Re: Rene can't handle AoA's Success

On 21 jun, 15:01, "Wolfgang Kern" <nowh...@xxxxxxxxxxx> wrote:
randyhyde wrote:
Sure, only those who understood what they read in the CPU manuals
get a chance to join in the game, regardless if they also read
Knuth, Pispalis, Dunteman, and/or similar books.

And they have a much greater chance of understanding what's in the CPU
manuals if they first read a book like AoA. That's the bottom line,
whether you like it or not.

That's just what 'you' believe in.


....and with him a whole lot of ppl, including those who determine
what's in the curriculum of a CS course and what's not. Granted, they
may not use AoA as the teaching material of their choice, but you can
be pretty damn sure that all of them start with a class or two in a
HLL, mostly an imperative one like JAVA or C++.

Randy's attempt to teach newbies will fail because his prerequisites
are HLL-knowledge which is a not required detour for true ASM coders.

It may not be a "required detour", but it is a very good one to make.

Call it whatsoever it is a detour.


whenever someone states that "real <insert random group of ppl>
(don't) do <insert random attribute commen to aforementioned group>",
they do tent to make some gross simplifications.

You may excel at knowing what each machine instruction does on a micro
level, but I've seen your code samples around here and from a
*programmer's* point of view, they're not something to be especially
proud of. Oh, you count the cycles *real well*, so that your lousy
algorithms and terrible data structures consume as few cycles as
possible.

Your readability argument again ?
You attack things like KESYS algos/data structures you never saw ?


The same goes for you; you attack HLA, and more notably, AoA without
ever having read the book completely even once.

In the meantime, a student in a HLL programming course,
using techniques they've learned there, write HLL code that compiles
to something that is better than what you've written.

Perhaps, but ONLY in terms of 'your' demands on maintainability
and readability, but never ever in terms of (size/speed) performance.


This may come as a surprize to you, but Randy isn't the one who's
grading those learning asm with AoA, not by a long shot. In fact, a
greal deal of them are using AoA for self-study and move on to more
bare-bones works later.

You say "Randies [sic] attempt to teach newbies will fail..." yet
there is nothing "Randies" about this.
When I was in *elementary* school, the universities were teaching
HLLs before assembly language.

And you think this meanwhile also anchient order have to last forever.


....cos it's that order that has been proven to be the most effective
one.

If you consider "true assembly programmers" to be the people such as
yourself out in the world -- self-taught, undisciplined programmers --

:) disciplined to what? to Your personal view of the universe ?
or would it be more 'discipliened' to follow all the C-convention ?


Try common good coding practices and standards among the software
engineering community for that one.

well, okay. But despite the small number of people such as yourself
(and your apprentices who "still make money writing low-level code")
who've figured out what the machine instructions do, in the real world
(using tests to measure knowledge),

Who do you think work for all the single chip market
(consumer electronics,phones,toys,household equipment,...)
where engineers have to design the hardware(CPU) 'and' the program,
C-programmers ? :):):)


How about embedded JAVA or muC(++)?

I'll guarantee you that the average person will learn assembly language
*much* better and be able to produce far more sophisticated programs
if they first learn a HLL.

What got that to do wiht each other ?
Complex programming needs detailed knowledge of the matter, that's it.
Granted: HLL-coders might use prepared C-libs and don't care the 'how',
but prebuilt code modules are also available for plain ASM.


Because, if you choose assembly language as your first programming
language, you have to contend with 2 aspects simultaneously, namely
1.) you have to master how to choose and work out algorithms to
achieve a goal and 2.) meticulously put the instructions and operands
in exactly the right order to implement said algorithms. A HLL is more
lax/easy WRT to that second aspect, allowing beginners to focus on the
first aspect. Once they have enough coding skills and experience under
their belt, you can bother them with the exact machine instructions
that make up said algorithms.

Sadly (and it's obvious in this newsgroup whenever a self-taught
programmer posts some code), programmers who learn assembly language
as their only (or first) programming language tend to exhibit some
*very bad* programming habits.

Yes 'very bad' for your limited ability to read code which looks
different to all you ever learned.


That's a very bad cop-out; this is not as much about code readability
but adhering to common coding practices that have stood the test of
time.

It's people like this who give assembly language a bad name.

No, it's the alive creativity which makes ASM interesting.
It's the boring 'traditional' HLL-styled coding which seems to
be burned in your brain, makes ASM look useless.


....and yet Randy took it upon himself to devise teaching method to
speed/ease up the process of learning ASM. Hardly something that seems
like making ASM look useless. But go ahead; if you think you can code
enteprise apps or databases smaller and more efficient, faster with
asm than with java resp. sql, by all means be our guest.

.

Relevant Pages

  • Re: What do I do with Art Of Assembly?
    ... > But, in that regard, AoA is clearly signposted as "previous HLL ... dogma and not valid as part of an answer about the best way to learn asm. ... I learned asm as my first programming language. ... > if someone intends to write Windows applications then learning DOS is ...
    (alt.lang.asm)
  • Re: A good thread there
    ... > you also know in German with the German version. ... > want to learn assembly programming, ... and apply those same notions to assembly language programming. ... You seem to assume that people who already know some HLL are incapable ...
    (alt.lang.asm)
  • Re: What do I do with Art Of Assembly?
    ... > could eventually learn something with AoA. ... People already knowing _HLL_ programming could eventually learn ... "unlearn" it all again to learn ASM... ...
    (alt.lang.asm)
  • Re: a turbo assembly language
    ... >>that ASM is no more time consuming than an HLL. ... >of the programming SoftWares has gone wrong. ... language representation is, by definition, attempting to create an HLL. ...
    (alt.lang.asm)
  • Re: Rene cant handle AoAs Success
    ... They're willing risk a *lot* of money publishing books like AoA ... that they have published dozens of x86 assembly language books over ... There is no question that I could have written my code in a HLL like C ... The thing they share in common is "programming". ...
    (alt.lang.asm)