Recently, a former colleague mentioned a variant of C which rewrote the language as S-expressions, but otherwise left it as C, so a translator could easily convert it. This gives you more than you might think, as it allows the whole of the lisp macro facility to extend the “SC” language, as it turns out to be called. My colleague either did not remember the name of the language or it slipped my mind, so I eventually aksed her, and getting no reply, (she is quite busy these days trying to create something) I started looking around for the quarry using Google.
I initially found BitC and was quite disappointed because it is the usual architecture astronomy, entities multiplied beyond any need and all constraints abandoned, war horses breeding on the outskirts of town, etc.
Increasingly, interesting ideas in software (in my opinion) are coming from Japan. For example, Ruby, which as we all know is an adequate lisp. And now, SC.
Zees guy, you see, had the idea: what if we just wrote C in s-expressions.
Simple. Constrained. Elegant (IMO). Eminently doable.
AND miracle of miracles, he didn’t turn it into a typical CS debacle, he and friends just … implemented it, and looked into how it could be used.
Anyhow, after several deciding to try again (BitC couldn’t have been what she was talking about. It was too… much) I finally found some papers about a language called “SC”.
Now let me digress a bit and talk about a former life where I wrote commercial applications in Common Lisp, which we translated into C, compiled on 16 different platforms including Vax VMS, and produced industrial quality software which was itself a development system called “G2″, a product of the late lamented Gensym corporation. Gensym was itself a spinout of the death throes of LMI, Lisp Machines Incorporated, and when I worked there we were located int he old offices of Infocom. Talk about history.
Anyhow, one of the last projects that my good friend and former colleague Jim Allard led before the salad days at Gensym had passed was a new lisp to c translator. He and Ben Hyde eventually persuaded Lowell Hawkinson to release this work into the public domain (YAY Lowell!) and so the Thinlisp project came to be.
Thinlisp is an interesting idea, although to my knowledge noone has used it for much. It’s kind of an elegant hack, because it … simplifies the LISP programming model to allow translation to C, and many LISP hackers would consider this anathema.
But now that this idea of SC is in my head, I wonder how it and thinlisp might play together.
Here’s the link to the source code for SC, which is tricksy tricksy tricksy to find:
http://www.yuasa.kuis.kyoto-u.ac.jp/~hiraisi/sc/sc080709.tar.gz
