On 3/8/2012 9:32 PM, David Barbour wrote:
Bret Victor's work came to my attention due to a recent video,
Inventing on Principle
http://vimeo.com/36579366
If you haven't seen this video, watch it. It's especially appropriate
for the FoNC audience.
although I don't normally much agree with the concept of "principle",
most of what was shown there was fairly interesting.
kind of makes some of my efforts (involving interaction with things via
typing code fragments into a console) seem fairly weak... OTOH, at the
moment, I am not entirely sure how such a thing would be implemented
either (very customized JS interpreter? ...).
it also much better addresses the problem of "how do I easily get an
object handle for that thing right there?", which is an unsolved issue
in my case (one can directly manipulate things in the scene via script
fragments entered as console commands, provided they can get a reference
to the entity, which is often a much harder problem).
timeliness of feedback is something I can sort of relate to, as I tend
to invest a lot more effort in things I can do fairly quickly and test,
than in things which may take many hours or days before I see much of
anything (and is partly a reason I invested so much effort into my
scripting VM, rather than simply just write everything in C or C++ and
call it good enough, even despite the fair amount of code that is
written this way).
most notable thing I did recently (besides some fiddling with getting a
new JIT written), was adding a syntax for block-strings. I used <[[ ...
]]> rather than triple-quotes (like in Python), mostly as this syntax is
more friendly to nesting, and is also fairly unlikely to appear by
accident, and couldn't come up with much "obviously better" at the
moment, "<{{ ... }}>" was another considered option (but is IIRC already
used for something), as was the option of just using triple-quote (would
work, but isn't readily nestable).
this was itself a result of a quick thing which came up while writing
about something else:
how to deal with the problem of easily allowing user-defined syntax in a
language without the (IMO, relatively nasty) feature of allowing
context-dependent syntax in a parser?...
most obvious solution: some way to easily create large string literals,
which could then be fed into a user-defined parser / eval. then one can
partly sidestep the matter of "syntax within a syntax". granted, yes, it
is a cheap hack...
Anyhow, since then I've been perusing Bret Victor's other works at:
http://worrydream.com
Which, unfortunately, renders painfully slowly in my Chrome browser
and relatively modern desktop machine. But sludging through the first
page to reach content has been rewarding.
One excellent article is Magic Ink:
http://worrydream.com/MagicInk/
In this article, Victor asks `What is Software?` and makes a case that
`interaction` should be a resource of last resort, that what we often
need to focus on is presenting context-relevant information and
graphics design, ways to put lots of information in a small space in a
non-distracting way.
And yet another article suggests that we Kill Math:
http://worrydream.com/KillMath/
Which focuses on using more concrete and graphical representations to
teach concepts and perform computations. Alan Kay's work teaching
children (Doing With Images Makes Symbols) seems to be relevant here,
and receives a quote.
Anyhow, I think there's a lot here everyone, if you haven't seen it
all before.
nifty...
(although, for me, at the moment, it is after 2AM, I am needing to
sleep...).
Regards,
Dave
_______________________________________________
fonc mailing list
fonc@vpri.org
http://vpri.org/mailman/listinfo/fonc
_______________________________________________
fonc mailing list
fonc@vpri.org
http://vpri.org/mailman/listinfo/fonc