HALLAM@vax1.physics.oxford.ac.uk ("Phillip M. Hallam-Baker") (02/14/91)
In hin arrticle Steven Ericsson writes :- > But anyhow, now you're waffling (and maybe I am too:). Yep probably. Anyhow the points I was trying to get across were 1. The article was unclear It is *NOT* pompous to complain about things being unclear. There is little point of carying on a conversation if you don't understand the other person's point. The problem I was having was that The original article was ambiguous. I can guess what is meant but that is not much good if I have two equally good guesses... (This does not however constitute a refutation that the manner of my complaint was pompous but I'm a kinda pompous sort of person...) [side note what is the point of denotational semantics if not to be precise and unambiguous] 2 The drift of the article was "oh occam is useless you have this copy penalty you see so let's forget it" >Well, I would like to see a study done of the claim in your second >sentence here Yep I would like to see one to ! You are right that someone should come out with some figures before making dumb statements - but I was responding to an article which had originaly been posted without supotting data so I figured that to produce some before posting a refutation was hardly appropriate. All I know is that in *MY* application the copy penalty is not a problem. 3 Where has your requirement for coding clarity gone? ... Out the window, that's where ;-) OK - Ill explain how I maintain coding clarity - soon hopefully It is the subject of my DPhil. Basicaly My premise is that to turn off usage checking you have to provide the checking at a higher level in your hiereachical program design. My source code maintains clarity by relegating the business of shared memory to a back end "optimiser". Basically I use Occam like yacc uses C - as the intermediate language in a code translation system. 4 I get totaly p****d off by introduction of jargon like `consumer orientated' languages. Especialy when the terms are either imprecise or make a valueless distinction. I get the feeling that these terms are invented mainly to make articles sound impressive - take out the consumer/producer stuff and the original article boils down to occam has constructs to handle parallel code. Parallel Fortran dosen't consider it a problem therfore use parallel FORTRAN (which the autour is on the committee for). It's committee speak ! The only value is to enable committee beuraucrats to look as if they are doing something clever. It also has the advantage that if you haven't kept up with the latest jargon you look silly and the jargon inventer looks like and expert. But then again this may just be me being pompous... Ho Humm enough waffle how about doing some work ?? Phillip M. Hallam-Baker PS Since this net seems to becomming a buletin board for prospective post docs - is there anybody on the USA side of the pond who is looking to fill a post doc position? Our beloved government has just axed Nuclear Structure research and is cutting back in Info Technology. I am working on methods of writting *LARGE* software packages for parallel (transputerish) systems. Am currently working on the 1000 transputer ZEUS data aquisition system in Hamburg Germany.