Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
heurist
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
56 ms
·
541.
▲
by
heurist
10y ago
The implication that he's not doing his job because of his heritage is still pretty bad. It minimizes the judge's experience and boils him down to a skin tone. He has a conflict of interest because of his heritage? Appeal, sue him
542.
▲
by
heurist
10y ago
Great, now make it a first class citizen in compose and mix in docker-machine's provisioning capabilities!
543.
▲
by
heurist
10y ago
I imagine it would be like code coverage. You don't always want to unit test every single branch and 100% doesn't mean you're bug-free, but it's still useful as a metric.
544.
▲
by
heurist
10y ago
After 40 years swinging hard to the right I imagine it will take some time before brash liberalism makes a dent in living standards.
545.
▲
by
heurist
10y ago
Seriously. I get more buzzed off a couple cups of coffee than vyvanse, but vyvanse helps me stay focused much longer. I'd have no chance of attaining the life I want to live without antidepressants and some stimulants, the weight of li
546.
▲
by
heurist
10y ago
The house would be the opposite of a fresh start, and would likely hand it to Trump (or whatever right-leaning 'independent' might happen to show up).
547.
▲
by
heurist
10y ago
Or the engineer's solution to a social problem :)
548.
▲
by
heurist
10y ago
Even piecing together systems out of existing components significantly benefits from knowledge of first principles.
549.
▲
by
heurist
11y ago
I was suspicious when every prediction was perfect and I figured it out after it started talking about the stock market. Read much more like fiction at that point.
550.
▲
by
heurist
11y ago
I'd definitely like to know what it adds to PostGIS.
551.
▲
by
heurist
11y ago
I'm guessing Jeff Dean is not a typical Google engineer.
552.
▲
by
heurist
11y ago
Robert Axtell has done a lot of interesting work on agent-based modeling of economic systems [1]. He has a new book coming out this year about simulating national firm growth patterns based on his recent research [2], which could provide an
553.
▲
by
heurist
11y ago
"Why try to start a business when if you succeed, you won't keep that much of the winnings?" Some people start businesses to solve legitimate problems. I suspect that higher corporate tax rates would encourage the average Sil
554.
▲
by
heurist
11y ago
That's why banks offer business loans.
555.
▲
by
heurist
11y ago
`eval $(docker-machine env docker)` is great, I can stick it in scripts and separate my development environment from testing, staging, and production for every project I need to track. It works on Windows, OS X, and Linux so all my develope
556.
▲
by
heurist
11y ago
Yet these social networks are still corporate entities. It's only a matter of time until powerful groups learn how to bend and twist them to suit their goals, if they haven't already.
557.
▲
by
heurist
11y ago
What would you say is a good amount of equity for being an early engineer (effectively technical lead) at a seed-stage startup? I'm at my first and don't know if I'm being compensated fairly, though I do enjoy the work greatl
558.
▲
by
heurist
11y ago
It was explained to me this way: in the US executive boards are unevenly weighted toward CO interests. In countries where wage experts and workers are more likely to sit on boards the ratio isn't absurd because boards are less likely t
559.
▲
by
heurist
11y ago
I did not misinterpret you, I understand fully what you are trying to say. I disagree with you. You and I are in the fortunate positions of being able to service debt and even avoid it entirely. Many people are not. Lenders have the ability
560.
▲
by
heurist
11y ago
It's more complicated than that. You need to live in a cheap part of town to afford rent but your city is un-walkable and has poor public transportation? Buy or lease a car so you can reliably get to work on time. You have a family and
561.
▲
by
heurist
11y ago
>Well you're probably right, but I don't think that you can do too much, realistically, to regulate early access to very fresh ideas through networks like that - and I think that's fine so long as there's no purposefu
562.
▲
by
heurist
11y ago
It's not that bizarre, the suburbs surrounding Detroit include some of the wealthiest areas in the nation. Also, downtown Detroit is booming. I visited in July and it was night and day compared to when I previously visited three years
563.
▲
by
heurist
11y ago
That's just enough for the YC social network to incorporate OpenAI researchers, and Sam knows it. YC folks will be at the water coolers and ping pong tables with OpenAI folks, and once the OpenAI folks move into another office they
564.
▲
by
heurist
11y ago
Sounds to me like YCombinator wants to fuel their growth by creating these tools for their companies to use. The YC business model is absolutely brilliant, but I can't see this being some purely altruistic mission. Or if it is then I a
565.
▲
by
heurist
11y ago
Modern liberalism leaves no room for evil. The contemporary social conservative-liberal axis reflects personal responsibility. Conservatives believe individuals are entirely responsible for individual happiness and liberals believe society
566.
▲
by
heurist
11y ago
IIRC third world countries are large antibiotic abusers and they have no regulations in place to control for these things.
567.
▲
by
heurist
11y ago
It's a still and quote from this movie, commonly used meme: http://www.imdb.com/title/tt0119215/
568.
▲
by
heurist
11y ago
There's a huge economic incentive in having this data available - predict markets, sell or provide predictions to businesses, understand what your citizens are thinking at any point in time... Facebook, twitter, et. al. are all sitting
569.
▲
by
heurist
11y ago
Strong notes of Pynchon!
570.
▲
Complexity science pioneer John Holland passes away at 86
(santafe.edu)
16 points
by
heurist
11y ago
|
0 comments
More ›