Re: Statement on Schildt submitted to wikipedia today

On 2009-09-11, spinoza1111 <spinoza1111@xxxxxxxxx> wrote:
Having left programming, in no little disgust at the personalities of
programming, after thirty years, I teach English and creative writing
in addition to intro computer science. I have learned that intolerant
students, who believe words are things, make the slowest progress.

Well, they make the slowest progress when you teach them. If you actually
teach them. If any of what you're saying is true.

Therefore, I will be the judge as to whether a text has ?intrinsic
worth?.

This is a non-sequitur. You have not offered a qualification to judge,
and if anything, you've offered a sort of non-qualification.

Welcome to the weird and wonderful world of C. This is why we have
Java, isn?t it.

Nope. Java wouldn't do us any good here.

Nonsense on stilts, Socrates.

Uh, no. Simple fact.

What you?re talking about happens to be a collision of the stack with
practices from the Ice Age which predated the stack, practices which
infected the C language, unfortunately, practices which do not show
intellectual superiority and are shibboleths.

I made no claims about "intellectual superiority".

I claimed a *reality* -- on real machines, with real compilers, believing
Schildt's crap would leave you unable to comprehend or debug a problem which
really occurred.

These include the use of finite and hardened registers which remains a
necessity at the low level because the MIPS kiddies actually
rescusitated this US-centric praxis in 1987, claiming, without much
evidence, that it was ?efficient? in a singularly sloppy use of
English.

When you can make a CPU do more work in less time on less power, please
let us know.

And you are saying that by teaching the stack (which is good practice)

You keep asserting that this is "good practice", but I've just demonstrated
that it's genuinely harmful.

Herb FAILS to teach arcana, and makes the tyro incapable of also
thinking at the more primitive (not ?better?) level of the single-
level register. Arguably, the student needs to think at both levels,
but the stack has priority because you can write slow OSen without
thinking in assembly language!

The student does not need to learn to think at either, but if you're going
to start thinking about how things "really work", it's crucial to think about
how actual machines work rather than about how Schildt thought some DOS box
worked.

You are creating a class divide which I reject.

No, I am pointing out reality. Which you reject.

You are saying that
bottom-feeding computer science *privat dozents* at DeVry and Borstal
may not speak of certain matters.

No, I'm not. I've merely said that people who make false claims are not
making true claims, and aren't going to get anywhere.

You?re here worse than student ?Otto? who sits in the back of the
class with twenty years of assembler and no health insurance,

Okay, so.

Basically, you have some kind of issue, totally unrelated to C, involving
health insurance and job security.

And in your mind this is all connected into one vast whole. But because
the connections occurred in a non-analytical part of your brain, you can't
tell anyone what they are, or show them to anyone, or offer any kind of
evidence for them.

I?m dead serious, Peter. The difficulty of programming creates
psychosexual noise even in the best brain, and the stack just bothered
people, being flexible and soft: whereas the registers are always
?there?, crystalline and hard.

If you're dead serious, you are schizophrenic. Not joking, not insulting,
just telling you what is fairly obvious to anyone with access to relevant
texts.

You emit "word salad" -- strings of words which aren't always even
grammatical. You experience huge and elaborate conspiracies which are
so crystal clear to you that you can't even comprehend people not being
aware of them. Every literary reference, every cultural reference, every
experience, is somehow tied to all the others by relationships which don't
exist in normal brains.

Schildt was describing the vast majority of cases for the vast
majority of programmers,

Even if true, so what? He described them incorrectly; even on x86, his
description of how things work is only true some of the time.

who didn?t go down to Oxocantabridgia or
HarvardusPrinceurbsania to banter about the Higher Sodomy. No, they
went to Iraq for money for school, mate, and they aren?t going to be
programming any bare metal any time soon, they are going to maintain
applications. If they are not introduced to the stack, they won?t be
able to debug.

This is simply untrue.

Only if you generalize. Use your common sense. People know that there
are different computer architectures. Learning is lifelong.

Well, that's the thing.

People who didn't read Schildt know that there are different computer
architectures. People who are steeped in Schildt's madness frequently
come out of it convinced that he actually described "how computers work"
rather than giving a specific example of one common implementation, and
are thereby gravely disadvantaged when they come to other systems.

In short, yes, there is a real risk of creating a class of disadvantaged
programmers who can't compete effectively or adapt to new environments.
That risk is Schildt's books.

Nobody is going to believe that Schildt is talking about ALL POSSIBLE
computers, nor do they seek to learn about ALL POSSIBLE computers. Use
your common sense and credit people with common sense, for the love of
Mike.

I have been reading Usenet far too long to credit people with common
sense.

You fail to notice that to understand the advanced situation, your
auditor has to know about?stacks.

But not about the things Schildt describes as "how the stack works".

So, Schildt is not training rocket scientists (nor people who fancy
themselves rocket scientists). Well, whoop de doo.

Not just that. Schildt is teaching people falsehoods, which will make
it harder for them to learn what they need to know to do their jobs.

Our only point of possible agreement is that having trained people at
Princeton and having been a *privat dozent* at DeVry, I believe there
is a class divide in programming education, and I think it?s most
unfortunate. But Schildt didn?t create this class divide.

He's certainly making it worse.

It is the height of absurdity to criticize Schildt for talking about
the stack without also talking about registers.

I didn't.

I merely pointed out that his strategy demonstrably renders people unprepared
to deal with real-world systems.

A more effective solution, by far, is to teach C without trying to explain
how it "really works" as though there were a simple answer, then get into the
details of implementations in more advanced courses. That works beautifully.

It would make most
introductory computer science classes impossible. I can only conclude
that yes indeed, you, Mr. Seebach, believe that only certain social
classes should learn of Eleusinian mysteries, and that the rest of us
should not be permitted to read turbulent and damnable books above our
station.

That would be an exceptionally stupid conclusion. I think everyone should
be allowed to read everything -- but that "everything" includes clear warnings
about significant errors in technical books.

No-one expects the Spanish Inquisition, Mr. Seebach. But after
Thatcher, Reagan, and the Bush-Clinton highjack of the public interest
for private gain, we?re seeing that reversion to barbarism can happen
pretty quickly. The trashing of Schildt was barbaric given the global
unsafety and global untruth of C.

Again, here's where you get off into schizophrenia land. You're wandering
off into unrelated politics, and come *ON*, the phrase "untruth of C" is
just plain crazy talk.

If you've got meds, go back on 'em. If you don't, see a doctor.

-s
--
Copyright 2009, all wrongs reversed. Peter Seebach / usenet-nospam@xxxxxxxxx
http://www.seebs.net/log/ <-- lawsuits, religion, and funny pictures
http://en.wikipedia.org/wiki/Fair_Game_(Scientology) <-- get educated!
--
comp.lang.c.moderated - moderation address: clcm@xxxxxxxxxxxx -- you must
have an appropriate newsgroups line in your header for your mail to be seen,
or the newsgroup name in square brackets in the subject line. Sorry.
.

Relevant Pages

  • Re: C: The Complete Meta-Nonsense
    ... such as telling Schildt not to say that C negatives ... In other words, you don't care about truth, you care only about status. ... But note that even CREATING those bug reports is often programming beyond ... was frightened by a stack when he was in her womb and who hates ...
    (comp.lang.c)
  • Re: Statement on Schildt submitted to wikipedia today
    ... You know that a great deal of this argument about needing a stack resolves round what that word is meant to mean. ... Schildt meets those two criteria but unfortunately when I first looked a bit closer I discovered that the kind of errors he was making strongly suggested that his mental model of C was erroneous even if limited to a DOS based machine. ... I regularly checked his books over the next decade and over those years I saw little if any improvement until right towards the end when the last couple of books I looked at did show signs that he had developed a better understanding of his subject matter. ... It is harder to pin down in the case of programming, but nonetheless bad programming based on a poor or erroneous understanding of the language being used is costing many innocent users a great deal. ...
    (comp.lang.c.moderated)
  • Re: The annotated, annotated, annotated C standard continued
    ... But we see Schildt making the same errors time and time again. ... His books on C++ do C'isms. ... Bjarne Stroustrup's The C++ Programming Language, ... I dont see anything like the same depth in any of Schildt's 4 erratas. ...
    (comp.programming)
  • Re: The annotated, annotated, annotated C standard continued
    ... on Schildt. ... And I posted stuff here on Schildt's flaws in "STL Programming from the ... standard then and what the C standard _REALLY_ says. ... The language of "Schildt's errors" as well as Clive Feather's strategy ...
    (comp.programming)
  • Re: Bloody Java again
    ... I passed my Java exams last year but really have ... Erlang's concurrent programming is rather abstract and high-level. ... He is not being asked to grade umpteenth different languages. ... Student: "It's all in the program, ...
    (comp.lang.functional)
Loading