Maybe someday Swift could be used both for speed critical code and high-level scripting in TeXmacs?
See: https://forums.swift.org/t/swift-and-c-interoperability-workgroup-announcement/54998
Swift and C++ interoperability workgroup announcement
I’m seeing you posting a lot on this board. You seem to have way more time on your hand than you claimed in one of your posts. Instead of spamming suggestions every day, why don’t you do something useful?
@hamorabi The aggressive tone of your message is uncalled for. This forum should be a place with a welcoming and constructive atmosphere, please keep it that way.
1 Like
I think we want to keep TeXmacs very independent of the platform, as far as I understand Swift is tightly linked to Mac and moreover I see some wise attitude from Joris to try to keep the development with tools which are well proven by time like C/C++ and Scheme. If Swift survive 10+ years we will see. In my short time as a person interested in programming I’ve seen many languages come and go, and some I see staying from the wrong reasons 
I’m pretty satisfyed with the current setup: Scheme is fun to work with and our C++ environment is quite ok for what we need to do, and it is easy to integrate other tools/libraries if we need. Our minimal dependencies is also what make possible to imagine to port TeXmacs to webassembly, for example.
1 Like
Apple planned to use Swift for 20 years. I use it all the time and think it might last even longer.
It’s higher level than C++ but can nonetheless generate fast code for release builds.
For Swift advantages and cross platform support, see:
For WebAssembly, see:
1 Like
Interesting. I was not aware of this. maybe someday…
Btw, Joris has developed his own language, Mathemagix, so he would like to rewrite everything in it.
But seriously: in TeXmacs we leverage a lot lisp-styles macros, and I do not know any other language apart from scheme & similar which has these facilities. I would be happy if we can put more of the code of TeXmacs in scheme and less in C++, for example for the UI part. I’ve tried to integrate Chez Scheme, which is very fast and compiled, but the interaction with C++ code is not easy to handle and would require more work and time than I can put in it. Personally I’m very happy of our current state of affairs, it is quite pleasant to develop TeXmacs in C++/Scheme, which actually means to develop it in the particular dialect you have to use to integrate well with the existing codebase. Another programming language will not bring anything to this stack. We allow some small parts written in Python (in the plugins) but in my opinion we should move them to scheme. If we run in the browser we will be forced to have some glue code written in Javascript.