home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.os.ms-windows.programmer.tools
- Path: sparky!uunet!mcsun!sun4nl!utrcu1!infnews!kervel
- From: kervel@cs.utwente.nl (Rinie Kervel)
- Subject: Revolution or madness??
- Message-ID: <1992Sep13.193749.12198@cs.utwente.nl>
- Sender: usenet@cs.utwente.nl
- Nntp-Posting-Host: rijn
- Organization: University of Twente, Dept. of Computer Science
- Date: Sun, 13 Sep 1992 19:37:49 GMT
- Lines: 82
-
-
- Revolution in Parsing ?? (C) August 17, 1992 Rinie Kervel
- e-mail: kervel@cs.utwente.nl
-
- Either I am insane or I have uncovered a flaw in compiler theory.
- As a result I have devised a new parsing method that has it all:
-
- - simple
- - small
- - fast
- - excellent error handling
- - reusable components
- - intuitive
- - extensible
- - applicable to weakly structure languages and computer files
- (shell scripts, X resource files, DBASE, COBOL or BASIC)
- - easily used in hand coded parsers
- - can be implemented with parser generators
-
- I know this sounds too good to be true. (At first I felt inclined to brand
- myself a fool, now I am prepared to take the test). Before you dismiss
- this message, I am not entirely new to compilation:
- - I have taken a (university) course in compiler construction
- - I read all the books:
- Aho, Sethi & Ullman's dragon book
- Fisher & Le Blanc
- Allen Holub
- Schreiner & Friedman
- - I have written yacc parsers, studied yacc & bison.
- - I have examined compiler sources & generated code
-
- The revolutionary effects mean that most phases in compilation should be
- reexamined systematically. However, evolutionary effects can be achieved
- within weeks. Resulting in faster compilers with good error handling, even
- in yacc generated parsers.
-
- Now I am faced with some conflicting goals:
-
- - I want the credit
- - I want to make it generally available for public domain or GNU-like
- software and end users
- - I want to get a fair share of the money saved in the
- development, and a share of the additional value of commercial
- products.
-
- I can explain the idea to anyone that has a basic understanding of
- compilation in 15 minutes. Therefore I will not even give a hint
- of how it's done before I know what to do with the idea, without getting
- ripped of.
-
- In my opinion, the only way to meet these goals (partially), is to contact
- some computer hot shot with vision, in a US company, say:
-
- - Borland's Phillipe Kahn
- - Sun Soft's Bill Joy
- or less commercially:
- - AT&T's Dennis Ritchie or Bjarne Stroustrup
-
- (Yes I know I am very very humble).
-
- My questions are:
-
- - How do I reach them.
- - What are my chances of getting ripped of.
- - How do I prevent this.
- - How can I protect the idea.
- - Are there better ways to achieve my goals.
-
- The new method has the simplicity of bubblesort and the power of quicksort,
- whereas traditional parsing has quicksort complexity with bubblesort
- efficiency. (Yes this is an understatement!).
-
- Please answer by e-mail:
-
- kervel@cs.utwente.nl
-
- or write to (please announce through e-mail):
-
- Rinie Kervel
- Campuslaan 27-206
- NL 7522 NC Enschede
- Netherlands
-