Forgot your password?
typodupeerror

Comment: To the core! (Score 2) 476

by Twinbee (#48035953) Attached to: Elon Musk: We Must Put a Million People On Mars To Safeguard Humanity
I love the idea, but I think it's been mentioned on Slashdot before that the best way to preserve humanity is to build a colony underneath the Earth's surface. Quite far underneath to protect against various threats, including medium sized asteroids and super volcanoes etc. We're talking about a self-contained, self-sustained system, to the furthest extent that we can manage.

Comment: Re:What, no positional tracking? (Score 1) 88

by Twinbee (#48011619) Attached to: John Carmack's Oculus Connect Keynote Probably Had Samsung Cringing
I did watch it. I was pleasantly surprised to see a focus on latency and fps throughout.

Perhaps G-Sync and a high frame rate is a simpler way to reduce latency, and then we can still keep the three frame pipeline. I don't think he mentioned G-Sync which was a surprise. JIT rendering also sounds good though.

Comment: Re:The 57% in the title is misleading. (Score 1) 182

by Twinbee (#48007801) Attached to: Breakthrough In LED Construction Increases Efficiency By 57 Percent
You missed the idea behind my post. I'm talking about the 'closeness' to 100%. From a particular perspective, something that's 99.9% efficient is 10x as efficient than something that's merely 99% efficient, since the former is 0.1% from 100% whilst the latter is 1%.

That means, heat loss will be 10x greater.

However, from the perspective of brightness (rather than efficiency), you're right, 99.9% is only about 1.01x more efficient than 99%.

Comment: Re:The 57% in the title is misleading. (Score 1) 182

by Twinbee (#48007185) Attached to: Breakthrough In LED Construction Increases Efficiency By 57 Percent
If you dig deeper into the math, it's a 55% improvement rather than 57%. That sounds like a trivial difference I know, but the logic is important for future comparisons. Let me explain.

Presume the new efficiency was 99% and the old efficiency was 38%. Naively, that sounds like 2.6x better. However, it's far more constructive to look at it in terms of how close it comes to 100%. So really, the formula is (100 - 38) ÷ (100 - 99) which means it's not 2.6x more efficient, but 62x more efficient. To get the percentage improvement, we add a little more to the formula: (100 - 38) ÷ (100 - 99) × 100 - 100 = 6100% improvement.

With the original numbers we calculate (100 - 38) ÷ (100 - 60) × 100 - 100 = 55% improvement.

Comment: Re:What, no positional tracking? (Score 1) 88

by Twinbee (#47998433) Attached to: John Carmack's Oculus Connect Keynote Probably Had Samsung Cringing
I hope small amounts of lag in VR causes nausea too, because if there's one thing developers don't seem to care about enough, it's latency/lag and frame rate (no, even 60fps isn't good enough). If they can get it below 10ms lag, then I'll start to be a bit happier.

Comment: Re:Black holes are real, we observe them all the t (Score 0) 356

by Twinbee (#47987203) Attached to: Physicist Claims Black Holes Mathematically Don't Exist
I believe it WOULD take miles if the string was frozen and couldn't 'move' once it had been lengthened. Your answer assumes the dog can pick the string up to go under it (which is a very fair assumption to make, granted). I think you basically were right but misunderstood what he was getting at.

Comment: Re:List the STL? Seriously? (Score 1) 471

by reanjr (#47979527) Attached to: Ask Slashdot: Finding a Job After Completing Computer Science Ph.D?

If you're looking for someone with STL experience who is expected to be up to speed and generating good, idiomatic code, then these types of questions can help weed out people who know C++, but don't have a lot of practical experience with the type of code your company wants. It's not the best interview question, but it's not terrible, either. Unless they expect you to actually list every single one. If you can spout off array, vector, and deque; that's a pretty good start.

Your code should be more efficient!

Working...