subroutine stack and C machine model

[Keith Thompson]

In


Oh dear, looks like spinoza1111 may well /not/ be getting it. What a
shame that sock puppets can't be brighter than their puppeteers.

And how exactly is this relevant to C?

 

[Walter Banks]

Object code generation wasn't fun.

How did word length make object code generation inordinately complex?


w..

 

[Colonel Harlan Sanders]

Your dishonesty is more global.

By "global dishonesty' you apparently mean I showed you up.
Whereas you flat out lied when you said you had never used a sock
puppet on Usenet.

For example, it was both dishonest and
stupid for you to pretend to understand the Oxford English Dictionary,
since you shat out all definitions of "clear" without even
understanding that we're talking about a text, not water or beer.

The point, as I've said three times at least, was that none of the
definitions, whether for water, beer, or text, support the definition
you claimed over and over was there, but everyone was "using the wrong
dictionary". So I quoted the whole damn thing so you could point out
the part you meant. And you couldn't.

There is no equation of "clarity" with "truth" in the English
language. Common senses tells everyone here that; the Oxford
Dictionary tells us that.
And your response: ducking and diving, insults and doggerel.

And, asshole, you've called me a liar repeatedly, yet cannot cite a
single statement where I have lied.

Fits your pattern of spewing out outrageous insults, slurs, absurd
conspiracy theories and abuse when you have no answers.

 

[Dennis \(Icarus\)]

If it's not clear, then you really can't assess whether errors are present
or not.

But Seebach was the careless writer, and his carelessness, unlike any
carelessness on Schildt's part, caused provable harm to a person.
Seebach labels as errors matters of taste and style. For example,
Seebach claims that the use of upper case on an obviously case-
insensitive file system is an error, and it's not.

It is an error, as has been previously explained.

<snip>
Dennis

 

[Seebs]

Sure you do Hektor.

Stop hectoring him.

-s
(What does this have to do with C? Well, I'm writing on a keyboard,
and surely no one can doubt that type punning is relevant to C.)

 

[Walter Banks]

You were the only one to pick up that 7-channel tape and 6-bit characters
suggest IBM BCD Card Image (026)? 6-bit BCD is directly translatable to 8-bit
EBCDIC (029) without fuss. "Our old 709 has 7-channel tape and our new 360 has
9-channel tape. What's the difference?" never came up in your group?

Perhaps I'm being unfair but we had this figured out in 1963 I think it was.

Paper tape came in 5,6,7, and 8 channel varieties. The real data zoo lived in
iron curtain countries with lots of data rates thrown in for confusion.

w..

 

[spinoza1111]

And how exactly is this relevant to C?

Hast thou not heard? Hast thou not known? Hatred and anger are always
relevant, and bile is always ontopic.
The bullies and thugs here must maintain their position or else all is
lost.
They fight as the wild sheep of Afghanistan defecate, at random they
lock horns:
For they have become Last Men, in the nacreous glow of their computer
screens.
Thus not for them is the delight in knowing, nor in learning, nor in
curiosity:
But instead the sour taste of bitter shadow fruit, the thought "yea, I
am right and he is wrong".

 

[Dik T. Winter]

>
> I have no doubt that this was a consideration, and the standard calls
> perforated tape one of the "principal media", and takes "physical
> limitations of media" as a consideration, but what it says in the
> section "A3. Set Size" is:
>
> A 7-bit set is the minimum size that will meet the requirements
> for graphics and control in applications involving general information
> interchange. Both a 6-bit and an 8-bit set were considered and
> rejected -- the 6-bit, providing only 64 graphics, could not
> accommodate essential format effectors, such as "carriage return",
> "line feed", "horizontal tab", etc; the 8-bit because it provides
> far more characters than are now needed in general applications.

Please note that the first ISO code was 6 bits... But indeed, the first
version of ASCII, as of 1963, was 7 bits, but did not yet contain lower case
letters. They were introduced in 1965 (but that standard was approved but
never published nor used). Finally in 1967 came common ASCII as we know it.

 

[spinoza1111]

By "global dishonesty' you apparently mean I showed you up.

If by "showed me up" you mean you took careful aim at your foot and
fired, hitting your target, why then I guess you showed me up.

Whereas you flat out lied when you said you had never used a sock
puppet on Usenet.

How old are you, and does your Mommy allow you to use the Internet?

The point, as I've said three times at least, was that none of the
definitions, whether for water, beer, or text, support the definition
you claimed over and over was there, but everyone was "using the wrong
dictionary". So I quoted the whole damn thing so you could point out
the part you meant. And you couldn't.

You didn't quote because a person cannot quote what he doesn't
understand. And you missed the part where I did point out the part I
meant. It has to do with texts, and texts are clear only if they help
the understanding.

There is no equation of "clarity" with "truth" in the English
language. Common senses tells everyone here that; the Oxford
Dictionary tells us that.
And your response: ducking and diving, insults and doggerel.

Yeah, you is Sonny Liston
And I is Muhammed Ali
Floatin' round you like a butterfly, chump
And stingin' you like a bee

And, asshole, you've called me a liar repeatedly, yet cannot cite a
single statement where I have lied.

You can't write, asshole, a positive lie. You are a lie.

 

[Colonel Harlan Sanders]

If by "showed me up" you mean you took careful aim at your foot and
fired, hitting your target, why then I guess you showed me up.

There really is no context here. No reasoning. You're just spewing
abuse and declaring yourself correct by divine right.

How old are you, and does your Mommy allow you to use the Internet?

Good comeback.
Relevant? Not so much.

You didn't quote because a person cannot quote what he doesn't
understand. And you missed the part where I did point out the part I
meant. It has to do with texts, and texts are clear only if they help
the understanding.

You pointed out a definition to do with "Understanding", yes. Which
wasn't what was in dispute, what you originally claimed: "Truth". With
the full definition all on the record, you have to backpedal and
pretend you meant something else .

And others here have pointed out the absurdities if one accepted your
equation: Any understandable statement, sorry, "text", must therefore
true. "Nilges enjoyed a good fisting on the weekend." This is clear,
understandable. Therefore, under Nilgean logic, true.

Yeah, you is Sonny Liston
And I is Muhammed Ali
Floatin' round you like a butterfly, chump
And stingin' you like a bee

Maybe in your mind that is a rebuttal.

You can't write, asshole, a positive lie. You are a lie.

"You are a lie". Well, so now I don't exist.
Nobody exists except you, of course.

 

[Dennis \(Icarus\)]

You were the only one to pick up that 7-channel tape and 6-bit characters
suggest IBM BCD Card Image (026)? 6-bit BCD is directly translatable to
8-bit EBCDIC (029) without fuss. "Our old 709 has 7-channel tape and our
new 360 has 9-channel tape. What's the difference?" never came up in your
group?

It wasn't IBM BCD card image.

Dennis

 

[spinoza1111]

There really is no context here. No reasoning.  You're just spewing
abuse and declaring yourself correct by divine right.



Good comeback.
Relevant? Not so much.



You pointed out a definition to do with "Understanding", yes. Which
wasn't what was in dispute, what you originally claimed: "Truth". With
the full definition all on the record, you have to backpedal and
pretend you meant something else .

And others here have pointed out the absurdities if one accepted your
equation: Any understandable statement, sorry, "text", must therefore
true. "Nilges enjoyed a good fisting on the weekend." This is clear,
understandable. Therefore, under Nilgean logic, true.

Texts are not even ordered sets of "statements": they are wholes which
have to be understood in toto, which is why there's no such thing as a
clear text which does not contribute to justified true belief. To
someone who understands texts, there's no such thing as a standalone
"statement", which is why Seebach made a fool of himself: finding 20
questionable turns of phrase and about five typo-bugs when drunk does
NOT make for a technical review, and McGraw Hill would hopefully had
refused to pay him had he accepted their offer.

Maybe in your mind that is a rebuttal.

No, it's poetry. You can't write it.

Si vous aviez un peu de lettres et d'esprit:
Mais d'esprit, ô le plus lamentable des êtres,
Vous n'en eûtes jamais un atome, et de lettres
Vous n'avez que les trois qui forment le mot: sot !

- Rostand

 

[spinoza1111]

If it's not clear, then you really can't assess whether errors are present
or not.

If it's not clear they probably are.

It is an error, as has been previously explained.

No, it's not. If the code works as intended on a Microsoft platform
and that platform is case-insensitive, the code is correct if it was
intended to run that platform.

Very few Microsoft C coders intend their code to run on non-Microsoft
platforms. If they were to want to write a portable program, they
wouldn't use C, because no matter what style C is written-in, the code
is not portable because it takes line by line analysis to make sure it
will run on a new platform, C being both "low level" and a series of
stupid errors in language design (of which making file identifiers
into Holy Writ is one of the stupidest and most ugly).

Microsoft coders have truly portable languages which don't even
require Windows because as we know .Net is portable as well.

Non-MS coders if they are professionals use Java and not C. If
efficiency is an issue they use best of breed algorithms and tune for
the most common cases and within the most frequently used code
pathways.

 

[Colonel Harlan Sanders]

Texts are not even ordered sets of "statements": they are wholes which
have to be understood in toto, which is why there's no such thing as a
clear text which does not contribute to justified true belief.

Why don't you just your time writing your own dictionary, then you can
make yourself (and Schildt) correct by definition, all the time.
That's basically what you're trying to do now anyway, might as well
formalise it.

To
someone who understands texts, there's no such thing as a standalone
"statement", which is why Seebach made a fool of himself: finding 20
questionable turns of phrase and about five typo-bugs when drunk does
NOT make for a technical review, and McGraw Hill would hopefully had
refused to pay him had he accepted their offer.


No, it's poetry. You can't write it.

Since you've disappeared up your own asshole now, farewell.

 

[spinoza1111]

Why don't you just your time writing your own dictionary, then you can
make yourself (and Schildt) correct by definition, all the time.
That's basically what you're trying to do now anyway, might as well
formalise it.

The half-educated and the auto-didact have an unwarranted faith in
dictionaries. Are you aware that there were no dictionaries in
Shakespeare's time, Milton's time, or John Bunyan's time, yet they
were better writers of plays, poetry and prose than anyone alive
today? The original dictionaries were essentially guides for social
climbers who wished to use "correct" spelling in order to impress
their "betters".

Both Seebach and Clive Feather, by confessing that Schildt was
"clear", tipped their hand, because they would have preferred him to
be unclear in the sense of producing a book that Standards experts
would find perfectly true, and which would have been unclear and
useless for its intended purpose. This is because C is not a modern
language: it is a series of mistakes. To become expert in those
mistakes stunts your growth.

 

[spinoza1111]

In

spinoza1111wrote:



Without troubling my memory too much, I can personally recall eight
different large projects on six different sites (i.e. for six
different companies), where we developed using Microsoft C but where
the target platform was a non-Microsoft platform (sometimes Unix,
sometimes MVS/OS390, sometimes set-top boxes).

Generalizing from failure, I see. There were probably thousands of
bugs in what you did, but based on your behavior in this ng, you
doubtless found other people to blame.

It was precisely because we needed portability that C was chosen.

Did you also design the rudder on the Titanic?

I say this because you don't seem to know what portability is. News
flash. It's not the availability of a compiler. It's whether the code
runs correctly when written naturally on multiple platforms. You
yourself have pointed out that a large amount of specialized knowledge
is involved in writing truly portable C and it would be best for you
to GET OUT OF THIS DISCUSSION and write a book or teach classes in
what you honestly know...you probably won't do so because you prefer
being a bully.

If Java or .Net was not available at the time you worked on this
project on some of the platforms, it probably would have made more
sense for you to port, not C code, but Java or .Net to the platforms
needing one of those tools, engaging a contractor as necessary.

Alternatively you should have used Cobol which despite its flaws is
far more portable than C, and more suitable to the financial
environment in which you worked.

You were in fact profoundly irresponsible with what Eric Hobsbawm
calls "the destructiveness of the lower middle class" because even if
everything worked, you left code that it easily changed, intentionally
or unintentionally, to something completely non-portable with
incorrect behavior (starting with memory leaks) that is never
detected.

You designed, I believe, the rudder on the Titanic and then you took
the money and left Belfast for London.

It's called "writing the code", and - where portability matters -
portability violations are considered bugs at code review.

I don't think you're qualified to attend a code review.

One of my clients had a half-million-line set-top box Web browser, of
which 495,000 lines were written in ISO C, and only 5,000 lines,
carefully isolated into separate modules, had to be rewritten for
each new platform. (I take no credit for that, by the way - when I
joined the project, the basic code had already been there for quite a
while, and we were just working on new features; my mentioning it is
intended not for false kudos but as a data point on portability.)
There was no need to review the 495,000 lines for each new platform -
we just ran the tests after each port and after each feature we
added.

And left a mess for the next fool in the same way as banks sold
tranches of toxic loans.

You don't seem to be aware that testing can reveal the presence of
bugs and not their absence.

These large software systems cannot, owing to their size, be properly
evaluated by anyone, least of all their developers, especially when
one of their leads proves here to be dishonest and incapable of
working in a group effort without disruption.

I say there was a need to review the 495000 lines and you didn't do
your job. At a minimum, did you

(1) Fix all warnings issued by compilers when the code was ported to a
new platform?
(2) Set warning levels sufficiently high (to "lint" levels), or
(3) Use a lint-like tool to evaluate portability?

Or did you just watch the stuff run with your mouth open and declare
victory as long as nothing crashed?

Most auditing firms would find you erred IF you made the decision to
use C after circa 2000 for by that time two tools (Java and .Net) were
available for write-once run-anywhere, and not to use them in the
financial context, using instead a language with known issues such as
C, was bad practice.

If you worked for Northern Rock then as an employee under UK law you
are probably indemnified because under UK and American law, it is a
legal axiom that no matter how incompetent an employee might be, he's
not to blame for decisions that are made or signed-off on by
management. Nonetheless it appears to me that insofar as you worked in
"banks and insurance companies" your decisions helped to create a
mess, and I wouldn't pat myself on the back I were you.

 

[Phil Carmody]

You keep saying this, but:

* The dictionary disagrees with you.
* So does the dictionary you said agreed with you.
* So does every other dictionary we've yet been able to find.

And *more* importantly - common usage amongst users of the language.

If something that was full of errors couldn't be clear, then the
phrases 'clearly wrong' and 'clearly full of errors' would make no
sense.

Phil

 

[Phil Carmody]

The code table is different from the physical representation. If I
store an ASCII value in a 32-bit integer, is it no longer an ASCII
value because I didn't use exactly 7 bits?

Completely irrelevant question. It's ASCII if you don't use 7 bits
but use more instead; but it's not '8-bit ASCII', '16-bit ASCII',
or '32-bit ASCII'. If you think the scenario you describe above is
worthy of the name '32-bit ASCII', then you have the right to show
support for Schildt's wording, but also have the right to be considered
very weird indeed.

Do you also have signed and unsigned ASCIIs?

Floating ASCIIs?

Complex digraphs?

Phil

 

[spinoza1111]

And *more* importantly - common usage amongst users of the language.

If something that was full of errors couldn't be clear, then the
phrases 'clearly wrong' and 'clearly full of errors' would make no
sense.w

"Clearly wrong" means "the statement or text x that declares 'the
statement or text y is wrong' is true and it is clear". It does not
EVER mean that "the statement or text y is clear and wrong".

In mathematics, clarity has only one meaning, and that is validity.
Mathematics considers the clear to be true and when it doesn't seem to
be so, mathematicians like to make it so. A good example would be the
expansion of the natural numbers to include zero, negative numbers,
rationals, reals and ultimately complex numbers.

To the Romans, zero was unclear and assertions about it false or
meaningless. Then arithmetic was further refined and now it's clear
what zero does.

"The square root of -1" was unclear and now is clear.

A proof in mathematics cannot be clear and wrong. This is because at
the point of failure, it is unclear.

A clear and true statement corresponds to a unique meaning (cf
Husserl) where a "meaning" is the concept or idea which lies outside
the text. An unclear statement, even in ordinary parlance, corresponds
to > 1 meaning such that they can't all be true.

p & ~p, 1+1=3, have no correspondence with meaning.

 

[Colonel Harlan Sanders]

The half-educated and the auto-didact have an unwarranted faith in
dictionaries. Are you aware that there were no dictionaries in
Shakespeare's time, Milton's time, or John Bunyan's time, yet they
were better writers of plays, poetry and prose than anyone alive
today?

Yes I am. And equally relevant, are you aware that Marilyn Monroe had
6 toes on her foot?

blah blah blah

So after weeks of asserting that the OED supported your definition, on
my proving it did not, now "Real Men don't need no dictionaries".

Just as "Real Coders don't need no C Standard".

What a transparently self-deluded fool.