Re: time to write a compiler

beckmann@das.harvard.edu (Gary Beckmann)
Fri, 9 Nov 90 11:41:59 EST

          From comp.compilers

Related articles
time to write a compiler roman@gaudi.ccsf.caltech.edu (1990-10-31)
time to write a compiler meissner@osf.org (1990-11-05)
Re: time to write a compiler holliday@csgrad.cs.vt.edu (1990-11-06)
Re: time to write a compiler marcel@vlsi.cs.caltech.edu (1990-11-06)
Re: time to write a compiler ge@sci.kun.nl (1990-11-07)
Re: time to write a compiler bright@nazgul.UUCP (1990-11-08)
Re: time to write a compiler pohua@polaris.crhc.uiuc.edu (1990-11-09)
Re: time to write a compiler beckmann@das.harvard.edu (1990-11-09)
| List of all articles for this month |
Newsgroups: comp.compilers
From: beckmann@das.harvard.edu (Gary Beckmann)
Keywords: design
Organization: Compilers Central
Date: Fri, 9 Nov 90 11:41:59 EST

This should maybe go into alt.hack, but since I didn't do it myself
and since I know how long it took I'm sending it to comp.compilers.


This was not a production quality compiler.


I introduced a friend of mine to the wonders of C a few years back.
His company (whose name and location will remain secret--for obvious
reasons as you read on) would not support the buying of a compiler for
his ibm-pc clone. All he had was basic. So he wrote a C compiler in
basic. Took him seven months working on the side.


He was dissatisfied with the performance of the compiler so he then
proceeded to write a basic compiler in C !! (My neck still aches from
the head-shaking I did when he informed me of his plans.) That took
him only three months -- he claimed that the speed was due to his
familiarity with basic and the relative simple flavor of basic with
which he was working.


He was pleased with this "project" and decided to port it all to an
Apple II which he had. This port was never completed-- I believed he
failed due to the radical differences in basic.


To repeat: this was NOT a production quality compiler (I think he even
had some restrictions on the syntax.


Hope I haven't horrified to many . . .


Gary Beckmann
beckmann@das.harvard.edu
--


Post a followup to this message

Return to the comp.compilers page.
Search the comp.compilers archives again.