joel@DECWRL.DEC.COM (Joel McCormack) (06/18/86)
Kevin Kleinfelter writes: > Gleaves book is based on some chapters from the Volition Systems > Modula-2 reference manual. (He wrote theses chapters while > working for Volition Systems.) Thus his descriptions are biased > in favor of his company's compiler. While Rich Gleaves did indeed round out the introductory portion of the Volition manual into his book, he nonetheless carefully maintained the distinction between the actual Modula-2 language and Volition restrictions/extensions. Niklaus Wirth has stated that the restriction of return types to scalars and pointers in the ETH compilers is just that, a restriction of his implementations, and not of the language. The first half of Rich's book was intended in part to make definite many of the fuzzier points in Wirth's book, not to slyly pass off Volition's implementation as the real language definition. Rich was in fact reluctant to state outright that structured function results were part of the language until Wirth had more publicly declared this so. Unfortunately, this clarification (and many others) did not make it into even the 3rd edition of the book. C'est la vie. (And how do I speak so authoritatively? Volition's ex-President, Rich's ex-roommate, and two trips to ETH.) - Joel McCormack {ihnp4 decvax ucbvax allegra}!decwrl!joel joel@decwrl.dec.com