From joel@.MISSING-HOST-NAME. Thu Jan 24 15:11:38 2002 Date: Wed, 23 Jan 2002 15:44:39 -0500 (EST) From: Joel Seiferas To: CSC484 Subject: Comment on formal model In resolving algorithms such as 2.2 with our formal definition, you may want to reread Section 2.1.3 (Pseudocode Conventions), especially the paragraph that spans pages 14-15. So that computation events remain deterministic, the ``arbitrary order'' of message handling should be a fixed one at each processor. (And, although it may not happen in Algorithm 2.2, what if there are many messages arriving on each link??) In the particular case of Algorithm 2.2, we seem also to need some assumption about how the ``bundling'' is done, right? Does somebody want to more carefully propose a better formal definition? (This is an open invitation.)