[WikiEN-l] Page freezing on "good" articles

Daniel Mayer maveric149 at yahoo.com
Thu Oct 27 23:58:51 UTC 2005


--- "steven l. rubenstein" <rubenste at ohiou.edu> wrote:
> The only thing I can imagine is this: when a page has reached this state it 
> is usually through the hard work of a few editors (and I am not trying to 
> deny the contributions of countless other people).  I suggest those editors 
> save that version of the article as sub-pages to their user pages.  If 
> there is ever major vandalism of the article, or if it seriously degrades 
> over a long period of time, those editors have a point of reference 
> (without having to go back through the edit history) of when they thought 
> of it as "done."

What we really need is to pay and/or bribe and/or beg a developer to fix the
performance issues of the 'rate this article version' feature. Then the highest
rated version of an article would be prominently linked from the top of the
most recent version of the article. It would also be neat if one could indicate
whether or not particular votes were useful and if the most recent version of
the article is still as good or bad (a cur diff would be displayed so
comparisons would be easy). 

Logged-in users would have the option of displaying the highest rated version
of an article if and when available. But I think it is *very* important that we
never have that as a global default; the big thing wikiwiki has going for it is
the instant gratification factor and vandalism should not be hidden behind a
higher rated but older version. 

Also, at the bottom of each page there should be an automatically-created
citation line that gives the exact page version you are looking at. I see there
is already a 'Permanent link' feature in the sidebar - but who the hell is
going to know what that is for? Citing a Wikipedia article is currently useless
w/o this feature (anybody checking such a cite could not be sure which of the
many versions on any particluar day were actually used). 

Another idea: I'd love to have the ability to rate other editors and trust by
proxy what users whose opinion I trust think of editors they have rated
(trusting by double proxy may also be useful). No data on any one person's
rating would be public; the data would only be used to filter out edits made by
trusted members of the community so that more attention can be paid to those
who are still unknown or who are known but not trusted. Much of the major data
processing for this feature could be done on the client's computer like this:
*raw RC/watchlist data would be downloaded to your computer
*your user ratings and the user ratings of the people whose opinion you trust
are also downloaded (this would be incrementally updated as needed)
*your computer would do all the sorting by user name
Of course, there will still be lots and lots of extra reads and writes with
this feature unless a completlely separate p2p editing client is used. 

The current system is just not scaling very well. If we want to stay open, then
we *need* to invest in some serious software and database improvements. All of
the above will have a very significant database hit, but I think it is all
worth if since it should push us toward the production of better quality
content.

-- mav


	
		
__________________________________ 
Yahoo! Mail - PC Magazine Editors' Choice 2005 
http://mail.yahoo.com



More information about the WikiEN-l mailing list