Compression - So, what's involved in...

Author

So, what's involved in...

Jules Gilbert

2004-06-24, 6:55 pm

Ta da...

aiTRAN, version 1.6, runon 6/18/04.
The aiSOFT and aiTRAN programs are the property of SymbTech. Also,
the
design, architecture and underlying engineering of this product is
the
property of SymbTech. This software is proprietary and is an
expression
of closely held trade secrets and/or intellectual property belonging
to
SymbTech. Phone (xxx) yyy-zzzz to report misuse of this software or
of
the related copyrighted materials. This software and the related
mater-
ials are Copyright 1994.

/* tuple on; Show parentheses encompassing a tuple. */
/* infix on; Show parentheses as supplied to operators. */
/* args on; Show parentheses at function invocation. */
/* hybrid off; Many character class tokens, like ".AND.". */
/* cshow off; C-character's in notation or as literals. */
/* exam off; Show various kinds of debug information. */
/* echo off; Echo aiTRAN program text. */
/* compile off; Compile aiTRAN program to C-based runtime. */
text: 1636 (string) + 10328 (structures) = 11964 bytes
dynamic: 0 (heap) + 0 (stack) = 0 bytes

[1] /* now running in FreeBSD. */ exit;

Now what's involved in putting this on the net?, so that someone with
a browser can try a few commands. The interaction is based on C, and
makes use of things like fgets as well as other nasty stuff.

Obviously, several routines will have to be completely rewritten.
Also bad is that I wrote it intending it for "dump glass tubes", such
as the much vaunted ADM3-A (in other words, not even as advanced at
the VT-100 -- and THAT's OLD!)

So, this may be a impossibly messy task -- maybe a complete rewrite is
in order.

But for projects such as coding up experimental compression systems,
genetic experimentation, or producing rules describing the actions of
an unknown compressor, based on instances of inputs and outputs,
aiTRAN is pretty good.

Right now the system does require gargantuan amounts of memory -- to
do any real work, you should 100M of RAM, and for serious projects
even 1GB of virtual memory isn't too much to have. So, maybe a
rewrite is in order...

Severian <severian@chlamydia-is-not-a-flower.com> wrote in message news:<lv35d0psjlu6ai1mtblbnlqa0u8oviuqrj@4ax.com>...
> On Thu, 17 Jun 2004 22:12:17 -0500, Earl Colby Pottinger
> <earlcp@idirect.com> wrote:
>
>
> Par for the course, ja?

Severian

2004-06-24, 6:55 pm

On 18 Jun 2004 05:57:11 -0700, julesg@financier.com (Jules Gilbert)
wrote:

>Ta da...

I have never been so impressed in my life!

>aiTRAN, version 1.6, runon 6/18/04.
>The aiSOFT and aiTRAN programs are the property of SymbTech. Also,
>the
>design, architecture and underlying engineering of this product is
>the
>property of SymbTech. This software is proprietary and is an
>expression
>of closely held trade secrets and/or intellectual property belonging
>to
>SymbTech. Phone (xxx) yyy-zzzz to report misuse of this software or
>of
>the related copyrighted materials. This software and the related
>mater-
>ials are Copyright 1994.
>
>
> /* tuple on; Show parentheses encompassing a tuple. */
> /* infix on; Show parentheses as supplied to operators. */
> /* args on; Show parentheses at function invocation. */
> /* hybrid off; Many character class tokens, like ".AND.". */
> /* cshow off; C-character's in notation or as literals. */
> /* exam off; Show various kinds of debug information. */
> /* echo off; Echo aiTRAN program text. */
> /* compile off; Compile aiTRAN program to C-based runtime. */
>text: 1636 (string) + 10328 (structures) = 11964 bytes
>dynamic: 0 (heap) + 0 (stack) = 0 bytes
>
>
>
>[1] /* now running in FreeBSD. */ exit;

Oooooooh. So impressive...

>Now what's involved in putting this on the net?, so that someone with
>a browser can try a few commands. The interaction is based on C, and
>makes use of things like fgets as well as other nasty stuff.

What are you blithering about? There's nothing nasty about fgets.

>Obviously, several routines will have to be completely rewritten.
>Also bad is that I wrote it intending it for "dump glass tubes", such
>as the much vaunted ADM3-A (in other words, not even as advanced at
>the VT-100 -- and THAT's OLD!)

You were targeting ADM3As in 1994? I guess we can assume your
stupidity is not a recent development.

>So, this may be a impossibly messy task -- maybe a complete rewrite is
>in order.

No shit, Sherlock. Of course, were it worth a green goddamn, you would
have rewritten it long ago.

>But for projects such as coding up experimental compression systems,
>genetic experimentation, or producing rules describing the actions of
>an unknown compressor, based on instances of inputs and outputs,
>aiTRAN is pretty good.

Sure it is. Show us some of its magickal results.

>Right now the system does require gargantuan amounts of memory -- to
>do any real work, you should 100M of RAM, and for serious projects
>even 1GB of virtual memory isn't too much to have. So, maybe a
>rewrite is in order...

I know of a few neurons that need to be rewired first.

--
Sev

Sponsored Links