Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
Cakez0r
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
28 ms
·
271.
▲
by
Cakez0r
12y ago
Out of curiosity, do you think that passion alone can carry a crappy idea to success? Is it better to be (truly) passionate with a trashy idea or have a solid idea but only be motivated by financial success?
272.
▲
by
Cakez0r
12y ago
Seriously? "At no point have you mentioned gender, but are you sure you're not being subconsciously sexist?" Perhaps n72 should also consider checking his privilege...
273.
▲
by
Cakez0r
12y ago
Find yourself a partner who is great at winning people over. You can make the best product in the world and it'll be useless to you if you can't even successfully pitch it to somebody. Soft skills are just as important as technica
274.
▲
by
Cakez0r
12y ago
I believe they're saying that a virtual DOM is created and updates that would cause a re-render are first applied to the virtual DOM (thus avoiding the over head of a render). The virtual and actual DOM are then compared to see which p
275.
▲
by
Cakez0r
12y ago
Thank you for taking the time. Ember's API stability made it the only viable choice for our web apps at work.
276.
▲
by
Cakez0r
12y ago
Does it have nuget (and / or npm etc) support? Pretty essential feature for a .NET IDE
277.
▲
by
Cakez0r
12y ago
All signs point towards Microsoft opening up their platforms rather than locking people in. I expect that in this case, it's more to do with the overhead of piping data in to and out of their data centers (I.E. paying for bandwidth) as
278.
▲
by
Cakez0r
12y ago
Would I like more privacy... or more accurate app reviews... Hmmmm...
279.
▲
by
Cakez0r
12y ago
I did not know either of those things! Thanks!
280.
▲
by
Cakez0r
12y ago
I really hope they implement this. It would make IO so much more efficient if you could pin a big chunk of memory and allocate buffers from it.
281.
▲
by
Cakez0r
12y ago
You can already get a long way towards channels by using reactive extensions (E.G. http://pastebin.com/h1xteunX ) Some syntactic sugar a la await could be nice though.
282.
▲
by
Cakez0r
12y ago
Outrageous claim(0) (0)This is a fact
283.
▲
by
Cakez0r
12y ago
They could ban only the accounts issuing the votes and revoke the fake votes when the account is deleted.
284.
▲
by
Cakez0r
12y ago
What if this bad government is already in power and are inching their way towards their full potential badness. 1. Do something somewhat unacceptable. 2. Wait for it to become the accepted norm. 3. GOTO 1 This is why issues like this have t
285.
▲
by
Cakez0r
12y ago
> Is a spy agency preparing ways to attack and cripple foreign infrastructure that unexpected or contentious? In light of current revelations, it may or may not be all that surprising. Surprising or not though, it is most definitely cont
286.
▲
by
Cakez0r
12y ago
Was thinking the same thing. Comments asking unbiased and valid questions are being downvoted...
287.
▲
by
Cakez0r
12y ago
I think offensive vs defensive is a very easy choice. We want to prevent our systems from being compromised, but don't have any business provoking attacks by breaking into systems. Invest in defense.
288.
▲
by
Cakez0r
12y ago
I don't understand why there are so many comments saying that the NSA needs offensive digital capabilities. What valid reasons does the NSA have to ever be committing a cyber attack?
289.
▲
by
Cakez0r
12y ago
But it's not inconceivable that in the near-future an AI could train itself to mutate (even if this means interfacing with mechanical turk or freelance websites and paying humans to do it).
290.
▲
by
Cakez0r
12y ago
The article mentions trusting HR with the company's deepest, darkest secrets... Company secrets are probably one of the biggest factors of my work satisfaction. If there are hush-hush meetings and rumors circulating, that's a huge
291.
▲
by
Cakez0r
12y ago
You know you're reading a quality article when one of the subheadings is literally just (quotation marks and all) 'Concern'
292.
▲
by
Cakez0r
12y ago
You stripped the context from the quote... They're not comparing going straight to sleep vs using an before bed e-reader. "A team from Harvard Medical School compared reading paper books and light-emitting e-readers before sleep.&
293.
▲
by
Cakez0r
12y ago
Several articles mention that the hackers have released stolen data, but I can't seem to find any download links. Does anybody have a source?
294.
▲
by
Cakez0r
12y ago
Curious about this too.
295.
▲
by
Cakez0r
12y ago
It was about 50/50 for the 6 samples I was given.
296.
▲
by
Cakez0r
12y ago
I wonder if they'd accept an acquisition offer from Google or Facebook!
297.
▲
by
Cakez0r
12y ago
There is a whitepaper about the details of TAO at https://www.cs.cmu.edu/~pavlo/courses/fall2013/static/papers...
298.
▲
by
Cakez0r
12y ago
Thank you!
299.
▲
by
Cakez0r
12y ago
Yes, that's the mechanism I'm wondering about. Let's say I have: Application -> TAO -> Cache -> Database I have a photo node (P1) and 125 comments nodes [C1, C2, ..., Cn] attached to P1 by the edges [(P1,C1)
300.
▲
by
Cakez0r
12y ago
Yes, I've read the TAO paper and referenced it heavily for my project. I'm curious about the implementation specifics of the caching layer. EDIT: More specifically, how do you efficiently cache your data such that an assoc_range q
More ›