We need to differentiate talks between those which bring a scientific contribution (something new & inventive inside) and talks which are helpful to bring the audience up to speed on a given topic (e.g. overview of botnets in the wild, or status of obfuscation...)

@cryptax Maybe also "engineering", as in "we know this _should_ work, here's how to make it _actually_ work"

@cryptax Can you create a 3rd category for talks that actually validate all those talks that claim “novel” results?

@thedarktangent to be honest, there's something "new" in every presentation, but sometimes it's just the way it is presented that's new ;P
I find it quite difficult to define what "new" is...