[Tfug] A question for somebody else...

[Rich]

On Aug 11, 2008, at 6:33 am, Tim Ottinger wrote:

> I once installed squid as transparent proxy for a church,

<insert witty joke about porn filters and churches>

> and I remember pulling a blacklist from somewhere.  I think that  
> it's a tough game, and you have to continually review logs for  
> false positives and non-rejected sites.

Or set up a cron job to keep pulling a blacklist you trust.  
Obviously, you'll want to test it first, and check things like  
breastfeeding or breast cancer sites aren't blocked.

> Blacklisting isn't going to be your best option.   We did some  
> filtering on regex, though I don't remember how that worked.

From: http://www.squid-cache.org/mail-archive/squid-users/ 
200301/0507.html

> If you use regex lists, try to have them as short as possible. The  
> regex based acl types are farily CPU intensive compared to other  
> fixed matches.
>
> Do not use url_regex for matching domains unless you have a strong  
> reason to as url_regex matches anywhere in the url (including query  
> strings), and this can give quite surprising results from time to  
> time if the regex list is not very carefully crafted.


You might want to search/ask that mailing list for advice.

R.