Freeman Dyson in the
talk linked above discusses the notion of pre-cellular pre-Darwinian
evolution where genetic material was passed around like tweets . He
points out that the evolution of knowledge has now returned to that
asexual promiscuous replication .
My number of Titter followers more than doubled when I
Tweeted
I ridiculously optimistically tweeted in response :
One reason why , when a path to my PhD in the psychology department at
Northwestern U closed , I refocused on
APL
was that I felt that for people ,
anyway : me , to
express the concepts
necessary to abstract and grok the brain's information
flows , would require a population fluent in an APL level executable
notation . The defining feature of APLs is that the
Array
or lists of lists , generalized from matrix algebra , is the
fundamental object . The notations are
generally as or more succinct than traditional mathematical notation .
I have lived my adult life in
notecomputing
environments I've created , first around 1984 in original "flat" APL ,
then , around 2000 , in Arthur Whitney's
K
. Upon moving to Teller County , CO in 2005 , I put in the extended
intense months necessary
for me to
create a reference counting dynamic object
vocabulary and the essential recursive algorithms for operating on
lists of lists and the essential naked lists of chars , integers , and
floats in Ron Aaron's
Reva Forth which was itself undergoing rapid
honing by a number of brilliant programmers scattered around the globe .
Arthur
Whitney's work has been the template and caliper for much of mine .
However , I came to decide that his 3 list dictionary or namespace
structure
( symbols ; values ;
attributes ) had the difficulty of referencing attributes from
values or vice versa . Motivated by
MinnowBrook
2011 , I added a cell to the headers of all CoSy dynamic objects to
allow a link to an object's attributes or
metadata . This would itself
generally be a dictionary .
While I use
4th.CoSy every day as my log , my
Shopping/ToDo list and my Reference
( passwords , account #s ,
etc ) list , I've not been able to steal the devoted weeks of
time necessary to convert the overall dictionary structure .
All
the family accounting is done in K.CoSy and will be for quite
some time .
I
was hoping to get the transformation done and a new "release" for this
NewsLetter . But that was terribly optimistic . Simply getting back
into the code is daunting . I have done some useful work tho , both
conceptually and in detail . I've started keeping a programming
log separate from my personal log to better organize the chaotic jumble
of experiments and mini-scripts which fill the first 1100 lines of my
log window .
A defining characteristic of
CoSy
environments since their earliest days around 1980 has been the mixing
of lines intended to be executed by the computer with ones just
intended for human consumption . In current
4th.CoSy
tapping
F6 will execute a
line in Forth displaying any output generated ; tapping
F7 will execute the line expecting
a
CoSy
object
or nothing to be generated , and
displays it as
CoSy
output . For instance , the last line executed in the screen shot ,
highlighted in blue transposes the result of the previous lines which
catinates a list of indices for the 3191 lines in the
text
window to the lines of text to produce a crude list so I can scroll
down and see my daily log for 2012 began at line 1127 . That was
executed with
F7 . Some
lines above , you'll see
nil DMP
.
DMP is a pure
Forth word which outputs the first 8 cells of a definition . That line
was executed with
F6 , then
F5 was used to insert it in the
text .
Note
that the great
Slava Pestov's early project ,
jEdit , fills most of
the
screen with , it happens , 7 windows open on various of the Reva Forth
code files making up
4th.CoSy . It would be great to
eventually have such an interface to
CoSy , but I really want to
interface it to HTML .
4th.CoSy
won't really be
CoSy
until normal definitions are saved within the dynamic root dictionary ,
R .
It's obvious there is a tremendous
amount , ie ,
several dedicated
me-months , left
to do . That's why I posted the current status as a
collection of useful APLish algorithms in Forth .
Reva
by itself is an open modern Forth with significant libraries for a
number common tasks . If some youngsters get interested , there's a lot
to be learned and contributed . I only worry that I won't be able to
keep up , that is , some paths will become dominate which I think are
IbM , not
optimal .
I plan to tweet
relevant announcements
with the tag #
4th.CoSy
. I am adding Disqus threads to new windows I create or revisit
as I have here .
I think
Yahoo could have done a much better job of
upgrading their mail groups into contemporary social networks , but
they
haven't .
I have also started a Google+ Community which may have its uses . If anybody knows of better options , let me know .