Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
gibybo
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
29 ms
·
331.
▲
by
gibybo
14y ago
Presumably they want to sell much more than 300 devices (enough that running them themselves would be less efficient than selling them) and turning on these 300 for themselves would make customers much less willing to buy more from them in
332.
▲
by
gibybo
14y ago
I prefer the term skeptic, and I think it's quite reasonable to be skeptical of a $1300 device capable of generating $210/day with still no proof. I think the skeptics will be much calmer once he plugs it in and verifies that it does what i
333.
▲
by
gibybo
14y ago
The $210/day for $1,300 only holds for a single incremental device. Since the network produces a mostly fixed number of bitcoins per day regardless of the hashing power thrown at it, there are severe diminishing returns for running a lot of
334.
▲
by
gibybo
14y ago
Their DOTA clone isn't a standalone game, or at least hasn't been so far. It is/was built on the StarCraft II engine, so if it were to support Linux, StarCraft II would pretty much have to as well. However, I think WoW is much more likely s
335.
▲
by
gibybo
14y ago
I think that depends what you're comparing. For CPU bound applications, I think you're right, $/cpu is fairly expensive on EC2. However, memory tends to be cheaper on EC2 than most other VPS providers: EC2 Small Instance with 3 year reserva
336.
▲
by
gibybo
14y ago
I don't think it's quite that simple. Here's some other things Paul Graham has said about that question: "There's one question that acts like a wildcard, at least for me: [the question] ... We're looking for people who like to beat the sys
337.
▲
by
gibybo
14y ago
Wow, that is a beautiful design :) I wrote something similar for http://giby.tv , though it's not really a core feature for me and it doesn't recognize geographical restrictions, which seems to be a primary motivation for you guys. You guy
338.
▲
by
gibybo
14y ago
It wouldn't just be one direction, it would be a sphere around the whole star. Stars are essentially just large fusion reactors, so it would take one the size of the star to compete with it (within an order of magnitude or so).
339.
▲
by
gibybo
14y ago
Maybe those people/countries think the same about people using the comma as a thousands separator instead of a decimal?
340.
▲
by
gibybo
14y ago
Heh, I've lost count of how many times I've written "document.onkeydown = function(e) { console.log(e.keyCode); }" to deal with this. This is probably easier, if I can remember it when I need it :)
341.
▲
by
gibybo
14y ago
I was looking forward to verifying the P = NP proof on level 3, but sadly I don't have access to DARPA’s 1000-node testbed, nor does my phone have any optical storage space. Sigh :( Anyway, love the challenge, the attention to detail is awe
342.
▲
by
gibybo
14y ago
Wait, 1.5mg per petabyte at large scales? Wouldn't that mean a gram could hold (1000/1.5) 667 petabytes and presumably scalable to many grams (eventually)? I understand it's only 650kB right now, but the density is obviously still incredibl
343.
▲
by
gibybo
14y ago
It says they stored 44 petabytes total (70 billion * 700 kilobytes), and 700 terabytes in a single gram (which is 5.5 petabits, not 4.4 petabytes).
344.
▲
by
gibybo
14y ago
At the top of the article, it says "Date: 05 August 2012 Time: 12:31 AM ET", lower in the article it says "Aug. 5; 1:31 a.m. EDT" but both of those seem to be mistakes since those times have already passed and nasa.gov says "Aug 6, 2012 1:3
345.
▲
by
gibybo
14y ago
How much is the fee? I suspect if given the choice between fee or commercials, many would choose the fee. Sadly that choice doesn't exist here (in the US) :/
346.
▲
by
gibybo
14y ago
I like this idea, but I'd probably opt for reading the stats out of memory instead of OCRing them. You'd get more data (since you don't need to have only one specific tab open), it'd be more accurate, and it might even be easier. If anyone
347.
▲
by
gibybo
14y ago
How are you generating the video files then? Surely you cannot run a game of SC2 on an EC2 micro instance?
348.
▲
by
gibybo
14y ago
If stocks moved up when earnings were good and down when earnings were bad, it would be pretty trivial to game the system and double your money every month or so. The reason is that quarterly earnings are not that hard to predict - everyone
349.
▲
by
gibybo
14y ago
I don't think payday loan shops are ever constrained by the amount of capital they can loan out, though. If they were, it would be trivial for them to borrow money at normal rates (say 10%) from a traditional bank and solve that problem. I'
350.
▲
by
gibybo
14y ago
I'm not sure why anyone would believe him enough to send him money. Perhaps they don't understand how absurd that claimed interest rate is? :x
351.
▲
by
gibybo
14y ago
Alright, so I spent another hour playing with these numbers and I think I understand it correctly now and agree with your original math. A combination of mistakes (F/I vs f/i, taking money out at the start of the year) and my imperfect intu
352.
▲
by
gibybo
14y ago
Ok I think I've got a counter example. First of all, I'm only going to consider values where N approaches infinity && r < 1 && r > 0 to make the math easier. lim [n-> inf, r < 1, r > 0 ] of S_0 = E_0 (1-
353.
▲
by
gibybo
14y ago
Sorry, I missed the F = (1+f)/R = (1+r) line. Pretty sure the rest of my argument holds though. Your formula merely approximates the correct answer for low values of i and f.
354.
▲
by
gibybo
14y ago
He's right. I'm not sure what your 'simulation' involves, but it's possible you just happened to pick numbers where the ratio produces similar results to the difference. Try inflation = 0.0001% and interest = 10%, for example.
355.
▲
by
gibybo
14y ago
It's not the ratio of interest to inflation that matters, it's the difference. The easiest way to conceptualize it is to consider an inflation rate of near-0, or even negative. The ratio would imply a months salary would be sufficient FU mo
356.
▲
by
gibybo
14y ago
Ok, requiring the attacker to download several terabytes could be an interesting advantage. Could we also achieve that by inserting billions of fake users in a traditional [user, salt, hash] setup?
357.
▲
by
gibybo
14y ago
I am skeptical that this method adds any protection to a brute force attack relative to a standard implementation of PBKDF2 hashes stored with the users. With proper salts, we have no advantage cracking a password in a database of 10 users
358.
▲
by
gibybo
14y ago
Diablo 2 actually had all business logic in the client itself because it needed to support single player and LAN play. They were able to just use the BNetD project to run Diablo 2 compatible battle.net servers with very little D2 specific l
359.
▲
by
gibybo
14y ago
I realize the 'profits' are donated to Teach for America, but doesn't using any part of the $10 warp your incentives? If you have a list of improvements to make to the site and it turns out one of them indirectly helps students avoid the $1
360.
▲
by
gibybo
14y ago
I don't believe it would be the possible to do the former with Bitcoin, short of telling them your private keys and erasing them from your storage I guess?
More ›