On 5/19/11, m.roth@xxxxxxxxx <m.roth@xxxxxxxxx> wrote:
>> As already said, this sounds really complicated. Coming from an IRC
>> and vBB admin background, I'll suggest moderation using the reactive
>> approach instead of a automated process.
>
> Not really. The perl script was written, um, around 1993 or '94, and it
> was based on one from talk.lang.russian? something like that.
> <snip>
<context>skeptic about automated censorship and 'communities' void of
human expression</context>
The primary concern was that it would take a lot of tuning to get an
automated filter working without bouncing perfectly legit posts. After
all, this discussion about a need for moderation may very well fall
into "obvious" off-topic. And being the continual victim of apparently
some kind of automated filter on several mailing lists (including this
one it seems as some of my posts never show up), I'm rather
distrustful of automated moderation.
For example, a perfectly acceptable, from my POV, joke about a
statement somebody made that lightens up the mood and give everybody a
good chuckle may be considered OK by most subscribers but again,
automated censors do not have a sense of humour.
If the filters are too lax, then the moderators would keep getting
verification requests. After a while if false positives rate are too
high, moderators will feel frustrated too or start ignoring anything
they didn't come across first hand.
Hence I feel it's better to rely on the human flagging process. After
all, if only the automated filter and maybe one person feels a
thread/post should be a bounce, is that really deserving of a bounce?
_______________________________________________
CentOS mailing list
CentOS@xxxxxxxxxx
http://lists.centos.org/mailman/listinfo/centos
