Actually I would still prefer a gnomesupport.org-like approach of the
restriction level in the gnucash wiki: Editing allowed for every
logged-in users and everyone can create a user account for
herself/himself. No manual grant of editing permission.
Iff this attracts too much spam, we can still consider setting a tighter
restriction level. And yes, I volunteer as one such 'sysop' user who
regularly checks for spam. Hopefully as one out of many sysop users.
Chris Shoemaker schrieb:
>>Mediawiki has a couple of quick-revert features that're generally
>>regarded as very good ... they were forged in the fires of Wikipedia,
>>after all. http://meta.wikimedia.org/wiki/Help:Reverting >
> Yes, they are very good. However, I'm surprised there's no multi-page
> admin roll-back. I mean, if the admin wants to roll-back a single
> page why not (optionally) roll-back *all* pages last edited by that
> user and why not (optionally) ban, or at least temporarily restrict
> said user, all in one operation?
In principle a single-click-rollback for all contributions of one (spam)
user might still help, but in practice I didn't need it on
gnomesupport.org. The most sophisticated spammers modified roughly 10-15
pages and each one several times, which makes reverting for the
non-admin users a PITA. But as an admin ('sysop') user, I click on the
"Contribution of user xy" page, and right next to each contribution line
there is the "rollback" link (only visible for admins), so I click on
each of these 15 rollback links (middle mouse button or whatever will
keep the original page open and focused) and that's it.
It really didn't bother me that I click 15 times instead of one,
especially since I need to at least spend one glance at the names of the
affected pages. So that rollback list will both inform me of the
affected pages and give me the link for rollback.
> I realize these features are over-kill for GnuCash, but for Wikipedia?
> Perhaps I over-estimate the volume of wiki-spam.
The problem on wikipedia et al is that the shear volume of users makes
it non-trivial to distinguish between actual spam contribution vs.
useful contributions (and due to the visibility it might even be
attractive for spammers to hide their activities by useful contributions
first). But the features are just fine for us as well.