[Foundation-l] Stable versions live on de.wp

[Ting Chen]

Nathan wrote:
> Oh right, that piece of data is there. Thanks! The other pieces of
> information are more important in determining whether it would work
> elsewhere, though.
>
> Nathan
>
> On Tue, Nov 4, 2008 at 5:57 PM, Gregory Maxwell <gmaxwell at gmail.com> wrote:
>
>   
>> On Tue, Nov 4, 2008 at 4:46 PM, Nathan <nawrich at gmail.com> wrote:
>> [snip]
>>     
>>> how many articles have been flagged,
>>>       
>> [snip]
>>
>> Follow the link!
>>
>>
>> http://toolserver.org/~aka/cgi-bin/reviewcnt.cgi?lang=english&action=overview<http://toolserver.org/%7Eaka/cgi-bin/reviewcnt.cgi?lang=english&action=overview>
>>
>>     

Hello Nathan,

I tried to forward your question to the german mailing-list to ask an 
expert for the answer. But for some reason I don't understand my forward 
didn't come through. So here my answer.

Philipp Birken, who is a member on the board of the German Chapter and 
who was the organizer of the development of teh stable versions made a 
very detailed and interesting talk on the Wikimania in Alexandria to 
this topic. The talk vastly changed my opinion on the stable versions 
(maybe you know during my election to the board I said I was still very 
reserved to the stable versions thing). Philipps conclusion was that he 
would recommend every big language version (that is with article counts 
more than 100,000) use the stable versions.

To my personal opinion I think stable version is especially useful for 
vandal battling. If you have say 80% of your articles checked, you can 
see very easily which article was changed in the last night and not 
checked again. Especially useful is this because you are even aware of 
the changes that are made on articles which no one had put it on his 
watchlist. With a few hundred enthusiastic voluntiers (I have no doubt 
that the english Wikipedia has them) this is a managable task. Sometimes 
there were lags on the checking. For example during the Wikimania the 
unchecked articles numbre on the German Wikipedia increased to a few 
thousand. After the Wikimania the german voluntiers started a 
concentrated check action to reduce that number back to a managable size.

There are some experiments and talks and votes on the German Wikipedia 
on how the stable version should be showed to the user: Show the checked 
version to all users if they request an article with a hint to not 
checked version, show the checked version to IP-user and new user with a 
hint to not checked version, ans show the actual version to users with 
check right. Show actual version to all users with a hint to checked 
version and so on. I personally would always use the actual user, but on 
the other hand, I am not a user who ONLY search for information and 
don't know the mechanisms. And the votes thus far made on German 
Wikipedia are also only from editors with vote right. So I cannot say 
how an only read user see this. Academic tests on this would be 
interesting but as far as I know there is no such tests planned until 
now. On the other hand, with one exception until now I had also not 
heard any negative complaints about this anywhere. That one exception is 
on the forum of the c't magazine and I think the user is generally 
unhappy with the rigid quality rules on the German Wikipedia.

As far as I know the more advanced function of the stable version, the 
proofed version (that is, marked as proofed by an expert) is not 
implemented on the de-wp until now. So from my view it is a good tool 
for vandal battling but it still doesn't fulfill the function it is 
thought of: to give a random user a proofed, qualitatively garanteed 
article.

Beside ther de-wp there are at least two other projects that use the 
checked version. One is ru-wp. I didn't see any reports of the practice 
there, would be interesting too.

Ting