CGI not novel, instant CAI gratification is garbage (was: JDEE/CGI/flashcards ...)

> From: Tim X <t...@xxxxxxxxxxxxxxxxx>
> > Are you referring to tricks with CGI itself, or just putting a strong
> > UI-transaction-based client/server application behind it?

> Neither - I'm talking about developing some new original application
> of CGI. ...

Then you're talking nonsense. CGI is just an interface to run
server-side applications. CGI is not an original application itself and
cannot be an original application in any sense.

I repeat in different words: What I do (except for the beginner's
tutorials) is set up server-side applications that run in client/server
UI-transaction mode: The user fills out a form in the client (browser)
program, submits the contents of the controls in the form across the
network to the server, the application (on the server) processes the
form-contents and performs some requested task locally on the server
(possibly making additional net connections to sub-contract some of the
work elsewhere) and returns a new form that shows the results of that
transaction and offers new options for the next transaction. (I've
skipped the part about encoding/decoding the form contents and
including MIME header and HTML framework on returned page with form.)
There's not much novelty in the basic CGI operation per se, and HTML
presents a restriction on what a form can contain and how it can be
used in ways that are universal across all browsers including those
without JavaScript or other active client-side software beyond the
basic emulation of the various controls allowed in HTML forms. All the
interesting work is in designing the application to work usefully in
such a client/server UI-transaction mode, and designing the form to lay
out the data in a way that is easy for the user to understand and
navigate. (Yeah, I admit my SegMat demo, the very first CGI demo I ever
wrote, has too many radio buttons which specific flavor of demo to run
before finally having the StartDemo button. If somebody wants, I can
break the form into several pieces, duplicating the HTML FORM framework
and the submit button and each of the general mode checkboxes in each
form, including just a small related group of radio buttons in each
form. Would anybody request that? In the 4.5 years since I wrote that
demo, not one person has made such a request, so that's why I haven't
bothered with the work of that tiny change so-far.)

One slightly illegal idea I've had recently is to apply a face over the
top of Google Groups broken-beta to fix the many bugs and broken-design
it has in user interface and basic operations. Whereas the original
Google Groups had a simple tree view that for really large threads
tended to exceed the 80-character limit and break lines into pieces
making it nearly illegible, at least it *did* show a true tree view in
ASCII-text which was, for not-too-large threads, very useful in
navigating followups and skipping sub-threads. The new Google Groups
Broken Beta has no connecting lines, just indented subjects, making it
very difficult to visually navigate, and it has a serious bug whereby
clicking on anything in the indented (not tree) view takes you not to
that particular article but to the first article in that group modulo
ten plus one (i.e. if you click on article 65, you get taken to article
61 instead, and you have to use your browser's local edit search
feature to search for "65." before you're at the correct article). My
idea is to return to the original Google tree-view but to make it avoid
going too deep and make the expansion/hiding of sub-trees selectable
just like in the directory-tree view in MS-Windows and Linux/Gnome, as
well as fix the bugs so that if you click on one article from tree view
you go to that article not first-of-group, and also fixing another bug
that was introduced in GG2, if you click on "view as tree" it goes to
the very top of the tree whereas GG1 showed the whole tree but set
cursor on the particular article you had just come from, I'd revert to
GG1 in that behaviour. If I developed this just for my own use, it'd be
a lot of work just to help one person, but if I allowed others to use
it, I'd probably get sued by Google for copyright infringement or
somesuch, or my ISP would be blocked from all Google access and when my
ISP's admin found out I'd lose my account, so I'm pretty much
restricted to doing all that work just for myself with no way to show
it to anyone outside my apartment.

> You are not going to get someone to see the wonders of your
> applications if it takes hours/days/weeks. You need to get their
> interest in seconds and minutes.

Unfortunately that instant-gratification mode of evaluation is not
suitable for CAI. A student can't learn the vocabulary of a foreign
language, and a pre-school child can't learn how to read and spell the
thousand most common words, or even the first twenty, in seconds or
minutes. It takes drill over a long period of time to reinforce
short-term memory and build it into long-term memory. (And the
statistical results of working with my algorithm show clearly that
there are actually not two but three different spans of human memory:
Short-term like your immediate train of thought, medium-term from a few
minutes to a couple days, and long-term of months or years. My software
effectively bridges both the short/medium barrier and the medium/long
barrier in one unified system. There's no way to demonstrate the
short/medium transition in less than ten to twenty minutes, and there's
no way to demonstrate the medium/long transition in less than a week.)

So how do I get the point across to potential employers/educators that
my method really works and my online algorithm really implements the
working method and it's available to demonstrate to prove I'm not
making this up?

By the way, IMO one of the major problems with both textbooks and
computer-assisted instruction (CAI) nowadays is that the only thing
anybody with money cares about is flashy
graphics/images/photos/animation, nobody with money cares about whether
the software (or book) really teaches anything to students. They
evaluate new software, or books, by looking at screen-capture images,
or leafing through the pages, for less than one minute, getting an idea
how "pretty" it is, and using that as the sole judgement for deciding
whether to use that book or program in schools. That's one of the major
reasons evolution is taking a hit: It takes longer than two minutes to
understand evolution by natural selection, whereas it takes less than
two minutes to say "goddidit" and sing the first verse of some hymn to
fire up the emotions. Johnny can play video games, but Johnny can't read.

(Breaking long reply again here...)
.