[Foundation-l] Forking the Wiki

Magosányi Árpád mag at bunuel.tii.matav.hu
Sun Jan 2 09:38:31 UTC 2005


A levelezőm azt hiszi, hogy Andre Engels a következőeket írta:
> Wikipedia and Wikimedia were once created to start an encyclopedia.
> Now we have more, we have wikis for dictionaries, books, source
> material, citations, multimedia files, biological species, victims of
> 9/11... Why?
> 
> The positive way to say it would be to state that we are extending our
> reach. But is that the case? Looking at the history, I have the
> feeling they are all weak compromises between deletionists and
> inclusionists. Some wanted to delete dictionary definitions, others
> wanted to keep them. No agreement. So we make a special wiki for that,
> and both can be happy.

I was thinking about the same. This is what I am thinking about it:

Each Wiki page can be viewed as describing a knowledge scheme. The
additions I have seen so far are of two types:

* "Structural" schemes, describing some kind of structure of other
schemes.

* "Informational" schemes, describing some information.

Examples of structural schemes are the species and the contents of
wiktionary. These both structure other schemes which are normally
entries of wikipedia. 

Examples of informational schemes are source material, citations,
multimedia files.

The structural schemes cannot describe the structures they want
to describe because disconnectedness. It is clearly shown in
wictionary: you can find multiple instances of the same word
across multiple language dictionaries (I mean you can find
words in each language in each language dictionary), and
these words are described independently of wikipedia,
which means much more inferior descriptions, in most cases
no description at all.

And what about informational schemes? We have a different place
for pictures. And we do have very nice, artistic graphics there.
But I would guess that there are much more graphics in wikipedia
as illustrations. Furthermore, content in common is also structured
according to some schemes: author, topic, type, etc.

One more thing to consider: the actual source of content is just
a technical detail.

Conclusion:

Wikimedia's approach to describe structures of schemes effectively
and aligned with human ways of thought needs further work.
So far only technical details have been considered, but the key
is the presentation approach.

Proposal:

There should be some well-established rules which widely acted upon.
The number of rules should be low, and they have to reflect sure
knowledge about human ways of thought. Some which might need
consideration:

* One concept should be described in only one place

* Relations of concept should be easily described. Every kind of
relations. (See wordnet for a very important experiment with this
in a very strict area: linguistics.)

* There might be separate wikis for separate approaches to gathering
 information, but at the end of the day knowledge should be presented in
 a coherent manner.

Now you can call me inclusionist if you want to miss the point.

-- 
GNU GPL: csak tiszta forrásból



More information about the foundation-l mailing list