carbonatedmilk

8 ms·

31.
by carbonatedmilk 8y ago
This makes entire sense to me: When you're going to stream the episode of Stranger things to 15million people, who really cares what the one-time cost to encode is? Surely you'd take algorithmic complexity that was 100x or even 10
32.
by carbonatedmilk 8y ago
Pull quote: We run compute on the cloud and have no real-time requirements. I was asked by the Chair, “How much complexity increase is acceptable?” I was not prepared for the question, so did some quick math in my mind estimating an upper b
33.
by carbonatedmilk 8y ago
This image delights me. "Jenkins, hit the track!" "Immediately, Sir."
34.
by carbonatedmilk 8y ago
It's probably because of new EU legislation. Everyone's just trying to make sure that they're Voight-Kampff compliant.
35.
by carbonatedmilk 8y ago
I've often wondered why I can't set my campaigns to, e.g. "Only show this ad to people who have bought something from some vaguely legit website in the last 2 years" You can assume that any impressions you were going to
36.
by carbonatedmilk 9y ago
Look, you can push your code to a public git repository or you can complain about people looking at your code, but you can't do both.
37.
by carbonatedmilk 9y ago
I've had lousy outsourcing experiences and I've had good ones. The best outsourcing experience I've had involved a very technically adept CTO, some solid pre-thinking about architecture and scalability, and (most importantly)
38.
by carbonatedmilk 9y ago
Gosh darn it Orville, I told you we needed a section 333 exemption with a part 61 pilot rating, not a Part 107 waiver! Maybe the drone guys should just start strapping cannons to the wings, then they'd be classed as assault rifles and
39.
by carbonatedmilk 10y ago
My takeaway from this article: Ingesting 38 different purported nootropics may cause you to write super long rambling blog posts and inadvisably apply statistics to poorly designed experiments.
40.
by carbonatedmilk 10y ago
Really, this couldn't have happened to a nicer bunch. Research staff need to get published in "A+" journals for tenure, promotion and all that jazz. At the same time, they know that these journals are exploitative, bankruptin
41.
by carbonatedmilk 11y ago
John Hempton, who is mentioned in the article, has an awesome, frank blog at http://brontecapital.blogspot.com . Of course he's often 'talking his book', but he's always well researched and (I think) surprisin
42.
by carbonatedmilk 11y ago
This isn't a story of drug company overreach, it's about how your government has driven the costs of compliance so high that it's not worth making a drug unless you can extract monopoly rent from it. The drug (pyrimethamine)
43.
by carbonatedmilk 11y ago
Lots of discussion about whether the '7 common plots' are an artifact of the low pass filter he used. https://annieswafford.wordpress.com/2015/03/07/continuingsyu... Both sides also seem to acknowle
44.
by carbonatedmilk 13y ago
So now you know your effective CPA for the ad you ran. You can count on a 3.7% install rate. Your app costs $1 For every 100 people who click your ad, 3.7 of them buy it - Making you $3.70 You need to pay less than $3.70 per 100 clicks So