Re: [stack] Concatenative Research

> I am still considering spending a Phd researching this area.

>  Motivation
Good, the motivation should be stated up front.

I fetched the cited http*tunes*, which I found I had already previously
started reading. And this is 'how' it reads for me:
] 1. John was born ...; has size X shoes; has ...
] 2. Mary was born...;has size Y shoes; has ...
...
] 37. Peter was born...

So by that stage I asked my self WHAT IS THE MOTIVATION?!
Where is this leading?
Haven't these people heard about what we learned in 60-70's
*Top down design*, goal directed design.
Computing is complex, and its tasks need a central/controlling plan, with
the goal always in sight/mind - to keep focus.
So I looked back to the TOC, and found the "Conclusion", which I hoped would
 state the goal/motivation.
The 'Conclusion' section was missing.
Why do you people want to torment me?!

> My interest stems from a gut feeling that the compositional model is
> more elegant than the applicative model.
'gut feeling' ??!!
AFAIK doing a PhD is not like 'I think I'll try a new flavour candy'.
What is it with this "get your self a full featured browser and
 have a *BETTER EXPERIENCE*" feely-feely US talk.
Have y'all/any read eg. Backus' ca. '78 analysis of HOW/WHY
<compositional model is better than the applicative model>.

And crucially we must define better.
I define it as 'getting me more results, with less pain/effort' --
in the long run, for the type of tasks that I do.

The items listed under the 'Motivation' heading are good/interesting
but need to be written as paths to pre-defined goals.

Have you ever dealt with law people?  Very strange thinkers.
But I've learned from them: this bloke wrote that "I shouldn't vague-out".
We techies, conditioned by math-studies to not explicitly spell out each step
of the inference chain, often leave too much unstated and fuzzy.

> I have the impression that variables are a difficult concept to grasp.
Our/techie problem is that if/why psychological/soft 'impressions' are valid,
must FIRST be established.  Ie. our goal must be to satisfy HUMAN requirements.
The goal must first be determined. The technical details of how to achieve
the supposed-to-be pre-defined goal [ie. what's discussed here and in the
other literature] is just detail. We are continually starting new journeys
without having defined our destination.
Like 'surfing is going nowhere, while waiting to fall off'.

> We spend years teaching students about variables in math class, and even in
> college many students have difficulty understanding that (lambda x. x) and
> (lambda y.  y) are identical.  This is certainly a double-edged sword -
> variables are a powerful reasoning aid - but I am nonetheless interested
> in glimpsing a world free of variables.

Admittedly psychology/cog-science is not hard science.
But we need to state [up front] that eg. the fewer variables/names/concept
that need to be carried/remembered the better.
And eg. that being able to read 'linearly' [with no forward references] is
less of a mental load. And the reasons why the successive refinement method is
economical and effective. Etc .....

I've only recently come to realise how economical [hence efficient] the:
A -> B -> C -> D -> ... -> GOAL; style of programming is.
And I'm annoyed that it's been hidden from me for decades by the unwillingness
of those who were able to advocate and explain, to lead the way.

A trivial example of this cascading/filtered programming style is:
 List all files in sub-dir ->
 less than N days old ->
 which contain the string <string1> ->
 and the string <string2> ->
 ...and the <stringN> ->
 goal: to find that file that you read last week about XYZ.

For *nix-boys this is a one-liner.
OTOH one-lining is immature look-ma-no-hands like.
And if we can separate the unfortunate baroque and irregular syntax of *nix,
we see the powerful/economical multistage-data-transformation programming
method which has proven a success over decades.

So, take the basic proven idea and clean it and refine it, is my advice.
But first make or get a ready-made justification of well defined goals.

Or otherwise just admit that human cognition is too subjective and fuzzy,and
we are just browsing around hoping to hit on something that seems to work.


== Chris Glur.