Re: [LogiLogi-list] [Citizendium-l] Fwd: RE: Encyclopedia

My mistake, I sent the message to you directly, I sent it to the list
again now.

> Nice to hear you are working on something like this.  I think that 
> multiple-versions is the new frontier in encyclopedias, and what 
> citizendium primarily seems to be doing is maknig a moderated version of 
> wikipedia (and creating the editorial community around it.)

Indeed. 

> Your idea needs a bit of effort on the part where you select different 
> versions. Instead of voting, there should be a simple way to determine 
> which version you want to see, so that users could select a pre-defined 
> selection of articles depending on their current needs (e.g. a scientists 
> encyclopedia, a philosophers encyclopedia).  Each group would branch (fork) 
> off its own version tree, and edit it to make it specialized for their 
> needs, and they would have to do the work of keeping it up to date vis a 
> vis other branches.

I see what you mean, but a disadvantage of what you suggest is that 
it does not allow the same text to be used by multiple peer-groups. 
This is a result of the coupling of the selection of texts for an user 
to view and their branching into different texts. 

Physically copying and modifying a text is something different from 
deeming it usefull for a certain peer-group.

Actually the peer-group based voting system has a few more advantages 
over the physically forking of the content:
* it allows competition between versions within peergroups (and thus
  it makes disallowing editing by anyone but the owner no problem,
  opening the gates for real intellectual responsibility)
* the peergroups can be self-organizing (voting-power determined by
  votes from the peer-group on one's own contributions)
* peer-groups can start off small and still be usefull, and can be 
  layered (philosophers of history, choosing 'philosophy of history' 
  as their first peer-group, and 'philosophy' as their second. Then 
  if no rating is found for their first peer-group, the philosophy 
  peer-group's rating is used for determining the version they see).

> I would be interested in collaborating with your project, as it sounds like 
> it is on the right track.

You're welcome. You could join the LogiLogi Manta mailinglist for a
start: https://lists.sourceforge.net/lists/listinfo/logilogi-list

LogiLogi is written in Ruby (on Rails), and largely along the
principles beautifully described in the book 'Getting Real', by 
the award winning web-development company 37 Signals (with the
difference that LogiLogi Manta is Open Source). 

Rails is really a bliss for the developper, so even if you don't 
have any experience with it, you probably will enjoy learning it.

> BTW Have you heard of Kim Veltman, from the 
> University of Maastricht?  He is also working on knowledge organization, 
> and is a friend of mine.  I worked on one of his projects 10 years ago.

I have heard of him, but I never met him in person. Would be an 
interesting person to meet.

> I think eventually if we can get this demo right, we could then patch 
> MediaWiki also, or extend it to support multiple versions, something 
> Rational's Clearcase does very well.

I already investigated applying the peergroup-system to MediaWiki, 
and the other improvements that LogiLogi Manta brings, but it seems
very hard, at best, not just technically, but also in terms of the 
re-structuring of the articles required (which is why we started 
from scratch).

greetings,

Wybo

> I am a software engineer, btw, and have a personal non-professional 
> interest in knowledge organization that dates back 12 years.
> 
> cheers,
> Hasan Murtaza
> 
> >From: Wybo Wiersma
> >
> >Hasan I fully agree with what you write below about the advantages of
> >"according to"-settings. I already championed for simmilar features,
> >but Larry's choice for a classical encyclopedia-style review process
> >and MediaWiki as a platform makes a real combination of open-ness and
> >quality impossible.
> >
> >LogiLogi Manta, a small project funded by the Philosophy Department of
> >the RuG (a Dutch University) that was already defined and started
> >before Citizendium was announced will implement what you propose in
> >the form of peer-groups:
> >http://en.logilogi.org/MetaLogi/LogiLogiManta
> >
> >In LogiLogi results from new research (due to the peergroup system) or
> >articles for different levels of audience (thanks to to the section
> >system) are no problem either.
> >
> >Of course I understand why a classical review-process is chosen for
> >Citizendium, namely because it is known to and trusted by the target
> >editors audience. The problem is however that it's not optimized for
> >the web as a medium, as it neglects possibilities that the web offers
> >like "according to"-settings.
> >
> >And MediaWiki is the only viable choice if one wants to build upon
> >(all) content of Wikipedia. Nice, but MediaWiki is not designed with
> >combining openness and quality in mind, and adding those features
> >will take much longer than starting from scratch...
> >
> >Expected release date for LogiLogi Manta: around the end of 2006.
> >
> >Wybo
> >
> >> (Just as a google is 10 to the 100th power, a googleplex is 10 to the
> >> googleth power.  By similar reasoning, Just as a the wikipedia is a 
> >single
> >> enyclopedia, a wiki-plex is an ensemble of encyclopedias where every 
> >article
> >> has an infinite number of possible versions associated with it.  
> >(Anybody
> >> can define their own version for themselves.)
> >>
> >> A complex set of preferences could be defined that lets the invidual 
> >reader,
> >> select whose version of the encyclopedia they want to see, subdivided by
> >> section or individual article.
> >>
> >> For example
> >>
> >>
> >> # sample user configuration for the CZ
> >> select latest version by "cambridge university philosophy department" 
> >for
> >> articles on Philosophy
> >> select latest version by nasa for articles on space
> >> select latest version before 2006 by jimmy wales for articles on 
> >wikipedia
> >> select latest modifications by citizendium_group for articles on 
> >anything
> >> else
> >> #etc.
> >>
> >>
> >> With luck, the community will be able to regulate itself and there will 
> >not
> >> be an N-squared explosion of versions defined by everyone and their dog.
> >> One immediate benefit will be that the questions of fairness, 
> >censorship,
> >> and credentials can be avoided in the near term, and these issues can be
> >> pushed far into the future.
> >>
> >>
> >> Everything I have seen in these discussions about the CZ seems to center
> >> around clearing out the trolls, attracting new authors and defending the
> >> realm with a new class of editors, essentially creating a moderated 
> >version
> >> of wikipedia.  Doing the things I mentioned above would truly make it a
> >> citizens encyclopedia.  Another way to understand this is that allowing
> >> "private wikis" to coexist within the larger public wiki would be 
> >equivalent
> >> to creating a merchant class in a medieval society, and it would be just 
> >the
> >> thing to invigorate it with even more activity, and also solve all the
> >> problems related to the lack of differentiation available in a 
> >monoversional
> >> encyclopedia.)
> >>
> >> Hasan

-- 

::Student:
- History, Informatiekunde (computer linguistics, IR, webtech) and Philosophy
- Member of the Center for Metahistory Groningen (http://www.rug.nl/let/cmg)

::Free Software and Open Source Developer:
- http://www.LogiLogi.org, innovative system for cumulative, shared commenting,
                           publication and idea sharing: Web as it should be...
- ComLinToo, a computational linguistics toolset written in Perl
- Lake (LogiLogi.org Make), a make-replacement using makefiles in pure C++

::Being:
- In the world, go figure (http://nl.logilogi.org/HomE/WyboWiersma)