[plt-scheme] Computers considered harmful

On May 7, 2009, at 09:51, hendrik at topoi.pooq.com wrote:
> And I'm sure debuging will never replace careful thought as a viable
> methodology for anything that has to be reliable.

Twenty years ago I worked on a verification project that, as it was  
described to me, involved a program I never saw that must never be  
run, but if it is ever run, must run correctly.  To ensure this, they  
had dozens of advanced CS people pour over the code, line by line,  
over a period of years.  The problem was that people would move on or  
die during the review, and after several years, people also forgot  
where they were.  My task was to help speed the process by hacking GCC  
to write ASTs to a database so all the code could be annotated and  
each reviewer could see the code the way he/she wanted, including  
personal renaming of things, and code coverage by reviewers could be  
tracked.

And with that, I think I'll go see the new Star Trek movie.