Saswat Anand

This is my research page – a place to store my random ideas, reports, links, and anything. My home page is: http://www.cc.gatech.edu/~saswat

I am interested in program analysis techniques for Java.

John McCarthy said:
"Instead of debugging a program, one should prove that it meets its specification, and this proof should be checked by a computer program."

Richard Hemming's three questions for new hires at Bell labs:
1. "What are you working on?"
2. "What is the most important open problem in your area?"
3. "Why aren't they same?"

Donald Knuth said:
"We should forget about small efficiencies, say about 97% of the time: Premature optimization is the root of all evil".

Corollary of Moore's Law( My 2-cents on effect of Moore's law on program analysis):

Scalability of a program analysis = K X ( Speed of Processor / Size of Program )

This equation is pretty intuitive ( K is a constant)! Now how fast the size of programs increase over time? I dont think it is anywhere near 100%, which is the rate at which the processors become faster in 18 months according Moore. Suppose we assume the programs grow 30% (someday I will do some empirical studies on that) in 18 months, it would mean a program analysis would scale by 150% in 18 months (or, 100% in a conference-year)! So should we be so much worried about scalabilty of a technique?

Interesting links: Moore's law, Probesting's Law, Yannis' Law

Problems and projections in CS for the next 49 years John McCarthy. 2003

ACM Digital Library IEEE Xplorer

[ Relevant Papers ] Theorem Proving + Static Analysis + Model Checking

Program Verification Techniques Program Verification Techniques II
[Mar 28]
[NASA]

Old Stuff:
[25 August][1 Sep][15 Sep][4 Oct][14 Oct][01 Nov][02 Nov][04 Nov][05 Nov][21 Nov][03 Dec][08 Dec][19 Jan][26 Jan][25 Feb][06 Mar]

Last modified 10 May 2007 at 12:35 pm by 144