Okay, so lisp is great for symbolic expressions because of homoiconicity… how come Maxima hasn’t come to dominate Mathematica, then?
Although I suppose Mathematica is a lisp, just one implemented with mexps instead of sexps…?
That seems like a strange argument to make. Mathematica is a lot more polished than Maxima, and you have a lot more resources to make something polished if you charge each user $1,095 a year :)
It’s not an argument, it’s just that lisp advocates routinely use CASes as an example of where lisp is absolutely amazing. If so, how come the actual CAS that was built on common lisp isn’t amazing? In fact, Maxima’s source code is kinda considered write-only and very hard to understand.
Repeating what I said, if your software costs $1095/user/year, you can do amazing things. I’m not saying Lisps are absolutely amazing, but you’re making a very unfair (and strawman) comparison.
Macsyma was expensive. It’s not just the money. Nobody has made a significant attempt to make a CAS in lisp again since then. Yacas is in C++, Sympy is in Python, Sage is mostly in Python too. If lisp is the natural language for CASes, why isn’t it being used for CASes?
Please note that I never said that Lisp was better to build CASes. I just said that with Lisp the CASes could be embedded in the language, while with other languages it has to be a separate program.