Time Machine
is it 1980 again? huh? Seriously, this song was released THIRTY YEARS AGO!
Posts by David Benoit
13 publicly visible posts • joined 9 Apr 2007
Men at Work appeal Down Under plagiarism ruling
Defects in e-passports allow real-time tracking
Wednesday 27th January 2010 16:32 GMT 
OR!
"It could be slid to the "on" position when the passport is needed to be readable via RFID and slid to the "off" position when the passport does not need to be read."
_OR_
the passport could, I dunno, be slid through a device that reads the information from it directly. Oh wait, that's what we had/have!
Yahoo! shares! hit! five! year! low!
Asus Eee PC 900 Linux Edition
World Bank chief: Ethanol cars run on human misery
Moroccan IT engineer arrested over fake Facebook account
Monday 18th February 2008 22:59 GMT 
Democracy has nothing to do with it
This is not a question of democracy. This is a question of whether this person impersonated someone else. It is a crime in most countries to impersonate people in positions of authority (police, fire, etc) let alone those as part of either the administration of the country or the monarchy of it.
How do you think the British government would react if the same thing happened with someone from the Royal family?
How do you think the US government would react if the same thing happend with some child of the president, or similar?
I'm sure whomever was involved would get some severe punishment.
Of opposable thumbs and software engineering
Monday 22nd October 2007 03:28 GMT
Massive Scale, even on Multi-Core...
If you're interested in exploiting the full power of your CPU, but aren't interested in the problems that come along with the "traditional" multi-threaded approach, or virtualization, or whatever, then I suggest you have a look at this article Brad posted a little over a year ago.
http://solvetherealproblem.blogspot.com/2006/08/multicore-singlethreaded-megascale.html
My colleague explains a bit behind how we attain massive scale without the constant need for more and more horsepower, or completely changing our techniques.
We have used this approach successfully over the last decade for telephony and "web-ish" systems, CPU intensive applications, network intensive, etc. We usually have the following problem: people in ops wanting to turn off our machines because we aren't using enough of the CPU, and us needing to keep it on for our redundancy model. We do, however, have a few CPU intense applications that run on quad-core machines with no problems. We run each app at 50% of a core, run 4 instances of it, and tell the "consumers" of that service about all of them, and they round-robin their requests among them.
Aussie scam artist rides bank computer glitch to $9m overdraft
Google, Yahoo! and Amazon sued over email patent
Thursday 30th August 2007 03:42 GMT
Prior Art
So, aren't the claims in the patent basically what "procmail" does? If so (if I read the patent correctly), isn't procmail (which was first released in 1990 - see http://www.procmail.org/procmail.HISTORY.html - it was up to version 3.11pre7 before the patent application was filed) an example of prior art, thus making this patent invalid?
If it isn't, doesn't that mean that anyone in violation of this patent would have to follow the exact flow described in the patent for the "automatic determination" of messages?
If this patent is valid, wouldn't it make all SPAM filters also violators?
/me hates trolls.
Biting the hand that feeds IT
