I would say that absent some kind of serious emergency, sysops should
never protect pages that they've edited recently.
> 1. Telling the list you protected the article
(and why).
> 2. Making ONE LAST EFFORT toward neutral POV.
> 3. Then butting out until a consensus emerges.
The problem I see with this is that once an article is protected, it's
hard for a consensus to emerge.
--Jimbo