Re:Kaspersky Labs: How scary can an old school programmer be?
by unimpressive on Wednesday March 06, 2013 @01:55AM
(#5329777)
I often upvote stories using the metric "Would I add this to my 'saved stories' if it was a stand alone feature?". Very often the answer is yes even though I disagree with large portions of an article/don't think it's a good fit for HN.
What is a good heuristic for upvoting articles?
Re:Kaspersky Labs: How scary can an old school programmer be?
by keefe on Wednesday March 06, 2013 @01:53AM
(#5329774)
it depends, graphics systems have some pretty hard real time constraints for latency. The best gfx coders are required to find unique implementations/approximations for extremely difficult problems. Think about doom and carmack and what you would have to do to reproduce that given a 486.
Stuxnet imho wasn't that ridiculously impressive, though it was impressive, because the evidence is that "somebody" made sure the people writing it had appropriate sources.
But ultimately, graphics routines are a completely different beast than finding weaknesses in a security system that you don't own. Many programmers on here, given a bit of guidance and some gfx/sound guys, could put together a respectable demo. I don't think the same about something like Stuxnet.