savannah-hackers-public
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Savannah-hackers-public] Revisiting hosted wiki software


From: Sylvain Beucler
Subject: Re: [Savannah-hackers-public] Revisiting hosted wiki software
Date: Mon, 5 Jan 2009 20:35:18 +0100
User-agent: Mutt/1.5.18 (2008-05-17)

On Sat, Jan 03, 2009 at 07:10:54AM +0000, Noah Slater wrote:
> On Fri, Jan 02, 2009 at 11:16:22PM +0100, Sylvain Beucler wrote:
> > On Fri, Jan 02, 2009 at 09:46:19PM +0000, Noah Slater wrote:
> > > > So there are several solutions, a few of them require coding, all of
> > > > them require long-term maintenance, and all of them require finding
> > > > something against spam.
> > >
> > > We were running a MediaWiki for CouchDB before we moved to the ASF and we 
> > > were
> > > having spam problems. We forced users to register and login before making 
> > > edits
> > > and the spam problem vanished overnight.
> >
> > I have a couple public (i.e. !intranet) MoinMoins and Mediawikis
> > around - both are spammed by registered spammers :/
> 
> That's pretty strange! Even our new Apache MoinMoin installation doesn't get
> spam and that is significantly more visible in Google. Not sure what could be
> the cause of this.
> 
> Both MoinMoin and MediaWiki come with spam filters, IIRC.
> 
>   http://moinmo.in/AntiSpamFeatures
> 
>   http://www.mediawiki.org/wiki/Anti-spam_features

I'm aware of that, but I'm looking for more constructive
suggestions. Take the mediawiki one:

$wgSpamRegex -> simple-minded, inefficient, hard to maintain
Longer spam blacklists -> partial, need a 3rd-party URL block list
Spam cleanup script -> after-the-fact, off-topic
CAPTCHA images -> accessibility issues and annoying
Proxy Blocker -> far from enough
rel=nofollow link attribute -> idiotic and inefficient
Lock Down (Poor Solution) -> at least this one works

And the unspammed mediawiki that I see usually opted for the last
solution.. Concretely I don't feel like there's something ready for
websites with an large loosely-knit userbase here. I can be wrong,
that's where the constructive suggestion comes.

MoinMoin has a TextCHA solution that sounds good though, keeping in
mind there may be a l10n effort to setup.

-- 
Sylvain




reply via email to

[Prev in Thread] Current Thread [Next in Thread]