John Scholes' derivations in http://www.dyalog.com/dfnsdws/n_derive.htm
in response to an email on Dyalog's Yahoo group shows the essence of my
goals in respect to computer language . I bet you have never seen any
computer language that looks anything like it . John is the retired
chief designer and author of Dyalog APL , and I consider to have the
best aesthetics of anybody .
There's a link at the bottom of the page to get Dyalog's font if the page doesn't look like this :
The little circle which would be better centered between its left and right arguments is compose which essentially glues its arguments so they act as one . The umlauted squiggle is commute and is one of the brilliances John added to Dyalog before anybody else recognized its necessity .
The whole exercise is to see whether a function like , as a basic example mean or average
, which uses just one argument , albeit more than once , can be
restructured to reduce that to a single reference , so it can be elided
leaving a pure functional ( verb ) string .
That starting expression for mean can be read as "divide the sum across the argument by the count of items in it" .
If you like like algebra , you can love APL . I think I gave you a copy of Dyalog they gave me in Montreal .
I've really been taking far too long too long on my article/proposal
for a reason to go to the second Heartland climate conference to be
held in March .
I'm hung up on making sure I get the conversion to integrating over solid angle patches on the sheer , right .
The text up to the equation I need to finish is at http://cosy.com/Science/HeartlandProposal0811.htm . I guess the integral will end up being over differences in 4th powers of temperatures in all directions .
The first example to compute is a disk , black towards the sun and
totally reflective on the night side , because that would determine the
maximum temperature any object in orbit could be simply heated by
the sun .
Finally , the same unfolding , to use the term that was popular with the catastrophy theory people when I was at NU , needs to be made for wavelength .
That should give the "null hypothesis" equation to calculate the effect of non-uniform albedo and non-uniform spectrum .
I look forward to your comments .
-- Bob Armstrong -- www.CoSy.com -- 719-337-2733 --