Steve Comstock wrote:
>Also, what about the shop that tries to locate the
>end date and zap it? You may say it doesn't happen
>but I have heard tales...

OK, I know this is going to turn into a religious issue, but here's my $0.02.

In 30+ years I've never seen this. I've seen ONE case where a customer was 
using licensed software illegitimately (on an extra CPU)-and that was an 
oversight: when it was recognized, it was a full-price bluebird for the sales 
rep, who was NOT unhappy at all.

I realize there are issues with Certain Countries whose reputation suggests 
that such practices might be common, but without hard proof, I'm not inclined 
to invest a lot of effort-initial and ongoing-in a CPUID protection scheme; it 
just doesn't seem worth it. And if the alternative is that those countries 
simply wouldn't buy the software anyway, then it's really not a cost to the 
vendor. YMMV.

Of course any sysprog worth his or her salt could hack such a scheme in 
minutes; this is the same as having unalarmed windows on your locked house: any 
thief who's interested can get in via a broken window. The question is whether 
it's also worthwhile to lock the front door and stop the casual intruder (and 
yes, I realize that seems to be arguing against my thesis-but locking your 
front door is low-cost and low-impact, as opposed to building, maintaining, and 
staffing 24x7 a CPUID mechanism).

SAS C used to have a nice, simple scheme: the compiler would say "SAS C 
LICENSED TO <company>" whenever it ran. You could run it on the wrong CPU, but 
it would whine (I forget whether it also said "Hey, dude, wrong CPU!"). The 
theory (I assume) was that no serious company would run software that basically 
yelled "Look! Stolen!" every time, but in a pinch, you could get critical work 
done. Seemed elegant to me.

As ever, I'd ask the question: What problem are you trying to solve with a 
CPUID scheme? That is, what's the REAL problem THAT YOU ARE HAVING? Or is it a 
theoretical problem? If the latter, consider the cost before investing: is your 
software so magical that it never needs support? If it does tend to need some 
support, then you'll find the cheaters eventually, and have 'em over a barrel 
price-wise...

...phsiii

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to