You could ask Spinny, who's never been able to answer direct questions
on issues like this (or at least, never able to do so coherently), or
you could look at the Java rationale.
My name isn't spinny.
I don't think you understand "the Java rationale", otherwise you'd
take the risk of explaining it. However, I think you cover up your
ignorance with phrases like "you could look it up" and when that
doesn't work, trashing other people.
I should point out that the decisions in question were made LONG before
I was involved with C standardization, or indeed, with C in any form.
I think he's claiming that I got involved with stuff I didn't understand.
(Which is arguably true, and one of the reasons I listened more than I
talked at most C meetings.) But your point stands -- Spinny's claims
about what I would have learned from a compiler development class appear
to refer specifically to a spectacularly bad compiler development class
in which the course material is full of errors.
Explain what justification you have for making this statement. You
see, you've said that you've never attended a CS class. Nor does what
you've done on the C99 effort, nor said here, show that you even
understand compiler optimization theory that recognizes that SOME
expressions can be reordered but that it's untrue that ANY can be
safely reordered. Because of this, you made an even bigger error, it
now appears to me. You made an unsafe language more unsafe than it was
already, by putting the onus on thousands of C programmers to figure
it all out, while conducting a campaign of personal destruction
against a computer author who was trying to make sense of the mess you
made.
You appear to have used a superficial knowledge of compiler
optimization to infer, incorrectly in light of Aho/Sethi/Ullman, that
ALL constructs can or should be reorderable in a way that violates and
insults the intelligence of the ordinary programmer even more so than
K & R C, because by your own admission you did not then (and may not
now) know what you were doing.
Were you aware in 1999 that some constructs can be safely reorderable
given dataflow analysis, and others not? Or did you participate in a
standard that made large classes of unsafe constructs reorderable not
knowing anything about the ability, then and now, of optimizing
compilers to identify safely reorderable code?
Did vendor salesmen, in fact, convince you that according to their
compiler experts, no constraint could be placed on reordering that
might not break their compilers although it was known at the time how
to do this? At this point, it appears to me that owing to your lack of
education, and mental laziness, you made a serious error.
You appear not to have read the literature in lieu of taking the class
since you worked on a standard in which this distinction is not as far
as I can see made. You have stated that you acted fraudulently since
we need people with qualifications and understanding to standardize
programming languages.
Please identify at least seven books on computer science you have
read, right here. Also, I need to know what you majored in and whether
you received an undergraduate or graduate degree, in what major, and
from what institution.
Reading alone would not make you a collegial person educated in
computer science, for you share with Richard Heathfield the ignorant
school-leaver's penchant for transforming ideas into personalities,
and spreading lies in lieu of being qualified to discuss ideas. In
fact, when you encounter prose beyond a certain level of complexity
that is nonetheless grammatical but not from a person with power over
you you mock that prose like a high school dropout. Are you a high
school dropout?
What, in short, were your qualifications for pulling your stunt "C:
The Complete Nonsense?" What, in short, were your qualifications for
being a member of the C99 standard effort?
