Re: New release of bookmark patch

[Beartooth <beartooth adelphia net>]

On Mon, 19 Jul 2004 14:55:46 +1000, Peter Harvey wrote:

> I'll have to admit I don't actually study FCA myself (I do constraint
> programming research). The main idea behind FCA as it was presented to
> me, and as I have used it in Epiphany, is that a 'concept' can be
> described by a set of attributes (eg. breathes, has four legs, has
> teeth, barks) AND by a set of objects (eg. Fido, Lassie and Snoopy).

Hmmm ... I haven't gotten around to studying the FCA website yet -- still
trying -- but on further thought this sounds very much like an approach to
what a librarian would call cataloging -- with an intriguing twist :
instead of a Sisyphean attempt to catalog the whole web, you simply
catalog one set of bookmarks.

I'm not quite sure whether the cataloging method depends on each specific
set, but it sounds like it (i.e., what you're really developing is a
meta-method).

Nor am I clear whether whether each individual catalog of a set *learns*
in some way what to catalog where (perhaps in some such way as some
spamfilters learn what each user considers spam), but it seems as if it'd
have to.

If I'm right about that, there's a vast assemblage of extant wheels for
you not to re-invent -- some because you can have them for the asking, and
many because they'd be third (or fifth or nth) wheels in your application.

If you're anywhere near a university that grants the MLS or MLSIS, you
might ask whether the cataloging professor is well thought of; if so,
arrange to have coffee with that person to talk about your work in her/his
context.

If that option isn't open, try looking at

http://listserv.acsu.buffalo.edu/archives/autocat.html --

perhaps after looking around

http://ublib.buffalo.edu/libraries/units/cts/autocat/

Caveat : Autocat is a *very* active list, with thousands of subscribers
and hundreds of messages per day. Otoh, many of the subscribers will be
fascinated by your undertaking, and eager to help any way they can -- and
among them they know most of what there is to know about arranging
information in intelligent and useful ways.


> http://www.dsl.uow.edu.au/~harvey/code_epiphany.shtml is a little out of
> date, but has examples and screenshots of how the menus end up arranged
> by my patch.

	I haven't gotten to that yet, either, though I have it bookmarked.
 
> Only thing I can say here is that the order in which the groupings are
> presented is out of your control. The only thing within your control is
> the groupings of bookmarks.

> So ... the exact location of bookmarks and submenus is out of your
> control. You are still forced to hve an alphabetic arrangement within
> menus (to the extent that it doesn't interfere with the groupings of
> bookmarks).
> option.

Again, that sounds *very* like a form of cataloging -- and of course
therefore pre-eminently welcome. Do take a peek at Autocat -- you can even
post without subscribing -- and expect vast helpful interest.

-- 
Beartooth Autodidact, curmudgeonly codger learning linux