[hashcash] camram 0.3 announcement
- From: "Eric S. Johansson" <esj@xxxxxxxxxx>
- To: CAMRAM <camram-spam@xxxxxxxxxx>,hashcash <hashcash@xxxxxxxxxxxxx>
- Date: Sun, 20 Jun 2004 15:13:16 -0400
One pill makes you larger,
And one pill makes you small.
But the pills that spammers sell you, don't do anything at all.
Go ask Alice...
(with apologies to Lewis Carroll and the Jefferson Airplane)
All -- If you have even a passing interest in anti-spam technology, give a look to
http://camram.org/ . The latest release of Camram, 0.3.14 (a.k.a the "Go ask
Alice" release), is now available for
public consumption. (Claims not validated by the FDA or any scientific
research.)
Large swaths of new functionality are in place. To wit, in the 3 months (to the day) since 0.2 was released, here's what's new:
- SMTP interception filtering. Camram can now be installed as a
"lump in the line" for both inbound filtering and outbound
stamping.
- Auto-adjustment of thresholds in the spamtrap. Camram applies
sensible algorithms based on both user action and the passage of
time to relax the scoring limits on email considered spam or
non-spam.
- Automated downloader / builder / installer. The new "Raging
Dormouse" takes much of the pain out of downloading and building
Camram and its associated pieces.
- Comprehensive install documentation, including several new
diagrams explaining Camram architecture, terminology, and operation.
- Improved user interface for spamtrap and user preferences.
- Layered configuration files.
- Spam filtering on a per-user or per-domain level, or combination
thereof.
- A better user experience while training the CRM 114 Bayesian
filter.
- Easier setup and operation of the Stamper component.
- Optional challenge-response using stamps. This includes a new Java
applet that runs in the sender's browser to generate a stamp.
- Additional error trapping.
- Reducing / eliminating the possibility of losing, or mistakenly
passing, messages due to spammer attacks on bypass filters.
In its current state of maturity, Camram can handle sites of a few undred users when filtering in the local delivery agent and about half that when filtering the SMTP stream. This means Camram works for small to medium scale enterprises. Some progress has been made towards implementing Camram on the desktop, and necessary changes for larger scale enterprises have been identified for future work. Given either volunteers or funding, future releases will push the scale limits both up and down.
Other related posts:
- » [hashcash] camram 0.3 announcement
