SpamBayes wins PCW Editors Choice Award for anti-spam software.

[Alan Kennedy]

Hi All,

If there any contributors of SpamBayes reading, Congratulations!
SpamBayes has won the Personal Computer World (pcw.co.uk) Editors Choice
award for anti-spam software, in a review of anti-SPAM solutions in the
October 2005 edition. (PCW, for those who don't know it, is sort of the
UK's equivalent of Byte Magazine, except that it's still publishing
after almost 25 years).

SpamBayes was one of two open-source apps in the group review, which
included commercial products from Symantec, McAfee, and half a dozen
other companies.

"""
.... SpamBayes 1.0.1 is definitely in a league of its own: during our
tests it obtained a 100% real success rate. It would have to be trained
for several months in order to check that it isn't too strict on a daily
basis and that it lets most of the "good" messages through. However, the
fact that it's free, offers a high level of efficiency and is compatible
with Outlook makes it ideal for anyone looking for a zero cost solution.
As such, we think it deserves our Editor's Choice award. As with all
Bayesian filters, it gets better with use, especially in terms of
detecting wanted mail.
"""

The only problem was they listed the "manufacturer" of the software as
SourceForge, so the product was known as "SourceForge SpamBayes". You
guys need to come up a team/manufacturer name. (But there is a
screenshot of the software in the review, and the "Python Powered" logo
is right there for all to see).

Congratulations!

Unfortunately, PCW don't seem to have made the review available online
(yet), so I can't provide a URL. Maybe someone else will have more
success finding a URL?

thought-ye'd-like-to-know-ly'yrs,

 

[Richie Hindle]

[Alan]

SpamBayes has won the Personal Computer World (pcw.co.uk) Editors Choice
award for anti-spam software

Yay! Do we get one of those cheesy medals to put on our website?

 

[Steve Holden]

Hi All,

If there any contributors of SpamBayes reading, Congratulations!

To which I add mine, even though I normally try to avoid "me too" posts.
The software is a great achievement, and deserves popular success.

[...]
The only problem was they listed the "manufacturer" of the software as
SourceForge, so the product was known as "SourceForge SpamBayes". You
guys need to come up a team/manufacturer name. (But there is a
screenshot of the software in the review, and the "Python Powered" logo
is right there for all to see).

While it may not adequately credit the implementation language, a Google
search for "sourceforge spambayes" results in the first hit being linked
as """SpamBayes: Bayesian anti-spam classifier written in Python.""".

The web page to which this link points unfortunately doesn't reapeat the
Python reference in the page body. But they only miss the Python
reference if they feel lucky and don't look at the browser title bar!

regards
Steve

 

[Paul Boddie]

Alan Kennedy wrote:

[PCW award to SpamBayes]

(PCW, for those who don't know it, is sort of the UK's equivalent of Byte Magazine,
except that it's still publishing after almost 25 years).

Hmmm. Even Byte at its lowest point was far better than PCW ever was.

[...]

The only problem was they listed the "manufacturer" of the software as
SourceForge, so the product was known as "SourceForge SpamBayes".

PCW may still be publishing after 25 years (half the magazine being
adverts probably keeps it just about economically viable), but they
clearly haven't yet managed to shake off that classic 1980s mindset
where "everything is a product by a company" (and, given the
superficial understanding of software licensing still likely to be
pervasive in the mainstream UK IT press, "everything else is public
domain").

As for URLs and other things, last time I looked at the PCW Web site,
it was all time-limited (or page-view-limited) viewing for
non-subscribers. If British print distribution wasn't such a lock-in,
I'd imagine PCW would have taken its place alongside Byte, staring at
us from the print media fossil record.

Paul

 

[Alan Kennedy]

[Alan Kennedy]
[Paul Boddie]

> Hmmm. Even Byte at its lowest point was far better than PCW ever was.

Well, I mostly disagree, but you've got your opinion.

I personally preferred Byte, particularly because of their orientation
towards what the PC market would become, not just how it currently was,
e.g. they would run articles on RISC vs. CISC, for example, when the
"battle" was just starting. But Byte went out of business: obviously not
enough people cared about what it had to say.

PCW is, and always has been, focussed on the state of the market as is,
which means they're always testing and reviewing the stuff you can get
off the shelves right now. Whenever I buy a peripheral, e.g.
scanner/fax, optical writer, digital camera, etc, I always check if PCW
has anything to say about it, or about the class of peripherals: chances
are they've done a reasonably thorough review quite recently. And they
are still around, after all these decades, because they provide
information that people want.

[Alan Kennedy]
[Paul Boddie]

> PCW may still be publishing after 25 years (half the magazine being
> adverts probably keeps it just about economically viable), but they
> clearly haven't yet managed to shake off that classic 1980s mindset
> where "everything is a product by a company" (and, given the
> superficial understanding of software licensing still likely to be
> pervasive in the mainstream UK IT press, "everything else is public
> domain").

Don't forget that the comprehension of IT journalists is generally a
good indicator of the comprehension of the general computer-using
public. But I generally find that the journos at PCW are a little more
enlightened than average: rather than copying and pasting corporate
product announcements, they actually use the stuff they comment on. I
personally put great store in the fact that PCW awarded the Editors
Choice award to SpamBayes, because it's based on actually *using* the
software, rather than doing a simple feature comparison.

95% of the people who read the review and download/install SpamBayes
won't give a monkeys what language it's written in. But they'll still
have a modern python interpreter installed on their system as a consequence.

IMHO, there is a great opportunity here for the python community:
SpamBayes is a "best-of-breed" product in a very important market,
anti-SPAM: it even beat commercial competitors. SPAM has become an
enormous logistical, financial, commercial and legal problem across the
world, purportedly costing billions of dollars(virus distribution,
phishing, scams, etc). If the community ever wanted to prove python to
be a serious language, here's a fine opportunity.

Surely that's worth a simple team name, for mnemonic purposes if nothing
else. Something different or unusual, like one of my favourites, "Legion
of the Bouncy Castle", who are a group of Java cryptography dudes

http://www.bouncycastle.org

(Also, I've often seen PCW refer to open source apps as "manufactured"
by individuals or teams: it's just that in this case the SpamBayes team
have made no name available).

> As for URLs and other things, last time I looked at the PCW Web site,
> it was all time-limited (or page-view-limited) viewing for
> non-subscribers. If British print distribution wasn't such a lock-in,
> I'd imagine PCW would have taken its place alongside Byte, staring at
> us from the print media fossil record.

I didn't notice any complaints when PCW ran a story this time last year
about Michael Sparks, python and python's use in the BBC's future
distribution plans for digital TV.

old-fashioned-ly'yrs,

 

[Paul Boddie]

I personally put great store in the fact that PCW awarded the Editors
Choice award to SpamBayes, because it's based on actually *using* the
software, rather than doing a simple feature comparison.

We can be thankful for that, at least. And you're right about SpamBayes
being a great opportunity for Python developers and the community.

I didn't notice any complaints when PCW ran a story this time last year
about Michael Sparks, python and python's use in the BBC's future
distribution plans for digital TV.

Well, I didn't even notice the story! ;-) But then, my exposure to such
newsstand magazines has diminished substantially over the past ten
years.

Paul

 

[Alan Kennedy]

[Alan Kennedy]
[Paul Boddie]

Well, I didn't even notice the story! ;-)

Here's the message I posted here at the time

http://groups.google.com/group/comp.lang.python/msg/4a33f07f11a0ef30

regards,