BleepingComputer.com: Media Discovers Spyware

We published another in the random series of PSC Newsletters. This one is about spyware, its evolution into the mess we all know too well, and some thoughts about where it's heading and the software to stop it from ruining your life.

Click here to read the PSC Newsletter: Media Discovers Spyware



//Mod edit: e-mail address removed to protect from spam bots

This post has been edited by KoanYorel: 14 May 2005 - 01:01 AM

That's an interesting advertisement. How many times can they say BOClean in one paragraph?

Enjoyed the history lesson though.

Of what help is publishing, moreover, a test which they themselves term "heavily biased" and whose "results are not reliable?" This is akin to a lawyer purposely asking a question that he knows the judge will instruct the jury to ignore, knowing that they have heard it.
Such tricks would lead the careful reader to discount anything in the article, and certainly give pause to the purchase of their product.
Regards.
John

This post has been edited by jgweed: 14 May 2005 - 02:54 PM

Because too many people get insufficent protection from software purchased/downloaded from a vendor focused on bogus comparisons and "testing" plastered about the net. Ever see a vendor test where the vendor *didn't* score 100%? Do you know if the "independent" test you're relying upon isn't a "pay to win"?

Above all, we were very straightforward about the how, when, what and whys of this little exercise. Every vendor test is a stacked deck, complete with contrived parameters and chosen specifications to best suit the testing vendor.

After years of vendor drivebys described above, universities testing "zoo trojans" that never saw a moment in the wild so they'd be a waste of time for a commercial product to detect, and hustlers with "pay to win" scenarios it's become time to illustrate just what these "tests" involve. We came right out and said it was "Heavily Biased Testing". Doesn't anybody get it?

Another thought- This test was done by an associate in the wee hours since *he* was tired of seeing his work ignored by largely everyone. He was tired of the slow and non-reponses by so many to the hundreds of samples he sends to a long list of vendors. Collecting and submitting as many as he typically does, lately several hundred daily, is a lot of work, only to see it ignored. He did it to make a point.

There are a handful of organizations that do conduct controlled, scientific and professional testing. If they were the only ones publishing tests there would be no need for this conversation. Sadly, the glut of bogus tests is drowning their voice.

Quote

Precisely, and publishing- - -even with caveats---biased and unreliable tests is certainly neither a solution or a contribution to the advancement of truth in providing a true account.

Regards,
John

Doesn't illustrating that these vendor tests are deceptive have value?

Yes, but from what I read in the link, whatever the purpose of the example, it was NOT to prove that vendor tests are deceptive. Nor, for that matter, would I venture to make such a reductive statement that all vendor tests, since conducted by vendors, are necessarily invalid.
Regards,
John

jgweed, on May 14 2005, 09:41 PM, said:

Please then, show me a vendor test where the vendor conducting the test didn't score 100%, without use of obscure parameters (there was one vendor that marked BOClean for not having an integrated right click filescan option, fully knowing BOClean is a real time system with no filescanner), provided notification and comment by all companies involved (so many have stated this is practice, I haven't been contacted about any test since 2002) and following acceptable scientific practice. Chances are, you won't.

My argument was aimed at the logic of the argument, and the reductionist use of "every":

Quote

Reading the argument further, the warrant for the conclusion is not clear.

Any test, especially when discussing spyware, has chosen specifications because the DEFINITIONS of what instances are spyware varies from one vendor to another. To call this process "stacking the deck" or to be "contrived" is to used emotively loaded words to describe vendors using their own definitions. While I am not sure that each and every vendor in fact does this, the reason would therefore appear to be systemmic and not attributable to unknowable motives.

Regards,
John

Quote

When vendors choose their own definitions, parameters and the software they test against, the results could never be imagined to be scientific or even sincere. Calling such contrived or a stacked deck is merely being factual in describing such a "test".