[WikiEN-l] Proposal for simplified credentials model

[Florence Devouard]

For the record, I am totally opposed to such a model.

If the english community decides to adopt it, it is the community 
decision. Fine.

But the Foundation should be entirely kept away from this. There is NO 
way our rare employees should have to deal with editors credentials. 
Doing so would on top further fuel the belief that we are in charge, 
putting us possibly at risk legally speaking.

At the same time, I believe this proposition will scare away some people 
and will result in people not telling what they are educated or trained 
in; having bad social consequences (people like the warm feeling of 
being in the small group of """this university""" or """that 
profession"""). This is I think a slippery slope toward requesting 
identification for various jobs.

Whilest I would agree we neeed identification for the jobs of stewards, 
checkusers and oversight, I fear the day we will request identification 
and credentials for the job of admin (yes, I saw the proposition 
mentionned in the press). Same for press contacts and 
business/partnership contacts.

This would entirely tip the very concept of our community. Trust build 
not upon someone credential but upon what the person does.

What counts is not the credential of the person, but giving a source for 
a controversial content. This is not because someone has a validated phd
that he should be more reliable than another.

Last, we should stop being the valet of the press. Each time there is a 
noise in the press, some feel we should respond, apology, change the way 
we are doing things.

I think that is a poor way to act. What counts is the values we share 
and the success we met. Not the noises done by journalists.



Anthere (for the general comments)

Florence Devouard, Chair of Wikimedia Foundation (for the statement that 
the Foundation did not ask anything, will not impose that, and definitly 
should stay out of that crap)



Erik Moeller wrote:
> also on:
> http://en.wikipedia.org/wiki/Wikipedia_talk:Credentials#Simplified_voluntary_model
> 
> After some more thought, I would suggest an almost completely
> voluntary model, where the only requirement would be that a user
> claiming credentials should put one of four templates on their user
> page:
> 
> # "This user's stated credentials have not been verified." => If
> placed by another user
> # "This user's verification of stated credentials is pending."
> # "This user does not wish to verify their stated credentials, and
> asks you to assume good faith."
> # "This user does not wish to verify their stated credentials, because
> they should not matter to you. Please judge edits on their merits."
> 
> (Potentially the last two could be generalized into an abstract
> template that lets the user provide an arbitrary reason.)
> 
> For verification process, I would suggest to keep the office
> completely out of the loop -- doesn't scale. Instead, verify
> exclusively by emailing credentials evidence to [[OTRS]] (mail from an
> institution address [requires reply to confirm], scanned diploma /
> PhD, etc.). This would be similar to the permissions queue we already
> have for copyright, or the general inquiries queue, and seems to scale
> reasonably well.
> 
> Using this method, we have a more obvious disclaimer present in cases
> where users do commit fraud (the Essjay page would have said "Does not
> wish to verify"), and at the same time, users with identified
> credentials can be found easily, which may be helpful in cases where
> you're looking for an expert on topic X (think categories).
> 
> I have some connections that could help to check for diploma mills
> etc., if we want to go that far; for now, a simple system should
> suffice.
> 
> Yes, this solution will also cause some conflict. That is, I believe,
> unavoidable. If we take our responsibilities seriously, we must be
> prepared to make a decision like this, even if it makes a small number
> of vocal people unhappy. I believe a position like the above could
> gain majority support, however.
>