thedufer

87 ms·

781.
by thedufer 12y ago
People go through a lot of cash in NYC. Most people I know pull $200 at a time, which is pretty close to that.
782.
by thedufer 12y ago
It looks like you're sending the same data to every client, but it sounds like Socket.IO/Faye aren't handling this well. Do they not have a `broadcast` method of something of that nature that handles your use-case?
783.
by thedufer 12y ago
Its dependent on sticky load balancing - a pretty big caveat. And as far as I can tell, it still doesn't work with the cluster module, which is what that issue is about.
784.
by thedufer 12y ago
That's an incredibly high rate of messages. What could you possibly need that many messages for? I'm in no way surprised that generic solutions don't work for your case.
785.
by thedufer 12y ago
It appears that the answer is "kind of". You can't use cluster, but you can put multiple processes behind a load balancer if the load balancer can do sticky routing. They should probably be a lot more explicit about this; t
786.
by thedufer 12y ago
That seems like a pretty big requirement that could frequently be deal-breaking. I would expect a big warning somewhere that you can't cluster if you're using socket.io.
787.
by thedufer 12y ago
By giving the wrong date, the author has all but ensured that the patent office is going to drown in completely irrelevant prior art claims.
788.
by thedufer 12y ago
> What matters more: the future, or your personal notions of what's classy? His point was about neither of those; it was about what is legal - something you seem completely uninterested in. Yeah, it might not be "classy"
789.
by thedufer 12y ago
It varies widely based on where you live, too. For example, in NYC state/local taxes are high (these are deductible on your federal return) so anyone with a reasonably high income is already beating the standard deduction, making the
790.
by thedufer 12y ago
> When someone releases an app to "connect people" or make some day-to-day taks easier, nobody seems to want academic papers to back their claims. Most people are qualified to evaluate whether an app is allowing them to connect
791.
by thedufer 12y ago
Yes, let's encourage people to pay ransoms. That's going to end in a rational system. Netflix switching would, in the short term, give them more bandwidth. In the long term, encouraging ransoms will cause them to go up and Netfl
792.
by thedufer 12y ago
That's a good point. I meant promise in the sense that that's how customers often interpret a bandwidth claim - legally, they're making no such promise (otherwise they'd be in a lot of trouble).
793.
by thedufer 12y ago
As bzbarsky points out, tests in the realm of 1e9 op/sec look like the entire function is being optimized away because there are no side effects. Something about the method allows it to do this, while it doesn't think its okay fo
794.
by thedufer 12y ago
The first two tests on that jsperf don't show the same behavior, though, and they differ in the same way.
795.
by thedufer 12y ago
> It sounds like Level3 wants to enter into bigger-money contracts with the ISPs As the article points out, peering rarely involves money (and shouldn't ever; as the word peer implies, its a largely symmetric relationship). The ISPs
796.
by thedufer 12y ago
If we're allowed to pick and choose our time series, the quarter century before that MIT saw inflation at almost double that of the grade inflation poster child, Harvard. Yeah, you can rationalize based on outside events, but without r
797.
by thedufer 12y ago
Yes, there are plenty of articles discussing MIT and grade deflation. Unfortunately, none of the ones I could find have any data - its all anecdotal from students. The hard data I did actually find ( http://www.gradeinflation.co
798.
by thedufer 12y ago
> Nope, I've carefully looked at MIT's history of grade inflation (the info is online) Citation? Every source I can find shows that a) they haven't released average GPAs since 2000 and b) from 1950-2000 they inflated grad
799.
by thedufer 12y ago
Either its an expense counted against his profit, and he never expected to make a profit (since he was already in the hole if it was exactly funded) - or it doesn't count as an expense and he made $20k. That's over 30% profit, as
800.
by thedufer 12y ago
> ...then use NaN = 0/0. Many languages throw an error when you divide by zero. As far as I can tell, if x != x didn't work you would have to resort to bizarre contortions like `x == float("NaN")` for such tests in
801.
by thedufer 12y ago
In many places you can't get rid of the bad teachers - largely because of overly powerful unions. People who would make good teachers are driven away from teaching by the perverse incentives. Yes, the number of good teachers is finit
802.
by thedufer 12y ago
> multiline console Shift-enter in the Chrome console creates a newline rather than executing. > live highlighting html as it changes In the Rendering tab (bottom panel) there's a toggle for "Show paint rectangles" whic
803.
by thedufer 12y ago
Soylent appears to mostly target the tech crowd, concentrated in SF/NYC. $4/meal is incredibly cheap in those places - raw ingredients of fresh produce will usually cost more than that. Not to mention that not everyone has time
804.
by thedufer 12y ago
I assume you mean NFIB? Asking small businesses doesn't really address the point, since the discussion appears to be about potential businesses that fail to find funding.
805.
by thedufer 13y ago
> var a = function(){throw Error('err')} undefined > try{a()}catch(err){err.stack} "Error: err at Error (native) at a (<anonymous>:2:26) at <anonymous>:2:5 at Object.InjectedScript._evaluateO
806.
by thedufer 13y ago
That double-naming thing is going out of fashion, since modern browsers + node.js all know how to figure out function names without it.
807.
by thedufer 13y ago
Not true as of 0.10.24, which is what I have installed.
808.
by thedufer 13y ago
What's the goal there? v8, at least, will give the name in a stack trace even if the function is declared like `var a = function(){}`.
809.
by thedufer 13y ago
Okay, that explanation sits a little better with me. The mathematical simplicity of splitting each dimension in half is very nice and clearly the simplest way to go when looked at in this way. Replacing the current non-answer with somethi
810.
by thedufer 13y ago
Cool visualization. I hadn't heard of quad trees and this presented them in a very clear way. Within a couple of minutes I understood how they work and what they are used for (the hit detection use-case particularly resonated with me

More ›