gwillen

If you are familiar with Ripple, it seems like a curious omission from your story.

If you are _not_ familiar with Ripple, you should learn about it before you get on board with Stellar. (Stellar is a clone of Ripple, so familiarize yourself with the controversy about Ripple in order to learn what the controversy about Stellar is going to be in short order.)

I am a fan of both the Bitcoin and Stellar concepts; I hold a bunch of Bitcoins as well as some Stellars. I used to hold some Ripples; I may yet buy some more.

At one time I had good luck with a streaming app called Bambuser. I switched from Ustream because the latter had issues (at the time) dealing with poor connectivity.

While I understand that giving people experimental drugs is extraordinarily dangerous, it still saddens me that, in all likelihood, these guys only got the rules bent because they were doctors. Anybody else would likely not have had the option.

You will be a bit hosed pursuing a CS degree on Windows. People do it, you can make it work, but you'll be constantly fighting to get your system to do useful things.

I swear to god, Kenton, the first thing I thought of when I saw this link was "hmm, I wonder if they would port this to Sandstorm."

A "normal" is a vector that sticks straight out of a surface, perpendicular to it. (This is the usual mathematical definition of the word 'normal'.)

When shading a 3d-rendered surface, the normals at each pixel are used to determine how bright the light on the surface is at that pixel. (If the light is shining directly at the surface, i.e. along the normal, it's bright; if it's shining shallowly along the surface, i.e. at a sharp angle to the normal, it's dim.)

If you make, say, a cube, and set all the normals "realistically" -- that is, you make them poke straight out of the squares that make up the surface -- and then you render a light shining on it, it will look like a cube. Every face will have all its normals pointing the same way, so every pixel on a given face will be lit the same and render the same brightness.

Now, instead, try _blending_ (or "interpolating") the normals smoothly -- so the center of each face has a normal that pokes straight out, but as you get closer to the edge, it gets flatter, so by the time you reach a corner, the normals on both adjacent faces are pointing the same way, straight out of the corner.

If you can picture that, you should see that this will look a lot like the normals on a sphere, because you have a continuous gradient of outward vectors that smoothly covers every direction.

If you now light the cube, using the normals as discussed above, so that a head-on light is bright and a shallow light is dim, you'll see that the faces are no longer lit with a single constant brightness, but rather a gradient that simulates the surface of a sphere.

So you'll see an object that has a cube-shaped silhouette, but a surface that's lit like a smooth sphere.

This will look very strange, but as you raise the number of faces you use, to 20 or 100, you'll have an object that looks basically just like a sphere, except the outside edges are slightly angular if you look very closely.

(You can see an example of this object here: http://www.twodee.org/weblog/wp-content/uploads/2011/02/sphe... , taken from this article: http://www.twodee.org/weblog/?p=594)

This is the same sort of cheating that the linked article is about, but the linked article is cheating even harder -- instead of merely blending the normals between adjacent surfaces, they're actually reorienting the normals of one surface all the way "around the corner", to point in the same direction of the normals on the adjacent surface. This means that when you look directly at the corner between the two surfaces, it will look like one surface extends around the corner, which gives the illusion of a notch in the corner.

Can you talk more about the extortion?

Given that Uber apparently drops drivers that get below a 4.5 or something, I wouldn't expect much trouble from them...

If you lie to someone in order to induce them to give you valuable private information, that's fraud, not "information you carelessly broadcast". If a Stingray says "Hi, I'm a cell tower, connect to me!" that's not at all the same as mere passive recording.

(Now, mind you, mere passive recording of cellphones without a warrant is _also_ an illegal search as far as I know -- or would be if there were still unencrypted data to passively record. But that's totally unrelated to what a Stingray does.)

The bitcoin ASICs I'm aware of can't be used for cracking SHA256, because the function they compute is SHA256(SHA256(f(params)), where f is a function that stuffs some parameters into a bitcoin-specific block template before hashing.

Are you sure that 1-year cliff is a good idea for founders? A lot can happen in a year, and founder disputes are A Thing; and it sucks for the person with 49.9% to be in a position where any unresolvable disagreement in the first year means he walks away with nothing.

So far I've been very underwhelmed by btsync; I couldn't seem to get it to actually start syncing between my devices. I've gotten it to sync test data and the like just fine, but as soon as I ask it to sync large folders of real data, it fails to connect, and it doesn't have any real logging or debugging features, nor any 'force' buttons, so there doesn't seem to be anything to be done about it.

You're absolutely right, I failed to read closely enough.

Note that your link doesn't appear to say anything about that. Might you be seeing an internal revision? (The current revision visible to the public appears to predate the policy change; perhaps a new version was meant to be pushed but has not been.)

I don't know if you'll see this comment; as far as I'm aware there's no way to passively monitor HN for replies.

But if you do see it, please check my reply to dang's comment for a thought on why I'm so angry about this.

I apologize for the unnecessarily confrontational tone of my comment.

(Because of HN's lack of comment notification system, I had no idea it was so controversial until I happened upon it again just now.)

Its significantly positive score does suggest to me, though, that there is a lot of sympathy for my position, and the angry way I stated it. In fact, it's one of my most upvoted comments. (And assuming it's received a significant number of downvotes for being unnecessarily rude, I suspect it's my most-upvoted comment of all time, in gross numbers.)

The claim that I'm complaining about is basically outright fraud, as far as I'm concerned -- they're advertising their startup as being different from other services, on the basis of having a quality they have absolutely no ability to promise (that of being guaranteed not to terminate service out from under the users without warning.) The only mitigating factor is that they are almost certainly misleading themselves as much as their users. Whether this is noble, I leave as an exercise for the reader.

The MUD I played back in high school (Frontier, found on frontier.mudservices.com:7680) has evidently been dead for some years now. When I first played it ~2004 it was already on its second wind, having died and been revived once previously by new management.

I vaguely wonder what happened to it.

I think the README mis-describes the performance comparison with the naive "huge array of samples" method.

In particular, both methods are O(1) after a precomputation step. What the Alias Method appears to get you is a significant reduction in space required.

Your claim that your product will not be discontinued is worthless.

If you're so sure of that, put your money where your mouth is, and offer an SLA with a contractual minimum duration you promise to keep the service up, and significant monetary penalties if you fail to meet it.

Doesn't sound like a good idea? Then shut up.

Perhaps it's time to forward your materials to the FDA? (Or whatever the appropriate regulator is for your device.)

I see you're getting downvoted for this, but the only thing that's wrong with it is that you're a bit late on the schedule. If you go back to the 2011 era it was absolutely true.

As an owner of a Bitcoin ASIC, I don't _think_ they're usable for general SHA-256. I think they're all specialized for SHA-256(SHA-256(X)), where X must be a packet in a bitcoin-specific format.

I think you've inadvertently explained part of the problem here.

The people who have the most influence on SO are not the ones most interested in using it for questions, but the ones most interested in improving their "score". It should be obvious that this outcome is both perverse and extremely difficult to avoid.

One should not under any circumstances deal with Mircea Popescu. See the other reply to you in order to understand why. (Mircea is the owner of BitBets, the creator of the "no refunds" policy in question, and also the owner of the MPOE-PR account that defends the policy on the forum. I think Mircea may deny that last part, I forget, but it's not actually in question.)

The engineers -- the ones down in the trenches -- practically revolted over it.

The decision was purely from On High, i.e. managerial. It came down from the highest levels.

Thanks for commenting with an explanation of your moderation (not just this time; I've seen you do it repeatedly.) It really makes a huge difference to how moderation on HN is perceived. (If you look at my comment history, you'll see that I have a history of somewhat passive-aggressive comments about the opacity of HN moderation. I really do appreciate the change tremendously.)

For most miners at the present time this is not true. The reference code does not do this, and most mining pools are running something close to the reference code. (And furthermore, most such txes will never transit the network in any case, because most nodes will refuse to relay them.)

In Chrome on OS X, either two or three fingers will both scroll and change page.

Right idea, but 4 is not enough.

  echidna:~ gwillen$ wc -l /usr/share/dict/words
    235886 /usr/share/dict/words

You get something like 17 bits of password strength per word, depending on the size of your dictionary. (The relevant xkcd estimates more like 11 -- which makes sense because /usr/share/dict/words has a lot of obscure words, shitty words, and alternate forms of words, that you would probably exclude when generating a password.)

So if you want a passphrase that's secure against brute force, you'd want more like 7-12 words.

The HN mods typically sanitize titles to match the title of the underlying article. (Except when they don't, of course.) They are generally unwilling to explain themselves, even when their edits are clearly making the title worse.

It's honestly a mystery to me why the system even lets submitters specify titles.

Hah! In all the times I've read this story, I've apparently always missed the punchline. Thanks.