On Wednesday, 14 February 2018 at 09:39:20 UTC, Luís Marques
wrote:
It seems that someone once again rediscovered the benefits of
component programming, in the context of OOP, but (as usual)
without the more mathematical and principled approach of
something like ranges and algorithms:
<http://geekswithblogs.net/theArchitectsNapkin/archive/2011/03/19/flow-design-cheat-sheet-ndash-part-i-notation.aspx>
BTW, I just wrote my DConf proposal. I've been experimenting
with a different style of not-quite-OOP in a real project and
so far I'm really happy with the results. I've been making use
of Jean-Louis' openmethods.d library, as well as other D
features and techniques. The result is a quite nice balance of
simplicity, expressiveness and performance. I'm really looking
forward to telling you all about it :-)
"Flow-Orientation is about tackling complexity at its root cause:
that´s dependencies."
That's an interesting statement (from that article).
Seems we are getting closer and closer to modelling programming
in accordance with how the brain programs itself. Kinda of makes
sense really - nature seemed to work this all out a long, long
time ago.
Small units. Data just flows in and out. The units don't 'know'
each other. Their are no explicit dependencies between the units
themselves (hence neuroplasticity).
Instead, what's important, are pathways by which they can
communicate (concatenate) their input and output, and the
subsequent data flows that arise from that collaboration.
In nature, increasing mass (complexity) arises from simple
components.
If only that were so in the world of programming.
