This is an extremely well-written paper which will probably be the major reference when it comes to the grid-following algorithms for drawing circles. However, here is my criticism: Is this really a very important problem? Who needs 26 pages on this very narrow subject? It seems to me that most of the mathematical analyses carried out in this paper is of the sort we come across in journals like the Journal of Approximation Theory. I personally find this boring.
To be fair, I applaud the fact that the author has included all his program listings (in PASCAL -- another nice point), which are carefully documented. But with all due respect to the high quality of this paper, I ask the same question: Don’t the readers of ACM Transactions on Graphics deserve papers on broader subjects in a journal which only comes out once every three months?