[Wikipedia-l] Do we really need a Sifter project?

[Daniel Mayer]

Ulrich Fuchs
>I do not beleive that this is a very good idea.
>You can do this, of course, the GNU FDL allows 
>it. However, I feel like Wikipedia should be cannibalized 
>that way, sorry. 

How would it be cannibalized? We are talking about creating a stable version 
of the articles. Most people, for example, don't install and use software in 
a production environment that is from CVS. Most people wait for a stable 
release of the software. Wikipedia is in a constant state of change - it is 
like CVS. Nupedia would be a stable distribution of that content that has 
been checked by experts. 

>I agree that stable and accepted articles are 
>important to have (for quoting and so on). I do 
>not agree that defining those versions must be 
>done by "people with baccalaureate degrees in 
>the subject area". It would be far more important 
>to get these experts *Writing* instead of editing.

They already are. I have a  baccalaureate degree in biology but that doesn't 
mean I'm staking my reputation on anything I write in the biology section as 
'maveric149.' But if I am checking facts in a biology article submitted to 
Nupedia I /will/ be staking my reputation on the factual accuracy of the 
article. As an added bonus Wikipedia gets an article whose facts have been 
checked by somebody who should know what they are talking about. There is no 
such stamp of approval on Wikipedia articles now. 

>I am not in favour of a system where a lot of 
>people drive thousands of articles to a certain 
>(excellent) state, and a few experts get the merits 
>by selecting the articels, making some smaller 
>copyedits and then calling that the "real" 
>encyclopaedia, implicitly stating that the Wikipedia 
>is not serious at all.

Having a "checked by" attrib is hardly giving all the credit to the Nupedian. 
The complete article history will also be linked from the static, checked 
version on Nupedia. And Wikipedia is the content development area - that is a 
serious and very important thing. Without it there would be no content. 

So people will have a choice between the stable version which has been checked 
by verified people with credentials or the more up to date version which may 
have glaring errors or omissions or outright false information inputed a 
second before they view the page.

I'm the type of person who likes to use stable distributions. 

-- Daniel Mayer (aka mav)