These things are undoubtedly cool. Pisonic is great, just vanilla supercollider with the JIT things like Pdef are really cool, Orca, as others have mentioned, is simply beautiful and work unto itself.
But with all these "live" programmable music things, I fail to find anything lasting in what they produce as works of music/art, I much prefer to see the software itself, and its potentialities, as the "work" that others continually express with them.
I love making music with computers, but there is so much infrastructure around things needing to be "realtime" and responsive, minimal latency, a million knobs on your controller, abstract grid controllers, high precision encoders (basically anything Monome has done), tons of ways to route and reroute data, all for the purpose of "performing" your toys. All so many things that are not really necessary to create something. Making digital music is so much more rewarding when your are not trying to plan a set to perform them. Thinking about the kinds of things computers can do ahead of time, and not just "just in time," opens up so much more creativity and ways of thinking about making music. And very importantly (for me), you can save your wallet, and not have to get lost in the dark money pit of buying equipment, or wanting to buy equipment, all the time.
These things definitely have their place and I think they are themselves beautiful, but there is a tension in them that is not really necessary.
If you want to hear some cool things you can make with just some extreme talent, imagination, and MaxMSP, without this emphasis put on being "performable" and realtime, look to Carl Stone [1]. He is just walking circles around everyone else, and really really showing what computers can do in the interest of art, rather than in the interest of being some techno-nerd rockstar.
Hi, if you're interested in Overtone, you might be interested in my projects as well, Scheme for Max and Scheme for Pure Data. They use s7 Scheme, a very clojurish Scheme implementation designed for computer music needs by Bill Schottstaedt at CCRMA, of Common Lisp Music fame. Part of my motivation for creating it was to overcome some of what I perceived as limitations in options such as Overtone and Pink. https://github.com/iainctduncan/scheme-for-max
Unfortunately Overtone is not really active anymore, the its replacement loses the lisp!
Other interesting Scheme/Lisp based systems in similar areas are Extempore (formerly impromptu), Nyquist (by Dannenburg, another godfather of the field), and Common Music.
There seem to be a lot of "music programming language" posts all of a sudden on HackerNews, many of which, like this one, have existed for years. Not that it's a problem, but I wonder why the sudden interest.
Wow, glicol looks amazing! One of the few music languages I’ve seen which actually tries to balance low-level synthesis with higher-level sequencing. (The only other one I know of is extempore: https://extemporelang.github.io/)
I use Extempore[0], but I have played with Haskell-based Euterpea[1] too. I bought the book by Paul Hudak and Donya Quick, "The Haskell School of Music". Common Music has Grace, an all-in-one, Lisp-based, cross-platform GUI [2]. They all have signal and note level music capabilities. That means you can synthesize sounds from scratch and also code at the higher level with notes and scores. Sonic Pi uses Supercollider as a server.
My projects, Scheme For Max and Scheme for Pure Data, use the same Scheme as Common Music (s7 Scheme) so you can use it to run Common Music code in Max or Pd. Taube's book "Notes from the Metalevel" on Common Music is really great, though sadly, now out of print.
I've played with a lot of languages for music, including the ones you mention, and for my tastes, Scheme is the loveliest. Eventually I plan to play more with Extempore too.
The author has said that if he would do it in Erlang if he was starting again now.
Here's a talk with him and the late Joe Armstrong. https://www.youtube.com/watch?v=4SUdnOUKGmo