NewsScan Daily, 3 June 2003 ("Above The Fold")

NewsScan Daily, 3 June 2003 ("Above The Fold")
***********************************************************
NewsScan Daily is underwritten by RLG, a world-class organization
making significant and sustained contributions to the effective
management and appropriate use of information technology. NSD is
written by John Gehl and Suzanne Douglas, [email protected].
***********************************************************

"ABOVE THE FOLD"
        Software Piracy Down in 2002
        TiVo to Market Viewer Data
        Law School Dishes Up Course on Spam
        HP Wants to Cut Corporate PC-Related Costs in Half
        Personalized Cancer Treatment
        Looking For Meaningful Patterns In Your Life? See DARPA!

FEATURES
        Flash Card
        Honorary Subscriber: Edgar Guest
        Mailbag: The World As We Find It

SOFTWARE PIRACY DOWN IN 2002
Global business software piracy declined slightly last year, thanks to 
increased education efforts and more aggressive tactics by the software 
industry, according to a report released Tuesday by the Business Software 
Alliance. The study estimates that 39% of business software installed on 
computers worldwide in 2002 was not legally obtained, with the most common 
violation being purchasing a single legal copy and installing it on several 
computers, said Robert Holleyman, president and CEO of the BSA. To combat 
that scenario, the alliance is continuing to circulate brochures on piracy 
and is conducting amnesty campaigns urging businesses to pay for additional 
copies without threat of prosecution. In addition, companies like Microsoft 
have instituted new restrictions, requiring special activation codes for 
its software that are tied to a single computer. It's too early to tell how 
well these strategies have worked in reducing piracy, said Holleyman. The 
country with the largest percentage of illegal software was Vietnam (95%), 
followed by China (92%), and Indonesia, Russia and the Ukraine, all at 89%. 
(AP/Wall Street Journal 2 Jun 2003)
http://online.wsj.com/article/0,,SB105459260651856100.djm,00.html (sub req'd)

TIVO TO MARKET VIEWER DATA
TiVo, recently described as "God's machine" by FCC chairman Michael Powell, 
has begun offering advertisers and broadcasters detailed information on 
what TV programs and commercials its users are watching -- or skipping. 
TiVo executives say they will be gathering information only in the 
aggregate, keeping individual user data anonymous. "Advertisers and 
programmers undoubtedly will want to understand how the technology will 
impact viewer behavior, and this is the first opportunity for them to get 
data to understand how viewers are using the technology," says an IDC 
analyst. Unlike traditional TV rating surveys that indicate which shows 
consumers are watching, the TiVo data reveals *how* viewers are watching -- 
at which points they're fast-forwarding or rewinding for an instant replay. 
For instance, during the Grammy Awards, TiVo noticed a spike in instant 
replays when Julia Roberts walked on stage to give an award. TiVo's 
measurement service will initially focus on viewer habits during prime-time 
programs and will be marketed as a quarterly subscription report. 
Customized data will be available as well. (AP 2 Jun 2003)
http://apnews.excite.com/article/20030603/D7RDUTBG0.html

LAW SCHOOL DISHES UP COURSE ON SPAM
Students at John Marshall Law School in Chicago will be able to bone up on 
the legal issues surrounding unsolicited commercial e-mail this fall, in 
what's reputedly the first law school course offered dealing exclusively 
with the subject of spam. "This seminar will investigate legal and policy 
issues raised by e-mail marketing and spam," reads the description for 
associate professor David Sorkin's seminar titled "Current Topics in 
Information Technology Law: Regulation of Spam and E-mail Marketing." 
Sorkin's course will address "litigation and legislation involving spam and 
e-mail marketing; the application of tort law and other traditional 
doctrines to spam; concerns related to constitutionality, jurisdiction, 
extraterritoriality, privacy, content and public policy; regulatory 
perspectives; issues faced by Internet service providers and legitimate 
e-mail marketers; legal aspects of blacklisting and other antispam 
measures; and other relevant issues." (CNet News.com 2 Jun 2003)
http://news.com.com/2100-1028_3-1012404.html?tag=fd_top

HP WANTS TO CUT CORPORATE PC-RELATED COSTS IN HALF
Hewlett-Packard has developed a strategy it claims will allow corporations 
to cut by at least 45% the cost of owning and managing personal computers. 
Shane V. Robison, the company chief strategy and technology officer, says: 
"There's still a lot of room for innovation in desktop personal computers 
if you look at the field broadly, especially the problem of reducing total 
cost of ownership." Studies have indicated that about 80% of the cost of 
owning personal computers is associated with such activities as training, 
help-desk support, systems management, and software upgrades. HP's solution 
uses clusters of so-called "blade" personal computers, with each blade a 
circuit board that handles all the processing requirements of a single PC. 
By organizing in this way, HP argues that a company can save money because 
not all the blades are used at once, and so will have to pay only for the 
number of blades actually employed. (New York Times 3 Jun 2003)
http://partners.nytimes.com/2003/06/03/technology/03HEWL.html

PERSONALIZED CANCER TREATMENT
A California startup company called Genomic Health has developed new 
cancer-screening tests to allow a physician to choose the best drugs for 
treating an individual patient's tumor. One cancer specialist, Dr. Ronald 
Natale of the Cedars-Sinai Center in Los Angeles, says: "In the future, a 
piece of patient tumor will be used to determine the best and most 
appropriate treatment for them as an individual." The same technology can 
be used by pharmaceutical companies to predict which patients are most 
likely to benefit from experimental medications. (San Jose Mercury News 3 
Jun 2003)
http://www.siliconvalley.com/mld/siliconvalley/6001801.htm

LOOKING FOR MEANINGFUL PATTERNS IN YOUR LIFE? SEE DARPA!
The Pentagon's Defense Advanced Research Projects Agency (DARPA) is taking 
bids on a new project called LifeLog, which will help someone capture his 
or her "experience in and interactions with the world" via cameras, 
microphones, and sensors; the goal is to created advanced software to 
assist in the analysis of a person's behavior, habits, and routines. 
Privacy advocates are expressing concern, but DARPA spokesperson Jan Walker 
says that the "allegation that this technology would create a machine to 
spy on others and invade people's privacy is way off the mark." LifeLog's 
software "will be able to find meaningful patterns in the timetable, to 
infer the user's routines, habits and relations with other people, 
organizations, places and objects." (AP/USA Today 2 Jun 2003)
http://www.usatoday.com/tech/news/techinnovations/2003-06-02-lifelog_x.htm

*****

FLASH CARD
      "The awareness of the ambiguity of one's highest achievements (as 
well as one's deepest failures) is a definite symptom of maturity." (Paul 
Tillich)

HONORARY SUBSCRIBER: EDGAR GUEST
      Today's Honorary Subscriber is the newspaperman and popular versifier 
Edgar Albert Guest (1881-1959), who amazed the publishing world with his 
unstoppable talent for producing a steady stream of sentimental and 
moralistic rhymes on such subjects as home, mother, and the virtue of hard 
work.
      For several decades as verse columnist on the Detroit Free Press he 
wrote a poem every day of the week. His first book, the best seller "A Heap 
o' Livin'", was published in 1916, followed by "Just Folks" (1917), 
"Collected Verse" (1934), and "Living the Years" (1949). Guest is thought 
to have written more than 11,000 poems.
      Guest was born in England and came with his family to the United 
States in 1891. At age 13 he went to work for the Detroit Free Press, first 
as an office boy, and then as a police reporter. There he spent his entire 
60-year working career, and despite his fame as a versifier he always 
considered himself mainly a newspaper man.
      His verse writing was an unplanned career development. In 1898 he had 
editorial responsibility for finding verses and other short filler items, 
and one day instead of clipping a verse, he decided to write his own. Soon 
he had a weekly column, "Blue Monday Chat," and then a daily column, 
"Breakfast Table Chat."  Ten years later, in 1908, Guest decided to 
distinguish himself by writing almost entirely in meter and rhyme. 
Eventually he got together with his younger brother, a typesetter, and 
began self-publishing small collected volumes of his verse. That was the 
start of a book publishing enterprise that would eventually sell his books 
in the millions.
      Guest acknowledged that he was no favorite of "highbrow, longhair 
intellectual critics and writers," attributing his journalistic popularity 
to this simple formula: "I take simple everyday things that happen to me 
and I figure it happens to a lot of other people and I make simple rhymes 
out of them."

See http://www.amazon.com/exec/obidos/ASIN/1568491085/newsscancom/ref=nosim 
for the Collected Verse of Edgar Guest -- or look for it in your favorite 
library. (We donate all revenue from our book recommendations to adult 
literacy programs.)

MAILBAG: THE WORLD AS WE FIND IT
      We'll limit the mailbag today to a single topic, the filtering by 
SpamAssassin of the May 30th issue of NewsScan Daily, as first discussed in 
June 2nd issue. It includes a response from SpamAssassin's Daniel Quinlan 
and a final comment from us. See
http://www.newsscan.com/cgi-bin/findit_view?table=newsletter&dateissued=20030602 
for the issue that started this discussion.

LOTS OF ALTERNATIVES FOR FILTERING SPAM
      Your readers seem to be a pretty sophisticated group, so I thought 
I'd just mention that there are other types of spam filters than ones like 
SpamAssassin that use pre-defined rules. For example, I use an open-source 
software filter named SpamBayes (http://spambayes.sourceforge.net) that 
uses bayesian statistical analysis to rate the probability that emails are 
spam. This type of filter is different in that you continue to "train" it 
as you use it. So it continues to get better at rating whether the emails 
you receive are spam or not. Note that this product is still in a 
pre-release version, so unless you like living near the edge you should 
probably wait on installing it. Yet there are other filters out there like it.
      And as long as I'm on the subject, there's another way of handling 
email that some people find satisfactory: if you use an email client that 
has its own filtering capability (like Outlook), you can set up rules to 
move incoming email from addresses you know into a special folder. In 
normal usage, you would spend most of your email checking time in this 
folder. Periodically you would look in the default inbox, delete the spam, 
and update your rules if required. The underlying assumption of this method 
is that most of the email you want to see comes from a limited number of 
people/organizations. What I'm trying to get at is that if you are not 
using a spam filter and you are frustrated, or you are using a spam filter 
and are finding it unsatisfactory, there are lots of alternatives to try. 
(Not that it wouldn't be better to get rid of this scourge altogether!). 
(Bill Winett)

WAG THE COMPUTER
      I can't imagine why anyone would ask SpamAssassin to examine their 
NewsScan Daily emails or why you think you needed to edit to suit? We're 
supposed to program the computers, not the other way around. (Gary Welles)

SPAMASSASSIN WAS DEVELOPED BY THE GOOD GUYS
      I just read today's issue, and your derisive comments regarding 
SpamAssassin drew me to respond. While the system isn't infallible, in my 
own experience I have certainly found Bayesian filters such as SpamAssassin 
to be infinitely more effective and less error-prone than any other method 
I've tried. I began using SpamAssassin on a personal email account that 
was, even after my ISP's own filtering, receiving 10-20 spams per day on 
average. Since I started using SpamAssassin two months ago, I've seen one 
spam in my inbox. One. In two months. Like the reader who pointed out the 
false positive, I save the messages SpamAssassin deems to be spam in a 
separate folder, so that I can double check for false positives. There 
haven't been any so far. Friends who have been using SpamAssassin longer 
than I tell me that they haven't seen quite the 100% success rate that I've 
had so far, but it's still been a vast improvement to whatever they were 
trying before. It's not perfect, but it works very, very well. The reader 
who reported NewsScan's SpamAssassin false positive may not be aware that 
it's possible (and recommended) to create a "whitelist" of email addresses 
who are automatically assumed not to be spam. While this is somewhat 
problematic in that it puts the onus on the user, in a case like "Above the 
Fold" -- which is published regularly and due to its nature can be expected 
to contain news about spam and the Internet now and then that includes text 
that spam-catchers find suspicious -- it's a good solution.
      As to your larger point, that good users are having to work harder 
and harder just to keep up with the spam industry (much less stomp it into 
the dust where it belongs), I couldn't agree more. But SpamAssassin and its 
like really are among the good guys in this fight. (Emily Kelly)

SOFTWARE IS A TOOL THAT NEEDS SHARPENING
      While I sympathize with you generally regarding NewsScan's being 
filtered out, I think the recent posting regarding SpamAssassin was a bit 
over the top. The problem that Ari Rapkin experienced is attributable to 
the fact that the spam threshold was set at too low a level--5 out of ten, 
rather than a more reasonable 7.5 or 7. Both the academic accounts that I 
use installed SpamAssassin a couple of months ago and I couldn't be 
happier. One has an address that I never give out--I visited it after a 
month the other day and found 390 spam messages in my spam folder. All I 
had to do was Select All and boom, they were gone. The other account, which 
I use daily, regularly has at least half a dozen messages a day in the 
spamscreen folder, which I dispense with easily, rather than having to 
filter them out myself.
      Think of the time that I would have wasted doing all that manually, 
whereas now I'm free to waste it much more productively ;-). A few spam 
messages a week still get through, and only a couple of commercial 
broadcast messages (which I didn't need anyway) have gotten filtered out, 
but I've never had NewsScan blocked. The lesson here is, such filters work 
if you use them intelligently. And if your system doesn't allow you to set 
up a spam screening folder, or to set your own screening threshold, demand 
that it does. A spam screening filter is just like any other tool--you just 
have to keep it well honed to serve your purpose. (James C. Morrison)

KEEPING ONE STEP HEAD OF THE ANTI-ANTI-SPAMMERS
      It's easy to ridicule SpamAssassin as being too clever by half, but 
I've been using it for over a year and it's an absolute godsend. It catches 
over 1,000 SPAM messages a month for me, with at most one or two getting 
through each week, and fewer than one false alarm per month (valid messages 
incorrectly redirected). Interestingly, my installation did not trap last 
week's NewScan: one of SpamAssassin's nicest features is its ability to 
"learn" individual characteristics, building per-user whitelists and a 
probabilistic map between certain words and high or low SPAM likelihood, 
based on the actual mail you receive. 
http://www.spamassassin.org/  However, now that SpamAssassin is becoming 
more and more popular, I have begun to notice SPAM that appears 
specifically constructed to trick it. You can work around this to some 
extent by modifying the weights applied to different rules in your own 
installation, but it's an ugly arms race. To avoid missing future issues of 
NewsScan, readers on Unix machines can simply add whitelist_from 
*@newsscan.com to their ~/.spamassassin/user_prefs file. I've done this 
pre-emptively, since filtering NewsScan would surely be a tragedy! Thanks 
for your excellent work. (Dan Ellis)

KEEP IT SIMPLE
      Sorry to hear of your publication being caught in the crossfire of 
spam filtering programs. Don't let it get you down.  It's interesting to 
read of all the trouble people are going to get rid of spam: special spam 
program filters, spam reports to review, special folders for the suspect 
emails to sort through. So much time spent preventing spam, compared to 
simply deleting them as they arrive, no worry about filtering the wrong 
email. Too simple for some people, I guess. (Corl DeLuna)

IF YOU USE A FILTER, YOU NEED TO CHECK FOR FALSE-POSITIVES
     You say "...in fact IT DOESN'T WORK -- i.e., it doesn't successfully 
do what it is supposed to do, which is to filter spam and ONLY spam." If 
you can come up with one that does, the entire e-mail-using community will 
love you even more than we already do. The fact of the matter, though, is 
that you simply need to check for false-positives that your spam filter may 
have caught. The other option -- not using a spam filter at all -- is just 
not viable for me. I get well over a thousand a week, being a 
multiple-domain administrator/postmaster.
      In the case of SpamSieve, which I use and which has previously 
flagged NewsScan as spam, you can click on the message after it's caught 
and designate it as "good," letting SpamSieve know that it triggered a 
false-positive and not to do that again. Problem solved. For me, anyway. 
(Ian Blackburn)

EDIT, MODIFY, ADD NEW RULES
      Far be it from me to disagree that the Internet world has gone mad, 
as indicated in your June 2 News Alert. However, as I was taught in 
statistics and analysis, beware n=1. While the SpamAssassin software 
running on Ari Rapkin's system filtered out the May 30 issue of NewsScan, 
the SpamAssassin software running on my system did not. Why? Because 
SpamAssassin is a Unix-based program in which "rules, weights and 
user-visible text are stored in text configuration files as much as 
possible, which the user (or sysadmin) can edit to modify or add new rules" 
(www.spamassassin.org). The "pre-determined point value" and rules (and the 
non-standard "know-it-all 'Suuurreeee'," message) were determined by Mr. 
Rapkin (or, more likely, his sysadmin) and can be modified to catch most 
spam while still allowing him to enjoy uninterrupted delivery of NewsScan. 
While I do receive one or two spam messages per week, I am delighted that 
my SpamAssassin configuration has never blocked an issue of NewsScan. 
(Carolyne Garcia)

JUST A TOOL
      The rather excessive vitriol of your invective directed at 
SpamAssassin makes it clear that you were more than a little upset by its 
trapping of your May 30 issue. I think you need to take a step back and 
engage in a more dispassionate analysis of the facts. Spam is a huge and 
growing problem on the net and any serious net user must employ tools to 
help combat it. SpamAssassin is a particularly good such tool. To make the 
accusation that "IT DOESN'T WORK" is unfair when you don't know what it is 
supposed to do but have simply *assumed* that it is supposed "to filter 
spam and ONLY spam". What it is actually supposed to do is to *mark* 
messages that are *probably* spam for further investigation. But how does 
it know what is spam? The answer is, it can't possibly be sure -- that's 
way beyond the capabilities of current AI techniques. But it can have a 
very good guess, on the basis of a set of rules. It will sometimes make 
mistakes, though. That's why you have the option, as taken up by Ari 
Rapkin, of checking its results. You also have the option of telling it (as 
I have) that all messages from named originators (like, say, NewsScan) are 
to be passed without comment. SpamAssassin is a powerful tool but like all 
tools it must be used properly to be useful. (Merlyn Kline)

RESPONSE FROM SPAMASSASSIN
      The installation of SpamAssassin that marked Ari Rapkin's message 
appears to have been broken, misconfigured, or worse. The scores are all 
1.0 which raised the score of your message higher than it should have been 
and the message was then marked as spam. The scores shipped with 
SpamAssassin are optimized by a genetic algorithm that heavily favors false 
negatives over false positives. Basically, the best rules are assigned the 
highest scores.
      However, on Ari's mail server, those scores were deleted, replaced, 
or somehow lost. I don't recall seeing this particular administrator 
mistake before, but it essentially turned SpamAssassin into a rather dumb 
piece of software. Considering how many people use SpamAssassin, I suppose 
it's not too surprising that this happened somewhere.
      I don't think it's very accurate or fair to judge SpamAssassin by 
other filters and a broken installation. We can't do anything about either. 
Given the nature of your article, I think it would have been appropriate 
for you to verify the problem or contact us. I'd like to request a 
correction from you since there were a number of factual errors in your 
article:
      1. It wasn't our software, rather an old installation of our software 
that was installed incorrectly. The logic was "half-baked" because the 
software was setup only half-correctly.
      2. By intentional design, SpamAssassin *cannot* delete mail. There 
isn't even an option for deletion. We only mark messages as probable spam. 
Someone must go out of their way to intentionally configure their mail 
server or mail program to delete messages.
      3. The message did hit some rules and would be somewhat high scoring 
for non-spam, but should not have been marked as spam except for the broken 
installation.
      4. SpamAssassin cannot and does not block email. A misguided 
administrator can add software to achieve that effect, but we don't have 
any options in SpamAssassin to allow outright blocking.
      Also, what crime did we perpetrate? Letting people download our open 
source software for free? Helping people who receive hundreds of spam each 
day to clean up their inboxes? Making it much harder for spammers to reach 
people with annoying, disgusting, and illegal messages?
      The truth is that we actually allow you to reach a sizeable number of 
subscribers who would be otherwise completely overwhelmed by spam and would 
have been long since driven off the Internet and away from email if not for 
intelligent filters like SpamAssassin. (Daniel Quinlan, of SpamAssassin)

A FINAL COMMENT FROM NEWSSCAN
      We didn't suggest that anyone perpetrated a "crime" or wrongly let 
people "download open source software for free" (where did that come 
from?). And we actually applaud SpamAssassin's pursuit of a noble goal 
(because we, too, regard spam as a pestilence). However, here's the bottom 
line: NewsScan Daily has been treated (for whatever reason) as equivalent 
to "annoying, disgusting, and illegal messages." Since we believe in the 
First Amendment, we are horrified to be treated that way. Are we being too 
sensitive?
      Please understand that we are not singling out SpamAssassin for 
special derision. SpamAssassin actually seems to be a quite clever piece of 
software, and is no doubt much better than many other pieces of software 
out there that have treated us as essentially equivalent to senders of 
"annoying, disgusting, and illegal messages."
      But, frankly, it doesn't matter to us whether we're filtered by a 
virtuous open source product or by AOL, Microsoft, an ISP, or anyone else. 
To us, the filters all tend to feel the same. Whether they block us, delete 
us, or deposit us in a dusty attic for later possible inspection, they all 
impede the free flow of information.
     But then maybe we're just taking the First Amendment too seriously -- 
and failing to understand that the burden of proving we're not spammers or 
pornographers is on us, rather than on others.
     Think about that for just a second. In the same issue of NewsScan 
Daily in which we brought up this topic, we included a non-story called "A 
Moment of Silence for the Mob," in which we gave up on the whole idea of 
trying to write a story about certain illicit Internet activities, because 
writing it would require the use of words that (in themselves harmless) are 
highly likely to mark us as spammers or pornographers. It's no fun, in an 
adult newsletter (an a*ult newsletter) to be forced to use asterisks, 
dashes, and euphemisms to say fairly ordinary things. Yet that's what we 
have to do if want to avoid rejection by some or all of the various filters 
out there.
      Ah, well. Too bad we're not the New York Times.
      Anyway, a lot of good points have been made in the mail included 
above. In the next mailbag we'll move on, unvitriolicly, to other topics. 
We wish Daniel Quinlan and SpamAssassin the best of luck.

SUBSCRIPTION INFO FOR NEWSSCAN DAILY:
      To subscribe or unsubscribe to the text, html, or handheld versions 
of NewsScan Daily, send the appropriate subscribe or unsubscribe messages 
(i.e., with the word 'subscribe' or 'unsubscribe' in the subject line) to 
the addresses shown below:
      Text version: Send message to [email protected]
      HTML version: Send mail to [email protected]
      NewsScan-To-Go: http://www.newsscan.com/handheld/current.html

TELL YOUR FRIENDS ABOUT NEWSSCAN DAILY! Send them our little frog: 
http://www.newsscan.com/newsscan/newscup.html

***********************************************************
Please visit the site of RLG, the great organization that makes NewsScan 
Daily possible. (RLG has no influence over, nor any responsibility for, our 
editorial content.) Created in 1974 as the Research Libraries Group, RLG is 
a not-for-profit membership corporation of more than 160 universities, 
national libraries, archives, museums -- and other institutions with 
remarkable collections for research and learning. Its major initiatives are 
long-term retention of digital materials, resource sharing among member 
institutions, and improvement of researcher access to primary source 
material. RLG supplies online research resources worldwide. Visit RLG at 
http://www.rlg.org.

*********

To subscribe or unsubscribe to the TEXT version of NewsScan Daily, send an 
e-mail message to [email protected] with 'subscribe' or 'unsubscribe' 
in the subject line. To subscribe to our HTML version of NewsScan Daily, 
send mail to [email protected], with the word 'subscribe' as the 
subject. (Subscribing to the HTML version won't automatically unsubscribe 
you from the text version; please unsubscribe yourself as explained above.) 
Please note: To unsubscribe from an account to which you no longer have 
access, send mail to at [email protected].

*********

We call our news section "Above The Fold" to honor the tradition of the 
great "broadsheet" newspapers in which editors must decide which news 
stories are of such importance that they should be placed "above the fold" 
on the front page. The NewsScan Credo: Be informative, have fun, and get to 
the point! See http://www.newsscan.com/, and send us mail: John Gehl 
<[email protected]> and Suzanne Douglas <[email protected]>, or call 
770-704-7517.

*********

Get a free 6-week trial subscription to Innovation Weekly, an executive 
summary of trends, strategies and innovations in business and technology. 
Send mail to [email protected] with the word 'subscribe' in the 
subject line.

What customized news and information services do you need for your company? 
Get in touch with us at [email protected] or call 770-704-7517 to talk 
over the possibilities.

Copyright 2003. NewsScan Daily (R) is a publication of NewsScan Inc.
***********************************************************