Excuse me?  You hope you do not have to read a thousand pages?  Why?  Are
  you illiterate or very hard of reading or something?  I have not kept
  record of how much I read every year, but it must be at least 12 000
  pages of nonfiction alone.  For the past 25 years or so, I have set aside
  2 hours a day to study or just read about something _unrelated_ to what I
  need to learn or work with, and probably spend even more on work-related
  reading.  This is in my view not a lot.  Learning to read and type fast,
  efficiently, and accurately is a requirement in modern society.

  With only two hours a day, the Common Lisp standard should take you no
  more than a month, with plenty of time to think about what you read, take
  notes, including deferring stuff for a later pass.  That is not, again in
  my view, not a lot to require of someone who wants to become a serious
  user of a programming language.  30 years of wisdom and insight in the
  Lisp community is made available in only 30 days.  Of course, it will
  take much longer to make use of all this, but at least you have been
  through it and have thought about it and provided a large number of
  "hook" on which you can hang various information you get when you ask
  people, and you can make connections that people who simply lack the
  knowledge cannot.

| I see emphasis placed on 1100 pages... r5rs is 50...  sorry if this is a
| problem...

  It means that R5RS is a short story you might chuckle or shudder at
  before you forget it while ANSI X3.226-1994 is a novel that enriches your
  character.

| really I want to implement the one that is hopefully simpler and
| hopefully have an underlying system flexible enough to include both.

  Why the hell do you want to implement it?

| when I was younger I used to use big books for bragging rights (I would
| say: I read this, and point to a 1000 to 1500 page book, but really I
| don't want to read that much...).

  Oh, so _that_ is the reason these idiot books were printed on cardboard
  with 36-point type.  Idiot bragging rights!  Good thing I never even
  thought of buying any of those.  My >1000-volume personal library seems
  to average about 400 pages to a volume, and a _large_ fraction of my
  books are more than 1000 pages thick, but they are printed on thin, high
  quality paper, not recycled cardboard, with fairly small print, not at
  all suitable for the fourth-grade reading level of most "computer" books.

| luckily the spec I got claims to be a simplified version, I do hope so...

  No, it is not, the HyperSpec (ANSI standard) is a standard.  Maybe Common
  Lisp is just not for you.

| oh yeah, I have no access to a real lisp machine so info about how one is
| put together might be helpful.

  Maybe you just need a Common Lisp environment that has already been built
  for you?

cl and scheme are similarly related.

Try any of these books:
In order of increasing difficulty

"The Schemer's Guide" (Available at www.schemers.com)
"How to design programs"
"Scheme and the art of programming"
"Concrete abstractions"
"Structure and intepretation of computer programs" (Esse é foda ;-)

Don't worry about GUI and database stuff right now. That's trivial, and as
soon as you master basic concepts such as data and procedure abstraction, or
common program design techniques, learning those things will be very easy.

Read this article, it might help you:
http://www.fernando-rodriguez.com/high-school-computing.html

Vai fundo! ;-)

  I strongly disagree.  Scheme destroys the minds of programmers much more
  and much worse than BASIC is rumored to do.  In order to learn to use any
  other programming language, a Scheme-trained programmer has to forget all
  of Scheme and relearn everything from scratch.  This is doable, but such
  a horribly inefficient way of doing things.

| "How to design programs"

  I have to admit, though, this looks like a _great_ book.

| "Structure and intepretation of computer programs" (Esse é foda ;-)

  I no longer think this is a great book.  It teaches differences where no
  basis for comparison has been laid.

``Lisp in Small Pieces'' by Christian Queinnec

Regards,
--
Nils Goesche
Ask not for whom the <CONTROL-G> tolls.

It so happens that I've been programming a little game in Corman Lisp
with my daughter.  It is true that Corman Lisp does not provide an API
like CLIM.  It does, however, come with some code that wraps the
windows message pump with some CLOS classes so you can simply
write some methods and let the generic function dispatch handle the
details.  This is much easier than interfacing to the raw windows API.

So if you have some familiarity with how windows programs work
(vis-a-vis messages, windows, device contexts, etc.) the Corman lisp
API will work just fine.

Perhaps you think other languages have not tried to mimic CL?
Or perhaps they had unlimited success?
Or maybe that Perl, Python, C++, C# don't incorporate any of
the features that give Lisp its power?
Or that their implementations aren't brittle?

I even once self belonged to the doomed group of people believing to
be able to produce an perhaps useful mini operating system. I still
believe one man is able to produce something like this, but not
without having spent some time with planning.

Regards,
Julian
--
Meine Hompage: http://julian.re6.de

Regards,
--
Nils Goesche
Ask not for whom the <CONTROL-G> tolls.

please don't call me illeterate.

by "simplified" I had meant: not massivly long...
I have before read long books, ie: my sql book was about that long, but I
took about 2 months...

I like smaller time frames, ie: a few hours...

  My education is from the University of Oslo, Informatics Department,
  which brags that its senior staff invented Simula and object-oriented
  programming.  They thought, for many years, that their students were not
  hampered by, indeed benefited from, learning Simula in their introductory
  courses.  Industry, students, etc, disagreed strongly.  Now they teach
  Java.  At the University of Bergen, the same story played out.

  For some reason, I do not generally trust people who would be seriously
  financially or prestigeously damaged if the argument they make turned out
  to be untrue, which is why I tend to push the prestige button on people
  to see if their stake is personal.  Some, like you, tend to take this
  very deeply personally, but prove to be untrustworthy because of it.

| There is no particular difficulty in moving on to other languages and
| other styles of programming.

  There is ample evidence here and elsewhere to the contrary.  That you say
  something does not make it true.  When you jump out to contradict me,
  there is reason to believe it not even true to begin with.  All this crap
  about writing your advisors and now your department proves that you lack
  the integrity and intellectual honesty that would have made you useful
  int his forum.  However, as long as you can be seriously hurt by being a
  retarded prick in your department, it is likely that you manage to fake a
  reasonably conforming student and obey those rules that are necessary to
  get your degree, but none others, just like that line about your boss
  that you claim people misunderstood.  So when you say something that
  simply contradicts something I have said, we have reason to believe that
  (1) you would be hurt if the truth I have spoken came out, and (2) your
  inability to deal with arguments leads you to attack the person and want
  to ambarrass and make unpleasant noises devoid of contents.  Instead,
  just provide valid and useful counter-information.  I do not think you
  can do that any more than ozan s yigit can, but feel free to change my
  mind -- I will if you provide different facts.  That is the difference
  between the two of us.

  <http://www.norvig.com/21-days.html>
  _Teach Yourself Programming in Ten Years_

Pretending that you can learn anything meaningful in an hour's worth
of reading is just silly.

ok, I could drop the idea of trying to support both, and maybe think about
it later.

  Then you are at such a disadvantage in contemporary IT that you should
  work really hard to improve your reading speed.  Double it, then double
  it again.  A regular paperback page of prose (say, something by Tom
  Clancy or John Grisham) should take no more than 15 seconds to read with
  no loss of comprehension, and the ISO standard for Ada should take no
  more than 5 minutes per page, on average.

| in my mind one can blame english more than text, as it is the mass of
| english like semantics that I don't like...

  As long as you do not like it, that dislike will always stand between you
  and mastering it.  Make up your mind to like it, work hard to get really
  good at it, and you will like it.  This applies to all things you want to
  learn well.  If you cannot like something, leave it behind and find
  something else that you can.  Life is too long to spend it on things you
  do not like or become good at what you have to do.

I really don't like reading fiction books either for similar reasons I
don't like long books... I can just watch movies...

luckily at least I can read code at a much faster rate than I can read
text, I don't really know why.
often when reading code I use page down comparred with scrolling, but text
is the opposite.

a little off topic, I once tried to write a small description of lambda
calculus (where I was trying to explain it in a much more compact way than
a book I was reading, which took a whole chapter...):
{
lambda calculus is a system of substitution. you will have a "lambda
expression" that will perform operations by being applied to expressions.
each will have the symbol lambda followed by some variable, this may be
several levels deep (here I will use L to denote lambda). after this there
is an expression (for which what is being combined with is substituted
into).  consider: Lx.x2 , this may be combined with some other expression
by enclosing it in parenthesis, then it is combined with something: say
dog, in which case you have (Lx.x2)dog which converts to dog2 , I will use
-> to indicate this process. this works because dog was substituted in for
x. each lambda term will have an associated variable (which whatever is
substituted in for in the expression). after this the lambda term
disappears (as it was used). these are combined in a left to right order.
if the variable associated with the lambda term does not exist, then
nothing is substituted in. if the variable exists within the expression and
is part of a lambda term then the variable is referred to a being bound
within the expression. variables that are part of a lambda term but do not
exist within the expression are unbound, and ones that are not associated
with a lambda term but are in the expression are referred to as free.
consider: LxLy.x and LxLy.y, in the first y is unbound, and in the second x
is unbound. consider each is combined first with dog and then with cat.
((LxLy.x)dog)cat -> (Ly.dog)cat -> dog, and ((LxLy.y)dog)cat -> (Ly.y)cat
-> cat. when multiple layers of parenthesis are involved the inmost are
evaluated first, then it moves outward.  consider if these were used as
true and false (true=LxLy.x, false=LxLy.y), then a statement could be made
if=LaLbLc.(((a)b)c) that could select between terms based on the condition
of a. thus: ((((if)false)dog)cat) would evaluate to cat, and
((((if)true)dog)cat) would be dog.  consider: LxLy.xz, in this expression x
is bound, y is unbound, and z is free. free variables are a hairy mess in
my opinion, as it is always possible for the variable to adversely effect
the expression (except if say it were combined with 0, then z would be
irrelevant).  lambda calculus can be quite useful.

  So was I.  Thosuands of hours of hard work cured it completely.  Dyslexia
  is the result a persistent problem in how reading and writing is taught,
  not a disorder.  All students benefit from a very different way to read
  and write than the stupid ways that only accidentally produce people who
  can spell correctly and read fast.  In brief, words are really images
  made up of strokes.  I consider the individual letter approach to those
  strokes vastly superior to the ideographic scripts.  If you have not
  learned to see words as images, not as individual characters, by the time
  you are ten years old, it takes an _enormous_ effort to fix it, but trust
  me: it is worth it.  Nothing beats the written word's ability to save you
  time in learning from somebody else's experiences, skills, and wisdom.
  Just make up your mind to like it, then work hard to actually like it.

    So was I.  Thosuands of hours of hard work cured it completely.

On Mon, 01 Apr 2002 20:30:49 GMT, the world broke into rejoicing as
"cr88192 sydney" <cr88...@hotmail.com>  said:

Getting an OS working involves a whole lot of fiddling around figuring
out how to interface with:
  a) Disk drives;
  b) Graphics cards;
  c) Keyboards;
  d) Motherboards, CPUs, memory buses

Keeping up with the regular "churn" of hardware changes means you'll be
liable to need to modify details on an annual basis, and you'll have to
very tightly specify what hardware must be used.

After all, if you wrote a nifty display manager for an nVidia RIVA last
year, you'd have found by now that the model you supported isn't
commercially available anymore.

If your goal is to continually struggle with the continual hardware
changes, then that's the nature of your project.  OS makers have been
known to "punt" and basically say "Nope; we don't support hardware other
than this tiny limited set."

For instance, you need not bother trying to use Plan 9 on a system with
an Adaptec SCSI controller; they don't support that at all.  And it
won't work with nVidia video hardware.

You'll be left with that very sort of challenge, of, on the one hand,
having to pick and choose some tiny set of compatible hardware, and, on
the other, regularly rewriting parts of the system to ensure that
there's _some_ current hardware on which your OS can run.

The alternative: Use Linux or FreeBSD as the OS kernel, and XFree86 as
the graphics "kernel."  You can thus avoid writing much if any kernel
code, and can run on the regularly-extended set of hardware that they
support.

The question is then of how deeply to start trying to have the system
"be Lisp."  

The "rather deep" choice would involve replacing Unix init with a Lisp
program that basically populates a system that from there on down looks
like Lisp, perhaps with an XFree86 process alongside to provide
graphical services.

  Typos are not caused by dyslexia.  Dyslexia is not even about typos.
  But I probably type faster than I should, because I try not to waste too
  mcuh time on trivial things.