I've been using Roon since 2017 and I have a lot of experience with it. I've built maybe 10 different Roon servers/cores, possibly more. Some Windows 10/11, some Ubuntu. I don't know MacOS well enough to try my hand at that.
In the end, what I've done is build 2 cores and I can switch between them when one is running the process that effectively kills Roon.
I certainly understand that many people don't have the same issues. My take on this, after a ton of experience and communications with other users having similar problems, is that it's a metadata/database update process. It's not network or hardware configuration.
Why does it only affect some people? I suspect it has to do with library composition - use of Roon's tags, and unidentified albums could be culprits.
The main issue with this instability is that Roon uses only one core for this process and it maxes out the core and makes Roon slow to unresponsive and it then ultimately restarts itself, which of course interrupts the music. That restart doesn't reset Roon to functional - the problem continues until the process is done.
Anyway, that's what I was referring to.
In terms of the forum, I've had posts edited to leave positive comments and remove negative...and lately just deleting negative posts.
This is not to say I don't love the product. Well, it's a love-frustration thing anyway. But I haven't found anything better. I just wish Roon would code away from this architecture that does this.