's at least 60 days away. The earliest a beta would be seen would be 45 days.
If you look within the source code,(when it's released), You can see some of the artifacts of the architecture I've been working on.
The voice communications platform itself has become almost as large if not larger than the entire tickets CAD system.
I really like the audio interfaces from proper physical CAD systems I worked with in the '90s.
I would like to have the ability to set the volume level for select and unselect audio paths, I want the ability to transmit simulcast on multiple modes by selecting more than one PTT to be active. I would like to support physical inputs for the PTT button, and still allow on-screen selection of individual channels for PTT.
I would like to have the ability to create a patch bus, and be able to assign different channels to either input or output or both to a particular patch bus. There are incredible complexities to this architecture. Echo cancellation and audio delay are a couple examples. In incoming stream, arriving at near real time, will need to have a delay inserted before connecting to a digital mode which may need a moment for an accept tone or the network to come up. This audio would need to be buffered, and in the event that you receive a bonk, a wait, or a reject, the system would need to know what to do with that audio.
I have given serious consideration to creating a competitor to the Zello service as an add-on to the platform. Most likely the protocol used underlying the mobile app would be mumble-based.
For analog radio interfaces, I had designed some physical hardware which I hoped would be inexpensive and I could sell for a reasonable price. But through consultation with my AI teams, we've decided that it would be best to connect with existing hardware. To keep costs reasonable for amateur radio and volunteer use. Use, I've looked at the All-Star Network. This is the one that is based on The Asterisk voip platform. I do need to do some additional research and possibly support other roip platforms due to commercial restrictions.
I have no experience with Tetra but would sure love to get my hands on the infrastructure and radios to give it a try.
The platform currently supports Zello for work and the Zello personal platform. I know Zello for work offers radio interfaces, this might actually work for you today, You can create a free work demo account, but it has a limited lifespan before you need to subscribe to their plans. Here in the US I believe it's about $8 per subscriber. I don't know the rate for the ROIP interface but theoretically you may be able to get by with just two accounts. One for Tickets CAD and one for the radio interface.
Additional interface ideas for the communications platform include SIP for IP telephony, rstp for video feeds which can include audio, I had again looked into designing custom hardware which would be a intercom station, with a controllable output. The concept would be things like gate access, and having the ability to patch these into other modes based on availability or escalation was an idea in the architecture. Some examples might be a gate into a facility, where you want to ring. Operator or guard station but if they are way or fail to reply within a time frame, you could ring digital devices or ring analog devices and support entering a code to remotely release the output thereby unlocking the door.
There are other iot devices which are common in the US. I don't know if you have them there, for things like call for help buttons. In our large warehouse stores or department stores, you may find a button that you can press on the wall for assistance and this rings. Rings department manager who, if they do not answer immediately, would create an overhead page. You might hear something like assistance needed in the paint department on the overhead page.
Of course I have a lot of ideas here, so as you can imagine, this blossomed into something gigantic and impractical to include in this release now.
As for mapping, the current best practice is to use an online map but enable caching. The caching service is built into tickets CAD right now. What this changes is that all clients of the web app point their queries directly to Tickets, CAD, and Tickets. CAD will serve a map tile from its cache, but if the cache is not available it fetches from the map source. This provides for high availability to the maps provided that you have loaded that map tile in the previous 90 days. Many map providers have been dealing with apps designed to siphon down all of their maps for offline use, and they take a heavy server load when people are downloading every map they think they might ever need, so all of them pretty much prohibit this practice now.
You can define your own tile sources. And there are instructions on how you can stand up a tile server. I'm not sure if you're familiar with the term, but generally online map providers are rendering tiles, and the Web UI stitches the tiles together for your view of a map. So the term map server and tile server are nearly synonymous.
I have not used it yet, but a beta tester suggested there be an ability to stretch an image over the top of a map as a layer.
This has been reported to be working and allows for things like taking the map of a fairground or a parade route and pinning it on top of a map. There are probably ways that you could also utilize this to be the actual map if you prefer.
I hope to have tickets CAD exposed either as a public release, or a public beta within the coming week. We're at a code freeze now, and working on cleaning up some prior commits which contains sensitive or confidential information. Most likely we will create a clean repository and post the necessary files without the previous history. The holdup at the moment is trying to decide how future development would go forward if we have two repositories.
I'll throw out here one more concern of mine. As the new tickets CAD is drawing attention, some of that attention is coming from commercial providers who feel threatened by the quality of what's being built here.. I have already experienced a couple of emails regarding possible patent infringement. There appear to be a lot of patents in this space.. I don't believe that all of the patents that have been issued are well deserved. The application of machine learning or AI or LLMs into a particular use seems to have received a patent. As I think about this, it's sort of like a knife and fork, there's not really a valid patent on those two tools, we all use them frequently to stuff food in our mouth. But imagine if somebody stuck a tax or license fee or has a patent on eating a chicken using a knife and fork.
With all that said, there may be capabilities in tickets CAD that are going to get stripped out of it, or never delivered, because of these patents.
Machine learning, statistical models, LLMs, they are a prior art and were designed for analyzing data. How does everyone feel about patents being issued because of the data that these tools are being used on?
I'm picking on just one use case here, this is not the only concern.
Push to talk on a mobile device, caching a map, there are others I won't list out here, but just wanted to mention that there are threats to open source software all around. Maybe the vendors will be excited about the fear this creates and can use that to sell you a very expensive product.