Re: Comment on Emacs Lisp Introduction

[Fren Zeee <frenzeee@gmail.com>]

On Mon, Aug 2, 2010 at 8:26 PM, Stephen J. Turnbull <stephen@xemacs.org> wrote:
> Fren Zeee writes:
>
>  > I am not complaining. Its just that I dont have it, and need to find
>  > source matching this old executable, which means that it will take
>  > some time to search for the particular freeze or tar file if it is
>  > still on the web.
>
> Huh.  Why are you so focused on "old"?  If you are an historian, you'd
> better enjoy such searches because that's what historians do.  If
> you're a programmer, though, old versions are of no particular
> interest unless you already know them well enough for forward
> differences to have meaning to you.

It appears that software is developed both outside-in and inside-out
at various times and the oldest most primitive version has the
simplest structure, and least features. IMHO it would be for
educational purposes, not just historical.

For a professor like you, who knows a lot more, perhaps, can start
anywhere in the field, but not for a newbie like myself.

> If you're interested in a particular function, M-x disassemble.

Show me an example where disassemble would help in understanding the
code. I think it gives VM code, not lisp code.

Give a walk-thru concrete example of several functions that illustrate
how disassemble would help understanding the code.

> Otherwise, why do you refuse to take the advice to use a modern
> version?  The basic architecture hasn't changed since the GNU rewrite.

OK, so give me a diagram which shows the scanner of LISP interpreter,
and in addition some comments for writing the primitives.

> Functions that are more complex now have become that way for a reason;
> those reasons are worth studying.  Many functions are *not* more
> complex in themselves than they were then, but have become simpler
> because they delegate subtasks that have increased in complexity to
> other functions.  It's easier to work backward from modern versions which
> are well-developed based on better abstractions (cf. Michael Stokes'
> comment about Green's Theorem: "It is trivial.  It is trivial because
> the concepts have been well-defined.  That definition took decades."
> -- or something like that, I don't have _Calculus on Manifolds_ handy).

I know what you mean but you cant stuff that into the mouth of a
newbie before knowing something of vector calculus.

Franz Xe

It would certainly help if I can get some debug sessions from you and
the other gentleman who gave me the initial advice on playing with
emacs using its own debug and gdb, a typescript file.