gregeganSF

gregeganSF, 1 hour ago to random

It makes sense that MacOS has a few words in its spellchecking dictionary that aren’t in the Oxford Dictionary used by the Dictionary app. “Genevieve” is one example.

But it makes no sense to me that some words the Dictionary app knows are redlined by the spellchecker.

gregeganSF, 1 day ago to random

Two Chinese editions arrived in the mail today. PERMUTATION CITY is the first of my novels to be published in Chinese; the previous books have all been short story collections (like the other one here, OCEANIC).

gregeganSF, 1 day ago to random

“exeme” sounds like some new-fangled technical term from the cutting edge of genetics or information theory ... but actually it’s just an archaic Scottish word meaning to release or exempt.

https://en.wiktionary.org/wiki/exeme

gregeganSF, 4 days ago to random

After more than three decades of development, you’d hope Mathematica could at least get the basics right … but no, here are three strings that sort into a different order if you append an identical suffix to all of them.

gregeganSF, 7 days ago to random

Breaking: OpenAI reaches deal with the estate of Douglas Rain.

luigithirty, 9 days ago to random

the file transfer program that comes with TriSoft CP/M-68K is so old it calls XMODEM the Ward Christensen protocol

lana, 10 days ago to random

Since you're not supposed to eat grapefruit while taking meds as it enhances the effects of some meds to dangerous levels, why aren't those meds sold in smaller amounts premixed with grapefruit?

gregeganSF, 10 days ago to random

Both popular science articles and many textbooks describe the Casimir effect — an attractive force between two closely spaced neutral conductors — as being due to the exclusion of some of the zero-point modes of the electromagnetic field, whose wavelengths are greater than the separation between the plates. In this account, the vacuum between the plates has less energy than ordinary vacuum.

A formula based on this model seems to agree reasonably well with the measured result. This has encouraged people to proclaim that “vacuum energy is real” and (sometimes) to suppose that this model can be pushed much further to predict larger regions of greater negative energy.

But in fact this force can be modelled as arising between the plates in a far more conventional way, related to van der Waals forces, which makes no reference at all to “vacuum energy”, and the formula based on zero-point modes can be seen as merely approximating a more accurate formula based on the material properties of the plates.

H/T Philip Ball on Twitter.

https://arxiv.org/abs/hep-th/0503158

gregeganSF, 12 days ago to random

I guess the best way to get spectacular improvements in the performance of your code is to do something spectacularly inefficient in the first version.

I’ve been working on a project that needs to perform computations on a quantity X along with several dozen of its derivatives wrt a parameter t, i.e. X'(t), X''(t), ... This is a form of “automatic differentiation”, but with a large number of derivatives.

There are simple rules for performing basic arithmetic on these vectors of derivatives; addition is trivial, and multiplication only grows linearly with the order of the derivative in the number of terms in each formula.

But the derivatives of 1/X and √X, written as sums of products of powers of derivatives of X, scale horrendously: the number of terms for the nth derivative is equal to the number of integer partitions of n:

1, 2, 3, 5, 7, 11, 15, 22, 30, 42, 56, 77, 101, 135, 176, 231, 297, 385, 490, 627 ...

And for a while, I thought, well, that’s just the way it is, I’ll have to live with it.

But luckily that turned out to be very naive! In fact, you can just write down the formula for the derivatives of a product, X(t)Y(t) = 1, and then solve that system by back-substitution to get all the derivatives of Y from those of X. Similarly for the square root, from Y^2(t) = X(t).

So the payoff for being dumb in the first place was the glorious feeling of seeing my code now running exponentially faster than before!

johncarlosbaez, 12 days ago to random

Chemistry is like physics where the particles have personalities - and chemists love talking about the really nasty ones. It makes for fun reading, like Derek Lowe's column "Things I Won't Work With". For example, bromine compounds:

"Most any working chemist will immediately recognize bromine because we don't commonly encounter too many opaque red liquids with a fog of corrosive orange fumes above them in the container. Which is good."

And that's just plain bromine. Then we get compounds like bromine fluorine dioxide.

"You have now prepared the colorless solid bromine fluorine dioxide. What to do with it? Well, what you don't do is let it warm up too far past +10C, because it's almost certainly going to explode. Keep that phrase in mind, it's going to come in handy in this sort of work. Prof. Seppelt, as the first person with a reliable supply of the pure stuff, set forth to react it with a whole list of things and has produced a whole string of weird compounds with brow-furrowing crystal structures. I don't even know what to call these beasts."

https://www.science.org/content/blog-post/higher-states-bromine

gregeganSF, 14 days ago to random

Wow, Thomas Hales and Koundinya Vajjha have proved Mahler’s First Conjecture! (That’s Kurt Mahler, not Gustave.)

https://arxiv.org/abs/2405.04331

Mahler’s First Conjecture says that the centrally symmetric convex shape with the worst possible packing ratio is made up of straight lines and arcs of hyperbolas, like the smoothed octagons shown in the animation below (demonstrating a 1-parameter family of slightly different packings all with the same density).

Whether these smoothed octagons are, as Karl Reinhardt conjectured, the actual worst case is still an open problem.

A bit more detail on Reinhardt’s conjecture in this article by @johncarlosbaez :

https://blogs.ams.org/visualinsight/2014/11/01/packing-smoothed-octagons/

and this thread by Koundinya Vajjha on Twitter:

https://twitter.com/KodyVajjha/status/1790211313618362848

A 1-parameter family of packings of smoothed octagons.

11011110, 17 days ago to random

After seeing several recent student presentations on quantum algorithms, it occurs to me (as I'm sure it has occurred to others) that we desperately need some way of communicating these concepts at a higher and more intuitive level than slide after slide of tensor algebra equations. If we tried to describe classical algorithms but our language was limited to circuit diagrams, how far would we get?

gregeganSF, 18 days ago to random

Feeling a strange compulsion to watch this again:

https://en.wikipedia.org/wiki/BrainDead

gregeganSF, 20 days ago to random

I enjoyed “Manhunt”, since I knew almost nothing about the history of the period, or Edwin Stanton. The mixture of 19th century and modern language didn’t bother me, but I did laugh when Patton Oswalt used air quotes

gregeganSF, 20 days ago to random

TIL that I cannot include a double quote or an asterisk in anything I write to the Australian Taxation Office in their secure messaging system, but backticks are OK.

I’m sure there are Reasons.

Valid characters are: A-Za-z0-9 - .? / ' ! $ % & _= \ : ; ( ) { }[ ] @ # ` | , space

gregeganSF, 21 days ago to random

“Dream Scenario” fell flat for me on every level: comedy, metaphysics, metaphor, social satire — there wasn’t a trace of energy or inventiveness in the whole 90 minutes. I would have had a thousand times more fun rewatching “Nightmare on Elm Street” or “Being John Malkovich.”