New Scientist, Feedback, 8 February 1997



INTEL held an informal party for the press recently at a London restaurant.

It was a relaxed event with no sales pitch. The only computers were a couple

of the new MMX Pentiums in a side room, with a pile of CD-ROMs of new MMX

software.



Inevitably, the computer hacks sidled off into the side room and

experimented with the PCs. Inevitably, too, someone pressed one button

too many and one of the PCs crashed and had to be restarted. The PC

was password-protected, so as the screen came back to life, up came a

request to enter a secret word.



Feedback tried all the usual words, like "letmein". But they all resulted

in the same error message, "Invalid password", with two options: "OK" to

try again, or "Cancel". One by one a series of computer experts clicked

OK and tried again and again, all without success. No one could find anyone

who knew the secret password.



After an hour or so a casual passer-by, who probably wasn't even part of

the party, leant over the PC and clicked "Cancel" instead of "OK".



As if by magic the PC sprang back to life, leaving Feedback wondering

whether the password-protection on all PCs is as easy to defeat.







New Scientist, Letters, 1 March, p 53

Fake security



I liked the story about the OK/Cancel keys (Feedback, 8 February).

I think that this is more common than most people realise, and that a

lot of barriers are purely imaginary.



If something looks impressive and formidable, people assume it's not

worth their time trying to break it.



Take, for example, Microsoft's 17-digit key for its Windows 95 discs (it

takes the form xxxxx-OEM-xxxxxxx-xxxxx) [sic]. Excluding the possibility

that some digits are predetermined, there should be less chance of guessing

the right code than there is of winning the lottery jackpot two weeks

running. So why is it that I can put any of our Windows 95 CDs into a

machine then pick any of their codes at random and it in stalls.



Yet Microsoft's desire to prevent software piracy at other times is

intrusive to the point of being ridiculous. Trying to install a new hard

disc, which the manufacturer had sent without instructions, I looked

in vain for help from Microsoft, whose operating system software it was

that required configuring. A keyword search of their online help found

three entries on how to tell whether preinstalled software on a hard disc

is legitimate, but none on how to set up a newly installed disc.



Bartie Wells

by e-mail





New Scientist, Letters, 22 March, p 54

Seven up



The reason the licence codes entered by Bartie Wells (Letters, 1 March, p53)

will work with any copy of Microsoft Windows 95 is that the only criterion

for these codes is that they are divisible by 7. Hence 0000000000000007

will work just fine.

The only justification for this that I can see is that Microsoft cares

more about making its software ubiquitous, than ensuring every copy is

licensed.



Jonathan Powell

Birmingham