i'm happy to let you know, that i've found the entry point.

which entry-point?

i'll tell you later.

first of all: you've maybe noticed a small change.

i became a little ironic.

this is to offend the hairsplitters.

and to offend garbage.

-

garbage is not something wrong that someone tells me .

garbage is an 'expert', who tries to teach me.

and who himself has not understood.

and who don't take back his didactical tone.

even if you show him that he is (al least maybe) wrong.

that's garbage.

-

human brain.

there are facts.

and the ability to process facts.

-

from now on, i'll treat with respect only those which treat me with respect.

this is cause i've to focus my energy on working.

thus i've not enouth energy for gentleness.

-

The Entry Point.

*year*:

well, it is 1958.

John McCarthy.

there i'll begin.

maybe i've to go back more.

will see.

i've already started.

-

*language*:

C or C++

i code in C++, although i use in the begining only C subset.

i'm not exactly sure if i keep that.

*functionality*

i know nearly nothing.

the reader-concept.

the macro a little.

backquote a little.

garbage collector. i've an idea what this is, cannot imagine that it is
so complicated. claim and release memory. but will see.

and of course all that knowledge from *using* many languages. but
nothing indeep, mostly fuzzy in memory.

-

the good thing: i can assimilate everything. no need to develope it.

i'm undereducated. that's why i can decide quicker what to assimilate
and what not.

how that, without knowing it? well, i'll know what i need. i give
someone the abstract description of what i need. and he tells me: this
an that you can use. i say: explain me 'with apples' the differences.
then i decide: that one.

you see: i'm depending on experts which are able to explain 'with
apples', which is nothing more than abstracting an essence down to a
general common sense level.

but before i loose 'virginity', i'll try to build a lisp without looking
at another code.

i've no imagination of how to build such a simple system. the first
coding has started. seems really difficult.

i continue now

-

to be able to appreciate other peoples work, which i can later include
with a small decision into my developement.

to see which  of the constructs used in lisp resulting 'out of the need'
and 'out of clear human sense' or 'the natural flow of things', and
which parts are the really difficult to realize creative works.

-

The Spirit of Lisp.

Also "Artifical Intelligence Programming: case studies in common lisp" chapter
23 has a simple lisp compiler.

I personally found Lisp a bit overwhelming at first. So I initially Learned
scheme, which is a lot smaller, but has most of the same concepts.

me: Mr. Salman Rushdie, can I?

SR: yes my son.

me: thank you!

"there is a fault in the specs. do not adjust you minds."

hehe!

life is to short.

the middle way:

realize the problem.

trying to solve ( a few hours or days )

look at the solutions.

feel the respect for the solutions.

assimilate the best.

-

The decision: which is the best?

With the entry-point 1958, 100.000 people will have 100.000 different
results in 2002 (or better: 2003).

But there is only one.

-

The Spirit of Lisp.

Why, with so many other newsgroups to ann...I mean languages to
learn, did you pick Lisp?

Why not Scheme?  Perl?  Modula-3?  Dylan?  Clipper?  Icon?  VRML?

Why not Fortran?  It's even older than Lisp, and has a much more
active following including many interesting people, who, I am sure,
will be more than happy to pan-fr...I mean spend their time
enlightening someone with such an obtu...I mean interesting approach
to learning as yours.

Think about it.  If CLL is such an unfriendly place, perhaps that says
something about Lisp itself.  Perhaps people who get to know Lisp well
become the kind of people who will flame, toast, roast and otherwise
barbecue innocent seekers after knowledge.  (My mouth is watering
already.)

It has become clear that CLL is not worthy of your attention.  I don't
think Lisp is either.  It simply has too much academic baggage for
someone with such a pristine intellect as yours.  I still think you
won't be really happy until you find your true home in some
Microsoft-related newsgroup.

A. Church: The Calculi of lambda-Conversion. Princeteon, 1941

will change to C++.

i'm to lazy to implement with C constructs.

but will only use few things of OO.

buttom-up-from scratch design of 'foreign thoughts' is not my strength.

additionally, i'm not so good in *creating* 'facts'.

much better in *evaluating* and *interconnecting* facts.

-

so, as i stuck, i looked at some C++ code.

but much to complex implemented.

i've set an limit of 10 pages for the core.

-

can anyone point me to a design guide?

or tell me the sequence of the functions i've to implement?

-

e.g.:

(please forgive me the naming. it follows strictly my 'internal'
terminology)

i start with:

typedef unsigned long atom;

struct node
{
   atom *first;
   atom *rest;

?

(i'm so 'virgin' in mind, i've *never* implemented a list)

-

i'm unable to continue my simpe C++ database project.

Hello, i hope you are fine.

Please read my post fully (you've quoted me 'off-context'). I'm writing
a lisp from scratch, to see how difficult it is. So i'm in topic.

I'm just a little confused, it is really difficult to find the entry
point to start coding. I know lisp only somehow fuzzy. I know cons,
pair, first, rest etc.! but i don't know exactly what they do and how
everything is represented in memory.

That why i've just printed some documents, which have more graphical
representation. so i think i'll understand until tomorrow.

i have problems with implementing a lisp interpreter in C++.

i'm a lisp novice.

My 60 days trial for Borland C++ Builder expires tomorrow.

I don't know if it is allowed to try for another 60 days.

If not, i must set up another C++ system.

I use windows OS, but i'd like to switch to a linux system one day. so
i'd like to have a cross-platform C++, but a nice one, with an
integrated developement environment.

can someone suggest me one?

I don't want to deal with complicated installations-sequences.

this is stuff for freaks that have nothing better to do than to manage
complicated things.

i have this need, too. I compensate it with 'spider', this is something
like 'pacienda' (this card game on windows). *very* complicated.
Althougth i do not always win, after a few game *i* am the winner.

when i see those lousy installation read me's, i shout: why does this
guy (the manual writer, which is in general the programmer or one of the
programmers) writes all this down, so thousands of people have to do
that, instead of providing a simple install.bat or setup.exe or so.

this small detail is one reason that companies win in competition.
Whilst the others exhaust their potential customers with dealing with
details, the others make things simple (not better, just simpler) and
win the competition.

this reminds me somehow... lisp.

whow.

in topic again.

-

Get the essence:

Try:

struct Node
{
    union NodeData
    {
        int Integer;
        int Symbol;
        int FunctionId;
        // Other data types go here...string, float, etc etc
    };

    enum NodeType
    {
        INTEGER, SYMBOL, FUNCTIONID, LIST
    }

    NodeType Type;
    NodeData Data;
    Node* pHead; //CAR
    Node* pRest; //CDR

--
Justin Johnson

during writing i hear a little cat screaming outside.

after 5 min i went out, find it.

i think 3 to 4 months or so.

i take it with me, i give it something to eat.

but this cattle don't eat.

just: miaaah, miaaah.

i cannot keep the cat here. impossible.

so out again.

takes me all the concentrations.

i feel like a little bad now.

conscience crisis.

espsecially as the 'miaaaa' was so loud, that i was afraid the
neighbours wake up.

water helps.

more concsience crisis.

-

conscience. the word is not logical.

science.

con.

con science.

anyway.

-

the reader. i've finished the skeleton.

the specs are really complicated written.

intuition was right.

things became clear due to coding.

It should be possible to shrink the reader-algorithm-description to
about 60% to 70% of its current size.

whilst keeping information value.

whilst increasing precision.

this way people can learn quicker.

-

and of course "The Scary Readtable" is fully correct.

http://groups.google.com/groups?selm=amsq4a$cr...@usenet.otenet.gr

how can anyone believe it's not?

funny.

-

the cat is still outside.

i'll got to play with it a little.

and then home sweet home.

-

lisp interpreter is easy to implement.

the challange is the raising complexity.

-

Borland C++ Builder 60-day-Evaluation is very gentle.

it shows since 2 days: 0 days remaining.

but it still continues to work.

am i illegal now?

i don't know.

it looks like a marketing thing.

-

write a lisp, instead of discussing in c.l.lisp

more fun.

--
Justin Johnson

and i got very much fun of how such a simple thing can be documented so
complicated.

it's really a game.

-

i was to early happy.

the Windows ME maschine crashed again.

after restarting: "thank you for evaluating borland"

-

my motivation is against zero now.

is there any C++ compliler for Windows OS which generates executables
for Linux / Unix / MAC ?

-

Mozilla seems to be overloaded with all this babbling in c.l.l.

i loose the saved news-content for the third time.

any suggestions for a good Newsreader, which is available for linux, too
(binary compatible)?

For macros, just get eval to macroexpand them (macroexpansion
is just treating the macro as a function applied to the
unevaluated arguments), before evaluating the macroexpansion.

The only problem closures and macros is keeping track of the
correct environment, but this is well understood.

Garbage collection is more problematic.   You can get
around the problem entirely by writing your interpreter
in a language such as Java which supports garbage collection,
but that of course is cheating.   A simple mark and sweep
collector is straightforward to implement in C, and the
algorithm is well known.   If you do this, you need to
remember to use an explicit stack, as if you naively
do something like

List append(List x, List y)
{
  List z = NULL;
  while (x != NULL) {
    z = cons(first(x), z);
    x = rest(x);
  }
  while (z != NULL) {
    y = cons(first(z), y);
    z = rest(z);
  }
  return y;

More sophisticated techniques (e.g. generational
garbage collectors) are probably overkill in a learning
exercise, unless you're learning about garbage collection
per se.

retrieved new eval keys with my *old* account @ borland.

installed.

works.

seems finally that borland tolerates  it