AZSlow

Testing is never boring, but if you prefer, we can meet online. I downloaded the file and am sending the log. The error window looks practically the same.

The error still occurs. I am sending the log.

Well, it is always fun to DIY something. I recommend reading X32 OSC remote protocol to understand details, the PDF is easy to find. Not sure if ChatGPT can answer such questions, but you can try. That pdf is long time available, so may be it is already learned. The command is /xremote which should be repeated periodically, then you should get notifications on changed. Meters are subscribed separately, since they constantly change. Getting complete status or just 100 parameters take significant time, so that is not an option for the purpose.

I am aware of this. Sure, it may not work, but maybe something will come of it. It seems to me that tracking in azaosc doesn't take into account the assumptions I made, so I'm trying to create something of my own.

AZAOSC has build-in tracker. So it is able to report changes. That was designed for X32 mixers with physical controls, not sure that is useful for rack mixers, till someone else is operating it and you want to know what is going on.

Note that ChatGPT pretend to be "smart", but in practice it is just able to compose information which someone has published in the Internet. For programmer which can easily distinguish between garbage and something which make sense, the tool can save time needed to find original information manually and adopt it for particular case. For non programmers, the result is more or less unpredictable. The result normally somehow works, but it can do something stupid. For small things the probability of good result is high, for complex things, especially if you try to do something no one has done before or done but has never published the algorithm, the probability to get reasonable code is zero.
That is one of biggest worries some people have about ChatGPT, it may be more stupid with time. Before AI, when someone had to code something, he was asking on some forum and answers was also on the forum. Such forums is the source for AI knowledge. Now people ask AI in this case, answers are obviously have no value for AI training. But periodically published in forums, even in case they are bad. So AI get the confirmation the answer is ok, even if not, and so learn more and more stupid solutions.