[Wikipedia-l] Re: Manning's beta/stable comments

[Daniel Mayer]

On Friday 24 May 2002 12:01 pm, Manning wrote:
> Beta-stable: In the words of Mr Horse (from Ren and Stimpy) "No Sir, I
> don't like it". One of LMS's early ideas was that when an article looked
> good enough it would get moved to Nupedia. Problem with that was that
> no-one ever actually went to Nupedia.

My opinion on the reason why this particular idea of Larry's didn't work was 
because there wasn't <i>any</i> practical (or otherwise) integration of 
Nupedia with wikipedia <i>and</i> the two projects represented two different 
cultures and methodologies -- not because the concept was inherently flawed. 
If there is tight integration of this proposed functionality and it is made 
very clear to viewers of the static pages that an editable version exists by 
"clicking here", I think it just may work. Although to clarify, I think the 
most current editable version of articles should be the one displayed by 
default -- a user would have to click to view the reviewed version (the 
difference between the two versions also needs to be made obvious and 
intuitive). 

> The same will happen here, if we create a Wikipedia that is static (can't
> be edited) then no-one will go there. Even the people who are truly only
> here to browse will gravitate to the dynamic site, simply because it is
> possible that that site "is more current".

Interesting idea, but I think this only really applies to casual visitors and 
not to researchers or students who need reliable data. For example:

If I needed information about a topic I would put far more trust in a static 
page that had been reviewed by 10 people with trusted status than I would 
place on an "up to date" yet un-reviewed version. How could I trust the 
validity of the world-editable version if I were witting a report? What use 
would there be in citing and linking to the world-editable version if I can't 
have a reasonable amount of confidence that the information I am citing will 
be there when someone checking my work follows the link? 

It would reflect very negatively on me if I cited something interesting I 
found in a beautifully written and informative - yet world editable - 
wikipedia article on say the Apollo Moon landings, and then a reviewer of my 
work checked the link and instead of seeing what I saw, sees a totally 
unrelated and unintelligible diatribe about CARROTS. Or worse, the version of 
the article that the reviewer sees was rewritten by a pseudoscience freak who 
is highly skeptical of NASA's "claim" that the Moon landings actually took 
place. How is that going to make me and my work look to the reviewer? 

> To prove this I'd recommend creating a one-off Static version, putting a
> link to it from the dynamic site and then monitoring web-traffic over a
> period of 2- 3 months. Even after the site has been around for a while I'd
> be willing to wager it would still be a very lonely place.

Sounds like a reasonable.proposal -- I too am not totally convinced that any 
proposed "review process" can actually work. But my concern is that the size 
of our current contributor base is not yet large enough to make this 
practical. There doesn't seem to be enough eyes yet to keep reviewed versions 
of articles reasonably up-to-date. And there is of course a HUGE backlog of 
articles that haven't been touched by any editor for multiple months. But 
then, starting this process wouldn't hurt the current "unreliable" status of 
wikipedia articles and may in time lead to wikipedia being viewed as being a 
more trusted source of information (if only for the handful of reviewed 
articles).  

> If I am wrong, and the site does get a lot of traffic, well then we should
> spend the time and develop a proper management and quality control system.
> But until we have proved its viability, we would just be wasting effort -
> effort better spent on creating/editing articles.

I agree in part -- if we decide to do this, then we need to give it a lot of 
thought and not rush into things (best to do it right than to do it fast). 
But because this idea has such great potential, I think it <i>is</i> 
something we should investigate and give a chance. This is something that 
won't really get a lot of traffic at first - there needs to be a critical 
mass of reviewed articles before traffic to them in general increases 
significantly (even a hundred reviewed articles would be totally lost in the 
current database). 

If done correctly this could finally begin to make wikipedia a critically 
respected source of human knowledge that people can depend on and come back 
to over and over again.

-- maveric149