Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!utgpu!water!watmath!clyde!rutgers!rochester!ken
From: ken@rochester.UUCP
Newsgroups: comp.text
Subject: Re: Comments on TeX - The Program
Message-ID: <2768@sol.ARPA>
Date: Tue, 29-Sep-87 19:37:35 EDT
Article-I.D.: sol.2768
Posted: Tue Sep 29 19:37:35 1987
Date-Received: Sat, 3-Oct-87 01:14:31 EDT
References: <3360@blia.BLI.COM>
Reply-To: ken@cs.rochester.edu (Ken Yap)
Organization: U of Rochester, CS Dept, Rochester, NY
Lines: 24

|It appears to me after a much too brief period of study that
|TeX could be a lot smaller and probably simplier if Knuth had been
|able to use a environment containing the function of something
|like the Unix Standard I/O Library. (This is only a hypothetical
|example; I'm not actually proposing this.)

I think the blame should be placed on the Pascal language. Pascal TeX
is one big monolithic program and takes ages to compile. Pascal has no
(standard) facilities for file handiling and separate compilation.  I
suspect that even if Knuth had C and Unix, he would have stuck to
read/write/open, etc. and written his own library on top, because some
of the subtler semantics of stdio can vary between systems (e.g.
behaviour of tolower on lowercase letters between systems).  Just a
surmise, no flames please.

As a matter of interest, Common-TeX is only about 500k of code so C
does help reduce verbosity.  Certainly TeX would be more modular in C.

What I'd like to know from those brave souls who have managed to
rewrite TeX in C is (1) if the translation was literal, how difficult
was it to translate the code, and (2) if only the algorithms were
retained, how difficult was it to rewrite TeX in C?

	Ken