[racket] Liitin screencast tutorial

Just comes to my mind the times, not so long ago, when people had to push
all names into 8.3 characters and learn to write so that computers could
understand them. Sounds like metadata is yet another such constraint; you'd
rather not let it restrict you (atleast, off work), but you still wanted to
benefit from it.

Not that I could easily think of the solution. Atleast, if anything could be
automated, extracted from the data itself or the context, it should be done
so.

I guess the principle of moving forward with what we've got and improve it
as we learn more, is valid as always. Except that, rather than starting from
new everytime we learn something and abandoning the old (or instruct 6
billion people to update), we should keep the old way still valid and leave
traces from the progress.

I'm not sure if this is a good analogy, but... think about some data/program
out there now (it can be something found in the web or any native OS).
- It works/exists today
- It it very likely to work/exist tomorror, as well.
- It will propably work/exist day after
- On third day it's 50:50
- after fifth day you are better off reading from frog's guts
(sorry, I took the days from wheather forecasting :)

Now, how likely is that _any_ data/program out there now will just
work/exist after 50 years? Yet, most of the data and methods are still
likely to be valid, or atleat of value to learn from.

br, jukka


> -----Original Message-----
> From: users-bounces at racket-lang.org
> [mailto:users-bounces at racket-lang.org]On Behalf Of Greg Hendershott
> Sent: 07 March 2011 04:48
> To: Neil Van Dyke
> Cc: users at racket-lang.org; Shriram Krishnamurthi
> Subject: Re: [racket] Liitin screencast tutorial
>
>
> > (If one has ever had the misfortune of reading YouTube
> > comments, one might despair that we'll ever achieve even basic
> literacy by
> > 20th century standards, nevermind what we think is required for
> 21st century
> > expression on the Web.)
>
> i likd teh comant you maid hear ur ideaz r reluvent to my intresd
>
> btw my boss says lisp iz 4 hippies
>
> ____
> {tags: epistemology; normative taxonomies; decline and fall of western
> civilization; #lang lolcat}
> {safety mode: off}
> Find your friends by giving us the credentials to your email accounts
> -- it's free!
>
> On Sun, Mar 6, 2011 at 4:32 AM, Neil Van Dyke
> <neil at neilvandyke.org> wrote:
> > Shriram Krishnamurthi wrote at 03/05/2011 02:15 PM:
> >
> > Also, in addition to the ambiguous "people lie," I'd add
> > "ontological/knowledge engineering is hard."  I think that malice and
> > incompetence should be considered two separate problems that don't
> > necessarily have exactly the same solution.
> >
> >
> > The "and" was meant as a separator.
> >
> > http://www.well.com/~doctorow/metacrap.htm
> >
> >
> > Then I say that laziness, incompetence, and malice are three separate
> > problems.
> >
> > And I see that Cory Doctorow identified this with his unquoted third
> > problem, "People are stupid."  He's talking about competence at
> expression,
> > which, when generalized to Semantic Web, might include ability to
> > distinguish concepts like "has-a" and "is-a."  I saw in the
> early '90s that
> > pretty simple static object-modeling confounded eluded a
> surprising number
> > of people, even after they received some instruction, and
> similar difficulty
> > was found in the fields of Knowledge Representation and Knowledge
> > Acquisition.  (If one has ever had the misfortune of reading YouTube
> > comments, one might despair that we'll ever achieve even basic
> literacy by
> > 20th century standards, nevermind what we think is required for
> 21st century
> > expression on the Web.)
> >
> > These problems, and others, are not insurmountable.  TeachScheme and the
> > like will help.  Smarter approaches for the Semantic Web will also help.
> >
> > --
> > http://www.neilvandyke.org/
> > _________________________________________________
> >  For list-related administrative tasks:
> >  http://lists.racket-lang.org/listinfo/users
> >
>
> _________________________________________________
>   For list-related administrative tasks:
>   http://lists.racket-lang.org/listinfo/users