http://lists.w3.org/Archives/Public/www-archive/2001Jul/0018
http://lists.w3.org/Archives/Public/www-archive/2001Jul/att-0018/logicerror1

Each of the nodes in the diagram represents one page on logicerror,
and each arc represents a link to another page within the site. The
site was very well knit together, even though it was large. Aaron put
a lot of effort into making sure that it was taftable, meaning that it
had a large operational navigational space for typical visitors to
click through.

But to me, the site was a failure. There was one major problem that
set it apart from Wikipedia, and it wasn't the size of the site or the
careful integration of the links. It was the fact that each page was
quite short and definitional. Wikipedia has a maxim that "Wikipedia is
not a dictionary", and logicerror showed why this maxim is very
important. You could click through a fascinating landscape, but along
the way you weren't getting much intellectual food to sustain you. You
had to, in a sense, derive it from the topology rather than the
content.

This, to me, is the biggest danger in setting up a personal wiki. I
call it the Logicerror Problem. Morbus Iff wrote recently that he is
planning to set up a personal wiki himself, called the disobiki, and
covers many interesting points related to the subject therein, but
doesn't touch upon the Logicerror Problem.

http://www.disobey.com/node/1866
http://www.disobey.com/wiki/

The main reason why I suspect Morbus doesn't broach this issue is that
for him it's not going to be a problem. He already has a lot of
content and data to load into the wiki, for a start, including huge
collections of various movie and subculture related ephemera, scans
and books and all kinds of things as long as the rights are
compatible. He's already writing stories on there, which are much
longer than a wiki dictionary entry, and full of great ideas.

So why was this such a problem for Aaron, and why would it be such a
problem for me? I'm not entirely sure, but I've been trying to model
it by thinking of what kind of content I would put on some kind of
personal wiki.

At the moment my main hobby obsession is astronomy, so I'd want to
document things typical of astronomy on a wiki. Setting up the
telescope that I got was kind of confusing in a couple of places where
the manual didn't explain what an extra part was for and so on, so
writing about that might help other users of the same telescope. I'd
like to publish my short summary of how to set up a polarscope, and
this sort of thing. Basically guides to things.

I'm not sure, however, how these would be taftwoven into a larger
site. On Wikipedia you'd have a category for telescopes, and you could
link the kind of telescope from the main telescope page, or a subpage
about reflectors. But you wouldn't host a guide on the page about the
telescope, it would be more close of a fit to a subsection about that
telescope. Of course a personal wiki does not replicate the
functionality of Wikipedia, and if you have material which would be
perfectly suited to Wikipedia then I suppose you'e better off adding
it there.

One of the potential articles in this category of things which may be
better added to Wikipedia is my idea about a richer description of M44
the Beehive Cluster. When I observed it yesterday evening, I found
myself wondering about the various stars and asterisms that I saw.
Afterwards I looked it up online, and found that there was almost no
information available of the sort that I was expecting. I've run into
this sort of problem before when researching astronomy. So I thought
it would be nice to provide a detailed map of the Beehive Cluster,
annotated with various interesting facts that I could find about it.

Even if this were published on a personal wiki, again the problem is
how to integrate it. An encyclopædic wiki contextualises an article in
terms of everything else which is described on the wiki, but a
personal wiki is not an encyclopædia, or at least is not best written
as one. To me, the context of the Beehive Cluster is my current
obsession with astronomy, so in a way I'd need an article on my
astronomy obsession. What I'm trying to think about is what
characterises an article within a personal wiki. We're not creating
articles to define abstract nouns, so much, but rather to talk about
things like enthusiasms, obsessions, social circles, possessions,
things that we're learning about, creative experiments, and so on.

Perhaps it would be possible to create a taxonomy of these different
article types. With Wikipedia, what you have is essentially one
encyclopædia, a very large one describing the kind of things that we
usually get in an encyclopædia. There may be many different *types* of
things described of in an encyclopædia, and you may for example
contrast an article about the political theory of civil war to an
article about a particular species of lichen. But on the personal
wiki, I think that the kinds of articles that I've been talking about
are divided at a different level. It's not the subject which has
different types, different categories, but the *mode* of the article
itself — the way that it's written, and the kind of use to which it
may be put.

If this be a useful model, then the greatest distinction between an
encyclopædic wiki and a personal wiki would be that a personal wiki is
a kind of federation or amalgamation of several different types of
wiki, the interactions between which may be quite poorly defined at
the moment since the personal wiki I assume does not have many
prototypes and studies into usage patterns so far conducted. The
problem is how to encourage taftability in a ecosystem which appears
at first glance to be very similar to the taftable site nonpareil, but
which may turn out to be substantially different.

On 15 April 2010 17:16, Sean B. Palmer <s...@miscoranda.com> wrote:

> I'm not sure, however, how these would be taftwoven into a larger
> site. On Wikipedia you'd have a category for telescopes, and you could
> link the kind of telescope from the main telescope page, or a subpage
> about reflectors. But you wouldn't host a guide on the page about the
> telescope, it would be more close of a fit to a subsection about that
> telescope

>So I thought

> it would be nice to provide a detailed map of the Beehive Cluster,
> annotated with various interesting facts that I could find about it.
>
> Even if this were published on a personal wiki, again the problem is
> how to integrate it.

Are you mixing two requirements Sean? Linking a page into the main
'flow' on subject A or B, then creating some form of hierarchy so that
a reader can find B or A given their view of the subject (not yours).

Hot links answers the first problem. IMHO you'll never solve the
second one to everyone's satisfaction. How about using
some tool to create a general index of the text, annotated
with the page, so you create an index rather than a toc?

Just a thought.

regards

--
Dave Pawson
XSLT XSL-FO FAQ.
Docbook FAQ.
http://www.dpawson.co.uk

> Perhaps it would be possible to create a taxonomy of these different
> article types. With Wikipedia, what you have is essentially one
> encyclopædia, a very large one describing the kind of things that we
> usually get in an encyclopædia. There may be many different *types* of
> things described of in an encyclopædia, and you may for example
> contrast an article about the political theory of civil war to an
> article about a particular species of lichen. But on the personal
> wiki, I think that the kinds of articles that I've been talking about
> are divided at a different level. It's not the subject which has
> different types, different categories, but the *mode* of the article
> itself — the way that it's written, and the kind of use to which it
> may be put.

On Apr 15, 6:16 pm, "Sean B. Palmer" <s...@miscoranda.com> wrote:
> Aaron Swartz has a site called logicerror.com, which was originally a
> kind of personal wiki. Most of the pages from the original site have
> gone now, lost in his great hard drive crash of 2004, but the scale of
> the thing is preserved in a survey I did of the site's navigational
> topology.
>

>
> Each of the nodes in the diagram represents one page on logicerror,
> and each arc represents a link to another page within the site. The
> site was very well knit together, even though it was large. Aaron put
> a lot of effort into making sure that it was taftable, meaning that it
> had a large operational navigational space for typical visitors to
> click through.
>
> But to me, the site was a failure. There was one major problem that
> set it apart from Wikipedia, and it wasn't the size of the site or the
> careful integration of the links. It was the fact that each page was
> quite short and definitional. Wikipedia has a maxim that "Wikipedia is
> not a dictionary", and logicerror showed why this maxim is very
> important. You could click through a fascinating landscape, but along
> the way you weren't getting much intellectual food to sustain you. You
> had to, in a sense, derive it from the topology rather than the
> content.
>
> This, to me, is the biggest danger in setting up a personal wiki. I
> call it the Logicerror Problem. Morbus Iff wrote recently that he is
> planning to set up a personal wiki himself, called the disobiki, and
> covers many interesting points related to the subject therein, but
> doesn't touch upon the Logicerror Problem.
>