Thanks so much!
(especially that the linked issue was introduced by my commit...)

• Maybe it would be good to add the documentation for the new method in the help...?
• Do you think this should be cherry-picked for 3.12 or is it fine to stay on develop?

I'm still usure whether connectToServerAddr is the best name for the method – it seems misleading when taken out of context. Whe only connect if necessary (for tcp), otherwise we directly call the onComplete message.

In fact, I would have thought it good to pack the startAliveThread into it, but the method boot passes this only when the argument startAliveThread = true (sorry, this is all a bit complicated to explain).

In fact, I would have thought it good to pack the startAliveThread into it, but the method boot passes this only when the argument startAliveThread = true (sorry, this is all a bit complicated to explain).

Yeah... the whole Server class and boot behavior was mean to be rewritten, but it seems it never happened. Not sure how much we want to change before the (hopefully) rewrite later on, and how much we should resort to minimal patching.

@adcxyz did a rather large rewrite long ago, but as far as I know, it somehow was never taken up and merged.

@adcxyz did a rather large rewrite long ago, but as far as I know, it somehow was never taken up and merged.

Yes, that was discussed here : #3310
and the related fork that sketched this out: https://github.com/adcxyz/supercollider/tree/server_ServerProcess

Ah good, thanks for pointing to this. As you wrote, the branches have diverged. What do you think: is the current PR doing harm (I am always ready to minimize impact at the cost of beauty :))

what do you think?