George Neuner <gneuner2/@/comcast.net> wrote:
| Threaded programs are not much harder to write if thought about
| properly. Raise your hand if you've read "Communicating Sequential
| Processes". Ok, that might be a bad example here in c.l.l., but if
| you were to take a poll in the general programmer population, I'd bet
| better than 90% have never heard of it, let alone read it.
The orange & white book by C.A.R. (Tony) Hoare that's on that
shelf over there not two meters from my right hand? ;-} ;-}
Great little text!! Loved it!
R.C. Holt's "Concurrent Euclid, The Unix System, & Tunis" is good, too,
as is Cheriton's book on "The Thoth System".
| Actually, apart from a few hands-on OS textbooks I've seen, most books
| that tackle threads suck with regard to teaching synchronization issues.
Not the above three, nor do Dijkstra'a various articles on the subject,
such as EWD464 "A new elephant built from mosquitos humming in harmony".
See also EWD508 "A synthesis emerging?", which starts out:
... It is exciting because it seems to open the possibility
of writing programs that could be implemented
a) either by normal sequential techniques
b) or by elephant built from mosquitos
c) of by a data-drive machine.
And in the Table of Contents of his collection of "EWD"s, "Selected
Writings On Computing", he links EWD508 with Hoare's "C.S.P." [the
1978 CACM article, that is]. Fun stuff.
p.s. Wulf et al's "HYDRA/C.mmp" also had some stuff I found
helpful once upon a time, but it may be a bit too thin in the
parallelism/synchronization area for this thread [pardon the pun].
Rob Warnock <email@example.com>
627 26th Avenue <URL:http://rpw3.org/>
San Mateo, CA 94403 (650)572-2607