leiroigh

21 ms·

31.
by leiroigh 7y ago
Of course it is trading one assumption for another. It basically comes down to "absent other specifications, one state is as probable as another". As good Bayesians, we should reject this sentence: We always, always need a prior;
32.
by leiroigh 7y ago
No, deriving the second law from Newton+Liouville is easy (in the form of "mathematically plausible pseudo-theorem", that's late 19th century physics). If you don't adopt Liouville measure as axiom, then it is impossible
33.
by leiroigh 7y ago
Ah, ok, with "purely classical phenomenon" I meant "the classical limit h->0 also exhibits an arrow of time; observing the arrow of time and thermodynamics is not enough to distinguish between a classical and quantum unive
34.
by leiroigh 7y ago
Yeah, QM offends my aesthetic sensibilities as well. I cannot but shake my head in disgust at creation. But it does not get better by attempting to derive the Born rule. Of all things to get offended by, why do people single out the Born ru
35.
by leiroigh 7y ago
It appears that a similar issue crops up in the interpretation of classical mechanics. That is, suppose we have our laws of classical mechanics. Hence, we have a nice way of propagating states along time. Now we want to understand thermodyn
36.
by leiroigh 7y ago
Not really. ZFS dedup is just brain-dead stupid. Rabin-Karp chunking plus spatial locality / opportunistic dedup allow you to gracefully scale down: you keep a complete table of successfully deduped blocks, plus a small fraction of oth
37.
by leiroigh 7y ago
One of the best policies I ever witnessed: There was a second guest network with internet and nothing else for guests/consultants and facebook/twitter/porn (the company just paid for internet twice). Employees had a second cr
38.
by leiroigh 7y ago
While I agree that the solution is brilliant (and obvious), I'd like to note that freeriding (Schwarzfahren) is not statistically more expensive in monetary terms (the fine is not that large, controls are not that frequent and tickets
39.
by leiroigh 7y ago
Phish tests need to be fair to people who actually understand something about security. "Opening an email" is not actually an issue (spearphishers that sit on drive-by 0-days in current browsers or email programs are not a threat
40.
by leiroigh 7y ago
Lot's of people using chromium or enterprise chrome. As long as ublock origin (then incompatible with mainline chrome) is not pulled from the webstore, and as long as there is a simple #ifdef that most linux distros can toggle, I am n
41.
by leiroigh 7y ago
Mathematica also has god-tier documentation and an interesting and very unique programming language. I personally hate programming in Mathematica, but it is interesting and deserving of an open implementation.
42.
by leiroigh 7y ago
Well, apart from the incredible standard library, there is an interesting programming language. What I think is missing is a stripped down open source implementation / interpreter. What Gnu Octave is to Matlab. That stripped down varia
43.
by leiroigh 8y ago
I know, and I am super grateful for all the cool things you and the other people in the scene develop, and for the effort you go to to actually ship out modkits. You do what's fun and I won't ask for more. Don't get me wrong,
44.
by leiroigh 8y ago
The old x60 mainboard was a single part. I don't know about the x61. It's not a big issue, but makes assembly pretty annoying because everything sticks out: There is no screw-hole from the lower casing in the extra attachement, an
45.
by leiroigh 8y ago
Damn, 15 minutes after getting pointed to your mods the disappointment :) Also, thanks for all your guides on the thinkpad forums. Would you mind writing a small idiosyncratic guide about "51nb procurement for dummies"? Like: Buy
46.
by leiroigh 8y ago
I am a happy owner of an x62. This article almost convinced me to buy an x210 next. Minor gripes: (1) The dispayport-hdmi thing, (2) the mainboard is split (sound + rhs usb are on a separate board) and the small part has pretty flimsy attac
47.
by leiroigh 8y ago
I don't get the complaints. As far as I understood (and ietf appears to agree), eTLS is not a protocol, it is a (server-side) implementation variant of TLS. And it is a universal construction: For any cryptographic protocol, one party
48.
by leiroigh 8y ago
And nobody uses this terrible mis-feature in practice, everything runs via 64 bit xmm registers. Rightly so, because programmers want their optimizing compiler to decide when to put a variable on the stack and when to elide a store/loa
49.
by leiroigh 8y ago
For some reason, julia has no ascii infix xor. That is, you can write `xor(a,b)` or `a ⊻ b` (`^` is taken for "raise to power").
50.
by leiroigh 8y ago
Above example is from julia. Julia structs have C layout and have no member functions. The analog of "classes" in julia are "abstract types"; e.g. `AbstractArray{T}` is anything that implements the AbstractArray interfac
51.
by leiroigh 8y ago
Good point!
52.
by leiroigh 8y ago
>Isn't this a fairly common pattern? Depends on the application. It is certainly not uncommon that queries fired at close times share a prefix. >How would we go about quantifying the preferring locality-ness of a query distributi
53.
by leiroigh 8y ago
>reasonable worst-case behaviour (think real-time apps, or adverserial scenarios) I'm not sure I agree. When faced with potentially adversarial input, a hash should be keyed. And, depending on the specific tree, you can probably cau
54.
by leiroigh 8y ago
It is unsurprising that a structure designed for range queries loses out against hashes for random point look-ups. Thanks for posting that addition and very nice link. It is completely unclear why authors in this field even attempt to compe
55.
by leiroigh 8y ago
FWIW, the Gault Millau is imho vastly preferable to the Guide Michelin. Cares more about food quality than decor, and is actually reliable on its lower spectrum (< 1 star Guide Michelin recommendations are worthless, and 1 star is alread
56.
by leiroigh 8y ago
One-based indexing obviously means that ranges are inclusive on the right, and you iterate by `for i = 1:length(x)`. That being said, I also abhor 1-based indexing in julia. Well, you get used to it, and it is imho a small price to pay for
57.
by leiroigh 8y ago
:) My super high-level partial understanding of Judy is the following: Start with ART, which is pretty simple. Then do the obvious improvements: First, we want fast access to "element number N in sorted order", i.e. you also store
58.
by leiroigh 8y ago
So I was wondering... are there sensible structures that combine hashmap and tree by having a double index? Reason: Ordered access or range queries need a (radix-)tree. So insertion and removal need to pay for the comparatively slow tree se
59.
by leiroigh 8y ago
Cool, looking forward. Please put Judy in as well :) Seriously, the 17 year old judy is still pretty good, despite it's lack of use of cool vector load / compare / etc instructions for quickly traversing tree nodes that are t
60.
by leiroigh 8y ago
Yeah, an easier formulation would be: noveil(...) to mark pathes we plan to use (fails on veiled pathes), and a subsequent veil() that irreversibly removes access to everything that was not protected by a noveil(...) call between this and t

More ›