cscheid

So this paper appears to have been published on 2017-12-23 (according to https://link.springer.com/chapter/10.1007/978-3-319-70972-7_...), which was ahead of the Specter and Meltdown disclosures. Truly unfortunate timing for the browsers implementing mitigation, huh.

It seems that one of the covert channels described here has been fixed (SharedArrayBuffers) but many of the other ones have not. The passive reading of data from timing-based side effects (like CSS animation as described in the paper) seems particularly hard to avoid in general. HTML5 video will have the same vulnerability, I bet.

yup, in cartography these are known as trap streets

> It's the extreme end of the situation, to be certain, but extremes are a great place to examine any system.

Although there's some dissent, in law the typical expression you hear is the opposite: https://en.wikipedia.org/wiki/Hard_cases_make_bad_law

neato uses distance embedding methods. The paper linked describes Sugiyama-style rank embeddings as implemented in dot. 1) They're incomparable. 2) the theory behind neato was published in 2004 (http://www.graphviz.org/Documentation/GKN04.pdf).

There are research groups (I'm thinking of Tim Dwyer's specifically) that are specifically trying to improve the situation. When you say "it's just a pile of incompatible heuristics", it sounds like you haven't taken the time to try and implement a better solution.

Graph drawing is the kind of thing that seems trivial, right until you try to work on it.

(disclaimer: I'm clearly biased here since I worked with the people who wrote graphviz.)

That's because the program is looking up every item. It's _right there in the text_:

> Note that this time is measuring the time taken to look up every element in the collection, rather than just looking up a single element.

http://www.lihaoyi.com/post/BenchmarkingScalaCollections.htm...

If you have a hard time remembering exactly how Markov inequality goes (like I do), there's a great mnemonic from which you can construct the general version:

- if the average person is 6' tall, than at most 10% of the people are taller than 60'.

Thanks - I should have known better than to blurt that out.

The first person to make a connection between categories and programming was arguably Moggi, in 1991. That's 20 years after SHRDLU.

> Can somebody please explain why the employers are being held legally liable?

Because this is... illegal.

Page 3 of the complaint document has references: https://www.documentcloud.org/documents/3722345-UploadVR-law...

Presumably, moving to an infrastructure like this (everything is incremental) is the biggest difference between Old compilers and New compilers, because of the ubiquitous IDE. I think I remember watching a talk by Anders Hejlsberg about this.

Reading the fourth entry on parenthesis matching made me wonder whether one could store, in the monoid, partials views into the table that is generated by CYK parsing: https://en.wikipedia.org/wiki/CYK_algorithm

I love the idea of using monoids like they're described in the blog series, but the examples suggest that there's a certain amount of non-generalizable cleverness that goes into defining each monoid. Could you do CYK subtables inside the monoid, so that people can define arbitrary CF grammars, as long as they're in Chomsky normal form?

(more spoilers)

You do make an interesting point which I didn't emphasize: Aurora also makes a moral argument against generation ships.

But I think things get more complicated still. If you accept the book's argument, accept cryogenic sleep, but also accept that Earth is something like the biggest generation ship we have, then where do we stand? If we accept that cryogenic sleep solves that part of the moral argument, then we kind of walked into Woody Allen's plot for Sleeper (that is, why not sleep right here on Earth until things improve?) These are really tough questions to consider.

Needless to say, I think Aurora is the deepest, most introspective work that KSR has produced.

The unfortunate part about Aurora is that it is "optimistic" only in the context of all of the Red Mars trilogy, and 2312, and really only in the last chapter, where...

SPOILERS below

the characters go "well, eff it, I guess we have to make it work here on Earth". It seems to me that, in the arc of KSR's work, Aurora is the effective acceptance that humans will have to make Earth work for them, for better or worse, even as they collectively wrecked the planet. And that seems a nice segue into New York 2140.

mocha has a naughty german word in its middle, I'm fairly sure.

Search for "B-splines knot insertion", which will enable what (I think) you want to do.

The best introduction to optimization methods book I've ever read is "Optimization by Vector Space Methods" by Luenberger. This is the book that opened my eyes to the generality of least-squares methods, the actual geometry of the problem, and how to apply it extremely generally. I read this before Boyd and Vanderberghe's "Convex Optimization", and I'm glad I did - it goes much more deeply into the intuition and exposition. It's not as well known as it should be, but it's a complete gem.

I'm pretty sure you're wrong. These asymmetric crypto keys, with which it takes a long time to encode/decode, are used only at handshake to securely negotiate a per-session key that's used in symmetric crypto, for which encoding and decoding is extremely fast.

You are right, of course, but from the perspective of an engineer, these two (admittedly broad) laws are very good things to have around.

"Essentially, all models are wrong, but some are useful" and all that.

You need log transformations on both axes for this to work.

Hey, so gamma is not a logarithmic response. You claim that the delta you use in Figure 2 is a ratio, but your code, https://github.com/johnnovak/johnnovak.site/blob/master/blog... uses a fixed power. These are not the same thing.

f(x+eps)/f(x) ~= eps f'(x)/f(x) + 1

f(x) = x^2.2 f'(x) = 2.2x^1.2

f(x+eps)/f(x) ~= 1.2 eps/x + 1

Human response to light is not particularly well-modeled by a logarithmic response. It's --- no big surprise --- better modeled by a power law.

This stuff is confusing because there's two perceptual "laws" that people like to cite: Fechner-Weber, and Stephens's. Fechner-Weber is logarithmic; Stephens's is a generalized power-law response.

(Minor nitpick. Presburger arithmetic can represent integers and is decidable: https://en.wikipedia.org/wiki/Presburger_arithmetic You were probably looking for Peano arithmetic.)

Jason Davies has a really great interactive demo of Bezier curves: https://www.jasondavies.com/animated-bezier/

Yes. The bigger your sample size, the bigger your ability to see small effects. A test with 1000 elements can only tell apart big effects; a test with 250000 can tell apart much smaller ones. This is the idea of https://en.wikipedia.org/wiki/Statistical_power

Just FYI, it would have been helpful if you had disclosed that you're the author of HTTPie.

Mods, can we instead link directly to Cal Newport's page? http://calnewport.com/books/deep-work/

This talk is great! And if you watched it all the way to the 6:30 mark, where he says "you can't have negative amounts of light", then you'll really enjoy chimerical colors: https://en.wikipedia.org/wiki/Impossible_color

The trick is you exploit retinal fatigue in order to perceive colors both "above 100%" and "below 0%". Wild stuff.

You'll have a lot of fun looking at https://en.wikipedia.org/wiki/Embodied_cognition

Huh, thank you. I missed it back then, and assumed that when the submission went through it meant that the dupe detector okayed it.

Yes, that is certainly a tension here. I think it's easier to think of the paper as two broadly different subparts.

The first part is: "regardless of what you think, if something you do rearranges power, then it will become political". This is more a statement of fact than anything else.

The second part is: "now that I hope to have convinced you that crypto is necessarily political, here's the moral stance I would prefer for you to follow."

We use "data cubes", as per Jim Gray's paper, http://web.stanford.edu/class/cs345d-01/rl/olap.pdf.

When I say "screen", I mean (e.g.) your laptop's screen resolution. If you're going to display a heatmap on a screen with p pixels, we (roughly) touch only O(p) memory cells on query time, independently of the dataset size.

When you say "you don't see the uniqueness", I'm not sure what you're comparing it against, so I can't say anything more. You mentioned gkd-trees earlier: is that the comparison you mean? In that case, these are two completely different data structures. For example, (at least as described on the 2009 siggraph paper), you can't subset on a categorical dimension of the pixels. In the case of the data structure we created, we can report (for example) a heatmap of all geolocated tweets generated by an iPhone, or all geolocated tweets generated by a windows phone, or all geolocated tweets irrespective of device, without having to scan the 200M tweet database.