To a certain extent, there is meritocracy in technology. Particularly when
it comes to proliferation. If it is easy to bootstrap, hello world, hack
useful demos, then expect a warmer adoption-rate than, say, gcc development.
from working through deep problems, like a compiler.
Post by Loup Vaillant DavidI wonder what would be the best way to spread such ideas.
One can build a product, but it risks being just that, watered down and
co-opted by the mainstream ways of doing things. Nobody change the way
they think, they just have this neat new tool.
One can show a nice demo, show the potential, but then *someone* would
have to make some product out of that, or the idea risks being
forgotten for the lack of a killer app.
This hints at a third alternative: making a simple product out of the
idea, *not* releasing it, but instead teach students how to build it
from scratch âuse a class project or something. I believe that's how
compiler courses work: at some point the student really writes a
compiler, and will be able to implement a DSL in the field if there's a
need.
Frank could serve as a basis for such a course. There's at least a
compiler course, a rendering course, a networking course, and a GUI
course in there.
Loup.
Post by Dan CookBrett makes it clear that his aim is to share ideas and get people
thinking about things like direct manipulation and better
visualization of mental models, etc. His demos were made to
demonstrate, rather than do make cool tools. This is similar to what
was said previously about why VPRI's Frank was not made code-
accessible.
On the other hand, Brett has a lot of really great articles that
encourage new ways of looking at software, especially shifting from a
mechanistic to a more humanistic view of what could be. His demos are
there as examples to show that it can indeed be done.
Having the source code would certainly be useful to see how he did
stuff; but I think he wants people to think more than just consume
his tools and call that good. Perhaps he's learning from history,
like how SmallTalk was a great idea that was supposed to transcend
any specific language or tool, but when it became commercialized, it
was reduced to just that, and the greater value of the big picture
was lost.
The trouble with Brett victors work is he hasn't actually published
any of his libraries (that I know of) or code.
Post by Dan CookI have seen Lively Kernel, and I think it's a great step in the
right direction, though it does nothing to make the language
extensible (it is all just JavaScript).
Anyone interested in the intersection of freely modifiable
language and freely modifiable environment should DEFINITELY look
at some of Bret Victor's work (worrydream.com). His "Drawing
dynamic visualizations" example is (among others) a great
example. If like to see the same technique applied to processes
(code).
I am working on a POC for a freely-modifiable
language+environment (though there's nothing complete to show for
it yet) at: https://github.com/d-cook/Objects
Post by John CarlsonHas anyone see programming by demonstration environment that is
extensible in demonstration mode? Lively Kernel perhaps?
What would be the best PBD environment for aspect-oriented
programming by demonstration?
_______________________________________________
Fonc mailing list
http://mailman.vpri.org/mailman/listinfo/fonc_mailman.vpri.org
_______________________________________________
Fonc mailing list
http://mailman.vpri.org/mailman/listinfo/fonc_mailman.vpri.org
_______________________________________________
Fonc mailing list
http://mailman.vpri.org/mailman/listinfo/fonc_mailman.vpri.org
_______________________________________________
Fonc mailing list
http://mailman.vpri.org/mailman/listinfo/fonc_mailman.vpri.org