On: Killing Yacc: 1, 2 & 3

[theoj]

Even if it turns out that the new parsing algorithm doesn't shine in all situations, I think it is nevertheless a welcome development. Compiler and interpreter makers can use this algorithm for those grammars and languages where it actually outperforms Yacc.

[sharkbot → theoj]

However, the article points out that we can only use the algorithm when we know it outperforms Yacc; read "know" as "proven theoretically".