Remarks on recent algorithms for LALR lookahead sets

We have read with more than a passing interest a recent paper by Fred Ives [ 1 ] dealing with algorithms for computing LALR(I) lookahead sets. In the paper a claim is made that our algorithm published earlier in Park et al.[2] has been improved. One purpose of this note is to raise questions about this claim. After trivial changes in notation, such as Path(C, A) for etz(C, A), q 6 Pred(r, a) for "a accesses r from q", etc., Ives' new algorithm reads: