Alan Kay talk ‘Is it really "Complex"? Or did we just make it "Complicated”?’. The talk was given some 9 years ago, but, this is the first time that I’ve seen it:

youtube.com/watch?v=ubaX1Smg6pY

My highlights:

51:21 suppressing the present to see different ways to address the future

51:35 “semaphore” is a “bad idea”, pseudo-time is superior, but never got used over last 30 years

54:10 Nile Graphics Language by Dan Amelang is a dataflow language and is very small. Expresses how to define shapes using pixels in only 45 lines of code.

54:51 - Bret Victor at VPRI

57:59 all compositing rules, including alpha-blending, is expressed in 95 LOC

59:16 whole graphics system, equivalent to what is needed on a PC, is expressed in 435 LOC

1:01:00 - Code has to be transformed into CPU-speak

1:01:38 - we need a “language transforming language” [[pt] At present, I use the term “t2t” (text to text)]

1:01:58 - OMeta2 [[pt] OhmJS is an ancestor of OMeta2]

1:02:58 Using OMeta to make Nile, parser 130 LOC, LLL 730 LOC

1:06:00 TCP/IP expressed in 160 LOC

1:20:24 POLs == Problem Oriented Languages [[pt] I use the term SCN - Solution Centric Notation]

1:37:37 hardware is just software that is crystallized early [[pt] I say it another way: software is just soft hardware]

On Interoperability of sync services for software designed to be Local-First: tonsky.me/blog/crdt-filesync

Here’s an article I wrote on a visual dataflow language modality where the execution of the graphical dataflow dynamically constructs the dataflow graph that is itself executing. (It’s similar to how Elixer functions can output AST for the function instead of the result and one can alternate between AST via quote/unquote) create.vi/programming-in-abstract-g-7e970f123f3c?source=friends_link&sk=a53e37d328f1fbc2a8020c94c1f4c0b0

youtube.com/watch?v=ieqsL5NkS6I