ezyang

9 ms·

91.
by ezyang 14y ago
Yeah, this happened to me when I was integrating Ur/Web and Haskell; Ur/Web generated C and wanted to handle linking, but to link against Haskell libraries I needed GHC to do everything. There's a happy ending to the story, though: we patch
92.
by ezyang 14y ago
I don’t think cascading is necessarily a problem; rather, HTML/CSS lacks good encapsulation features. Actually, this has been well-recognized by browser vendors, who want to use HTML and CSS to implement things like HTML5 video player UIs,
93.
by ezyang 14y ago
No videos. Also, while D3 is the "official framework" of CS448B, most of the lectures are not about D3; there was one tutorial on the topic and the rest is mostly theory. (Also, this was probably the last year the course will be run in its
94.
by ezyang 14y ago
The best way to cue that it is clickable is to have it "bounce" when you mouse over it, indicating that it is hidden on purpose and can be expanded.
95.
by ezyang 14y ago
“Words, they mean things!” I’ve historically been pretty bad about using the right words for things, and I’ve had colleagues chew me out repeatedly for it. But in this case, I think the intended meaning is pretty unambiguous.
96.
by ezyang 14y ago
Eratosthenes is basically NaCL taken to the logical extreme. NaCL itself doesn't go far enough, because it still exposes the traditional browser APIs, when really they should also be run in NaCL. They are fundamentally the same approach, bu
97.
by ezyang 14y ago
Because it's untrusted: it runs in the sandbox.
98.
by ezyang 14y ago
The difference is that Java doesn't let you use WebKit directly, whereas native code lets you use both, if you so please.
99.
by ezyang 14y ago
New processors are already at a huge disadvantage; one problem is that it takes a huge amount of engineering effort to match the performance of existing processors, even if you do a lot better, another problem is, well, you actually have to
100.
by ezyang 14y ago
The difference is that in canvas, you have to reimplement a rendering engine; in native code, you just run WebKit. This is key! Cross-website communication is certainly tricky, check out section 4 of the technical report.
101.
by ezyang 14y ago
x86 versus ARM is not a big deal at all; all you need to do is have cross-compiling setup, which for any VM you care about is already well supported. (Note that we are not recommending you start coding your web pages in assembly; your stack
102.
by ezyang 14y ago
The Java sandbox was actually really close to something that worked well. If you look at the Java virtual machine, it's actually quite simple, and very easy to ensure that nothing "bad" happens. Where Java lost the battle was in the provis
103.
by ezyang 14y ago
Sure, the web has given us an awesome UI toolkit. Ship that to your users. You're conflating a lot of issues in this post, many of which are orthogonal to Howell's idea, which is making it difficult for us to talk about the core technical i
104.
by ezyang 14y ago
I recommend checking out Section 4 of the linked technical report, which discusses how to recover traditional interface elements of the traditional web.
105.
by ezyang 14y ago
You don't need to reimplement it, you just use whatever library you were using before (but linked with the browser). The point is that the stack doesn't change, just who provides it.
106.
by ezyang 14y ago
This post received a lot more attention than I thought it would; if I had known, I might have explained a few things more carefully. If you didn't immediately dismiss it as stupid and found it at least a little thought-provoking, I highly r
107.
by ezyang 14y ago
Ubuntu is a Debian-based distribution, and receives many of its software updates from Debian.
108.
by ezyang 14y ago
I wrote a piece a while back, which I think retitled would make a good companion article, "Why I am tired of merging pull requests" < http://blog.ezyang.com/2010/05/i-hate-patches-confessions-of... ;
109.
by ezyang 14y ago
Hidden inside this fascinating screed is an announcement about the "Google Grok" project, which appears to be something of an Eclipse-killer for dynamic languages. It's good to hear that Google is working on this problem, and I'm interested
110.
by ezyang 14y ago
Are we still talking about Sybil attacks?
111.
by ezyang 14y ago
You could this, but you would then be unable to mix any further, so the first mix must be sufficient to give you sufficient anonymity for the transaction you want to carry out. The resource requirements are not well studied, since none of t
112.
by ezyang 14y ago
Correct, the usual problems with Sybil attacks still apply: the point is that the attacker doesn't gain anything from faking addresses.
113.
by ezyang 14y ago
Yes, transaction fees are a big problem. One possibility is to not pay any fees: currently, there is a social convention that if all outputs are 0.01 BTC or larger, and so if we're not mixing too many addresses together, we can do the trans
114.
by ezyang 14y ago
This doesn't work. If you know how many coins an output address is waiting to receive, all you need to do is look at which input address is submitting that many coins, and you have a matching.
115.
by ezyang 14y ago
The sentence is a little unclear, but it is a little clearer if you consider things from the perspective of an organized attacker who controls some number of parties in the transaction. Even if the attacker messes around with the addresses
116.
Secure multiparty Bitcoin anonymization(blog.ezyang.com)
70 points by ezyang 14y ago|15 comments
117.
by ezyang 14y ago
I agree that it is important to recognize the limitations of this approach. As a high schooler (and even as an undergraduate in mathematics), you learn all sorts of formal systems. As a mathematics researcher, one of your charges is inventi
118.
by ezyang 14y ago
In the case of mathematics, there is a fundamental sense in which the symbols and their manipulations are what you're studying. Think of it as if you were an archeologist: the symbols are the artifact of study. It's only half of the pictu
119.
by ezyang 14y ago
Brett Victor writes: When most people speak of Math, what they have in mind is more its mechanism than its essence. This "Math" consists of assigning meaning to a set of symbols, blindly shuffling around these symbols according to arcane r
120.
by ezyang 14y ago
I think Mercurial also misses the mark for uploading patch sets which might evolve / change over a few iterations. The workflow usually looks like one user is working on only one set of changes at a time, and in order to show them to upstr

More ›