3 ms·
I too do all these things listed by James, skipping articles and comparisons. But that's because I am now an expert: I've read far too much about dynamic vs. st
by inerte 14y ago
I too do all these things listed by James, skipping articles and comparisons. But that's because I am now an expert: I've read far too much about dynamic vs. static to have the patient to fully read the good old opinions about this subject.
The same about comparisons, I know it's 10% better than last year's because last year I actually read an article, and it was 10%-15% better than 2010, which was 10% better than 2009.
Two things helped: I stopped doing the engineering dance, dueling trivia with others. I don't care if the new iPhone does 5 times less network requests on maps when not on wi-fi but gets less POIs because the map provider is pulverized, on China is X while on Iceland is Y, but on San Francisco bla bla bla. Frankly, talking to a Wikipedia is boring.
The other thing that helped is that I gave up trying to live on the bleeding edge of anything. I don't sleep on the line to buy an iPad mini, I wait three days if and when I have the free time. I don't know what's the current best SSD, I google it when I am actually about to buy it. Slowing down is one of the best things I ever did. I have lots of friends that complain about being too "anxious" (all properly self-diagnosed). Just don't do something if you know you'll be complaining about it later.
- PaulHoule 14y agoI wouldn't take it for granted that it's 10-15% better than last year. Up until a few years ago, for instance, you could count on this year's CPUs being much better than last year's CPUs. After Intel came out with Nehalehm, the pace has slowed down. The really interesting developments are in details like virtualization support, SIMD instruction, and things like AMD's Fusion. And it took me plenty of talking with people to really understand that Fusion is nowhere near as exciting as I thought it was at first, since you can't get high-end (or even medium) GPU performance out of RAM designed for CPU's. The things I build today could be at their market peak three years in the future so understanding the hardware on the market and where its going can be the difference between making somethign that's commercially viable and something that isnt.