3 ms·
I've written in Common LISP. Here's my port of the Boyer-Moore theorem prover to CLisp.[1] I've used original INTERLISP. I've used a Symbolics LISP machine. I o
by Animats 3y ago
I've written in Common LISP. Here's my port of the Boyer-Moore theorem prover to CLisp.[1] I've used original INTERLISP. I've used a Symbolics LISP machine. I once did a lot of work in Franz LISP.[2] And I'm partially responsible for AutoCAD going with AutoLISP as a scripting language.
I've even taken a class from John McCarthy himself, at Stanford.
All that was decades ago. I haven't written a new program in LISP in decades. (In later years Python, Go, Rust, C++, and when forced, JavaScript.)
* Early thinking was that programs needed to be able to modify themselves. This turned out to be a niche feature. It was a holdover from the early days of programming, where storing into the code was considered a normal activity. Von Neumann was into that. To index into an array, load instructions were modified in place.
Once index registers were invented, that mostly became unnecessary. Languages can be too dynamic. You pay a price for that. Python suffers from the design feature that anything can store into anything at any time. This blocks most compile-time optimizations.
* Lists as a primitive are OK, but modifying lists all the time isn't what you really want. Most languages have settled on a useful set of collections today, and their implementations, especially hashes, are highly optimized. Python pretty much got this right, and other languages now use roughly Python's various collection types.
* Saving the entire state of the system, instead of having source files, was a really bad idea for anything that had more than one person working on it.
* The overly clever macro system makes code hard to maintain. (I tried to port [2] to Common LISP. Got about half way. It has the original Oppen-Nelson simplifier inside, the first of what's now called a SAT solver. Originally written in MACLISP (MIT's Project MAC, not Apple Macintosh), and ported to Franz LISP by Oppen and Nelson, it has too many clever macros that I wasn't able to completely figure out a decade ago.)
This is a generic problem with macro systems, of course, which is why C deliberately had a weak macro system.
LISP is a blast from the path. It's fun for retro reasons, but things have moved on.
[1] https://github.com/John-Nagle/nqthm https://github.com/John-Nagle/nqthm
[2] https://github.com/John-Nagle/pasv/tree/master/src/CPC4 https://github.com/John-Nagle/pasv/tree/master/src/CPC4
- mark_l_watson 3y agoThanks John, you make great points on the advantages of ‘moving on’ from Common Lisp. I still use Common Lisp a fair amount just because it makes me happy to do so. While I need to use Python, using Lisp languages is just a preference. Because I am retired now, I have thought of starting with something bare-bones like Chez Scheme and making a long term hobby of building up my own environment mostly just for myself. The author of Gerbil Scheme started with the excellent Gambit/C Scheme and did this, and now the Gerbil ecosystem is really cool.
- vindarel 3y agoyet no other language gives so many tools to the developer… quantum companies would disagree. https://github.com/azzamsa/awesome-lisp-companies/ https://github.com/azzamsa/awesome-lisp-companies/ (BTW: CL isn't Smalltalk which isn't uniquely that anymore, we do use source files and we can compile single-file binaries. My web app weights 35MB, starts up in 0.4s (or 0.01s without core compression))