@4raylee@mathstodon.xyz
@4raylee@mathstodon.xyz avatar

4raylee

@4raylee@mathstodon.xyz

Culture is kept in our conversations, for some loose definition of 'conversation'.

I work in software with a team of physicists, MEs, EEs, and software engineers, all working to protect the biosphere.

I have a background in physics and math to the BSc level. My posts tend to be a mélange of all the above, and a bit more.

This profile is from a federated server and may be incomplete. Browse more on the original instance.

dpiponi, to random
@dpiponi@mathstodon.xyz avatar

One of the weirder bugs I've experienced: you know how you're always being told to make sure caps lock is off when you enter your password? My Mac is currently enabling caps-lock at login and you can't disable it. It took a long time to deduce this was the problem but surprisingly I was able to log in after going round a few loops and realising an obvious trick...

A known problem with a venerable history: https://iboysoft.com/tips/macbook-stuck-on-caps-lock.html

4raylee,
@4raylee@mathstodon.xyz avatar

@dpiponi I always remap caps lock to be another control key, just to go easy on my little finger. Side benefit, I guess I get to dodge mysterious and annoying bugs.

ColinTheMathmo, to random
@ColinTheMathmo@mathstodon.xyz avatar

Well .... this is something ...

https://mathstodon.xyz/@susseratal@mastodon.gamedev.place/112378580389464806

4raylee,
@4raylee@mathstodon.xyz avatar

@ColinTheMathmo I love that it recognizes mouse clicks, pinch and zoom gestures, too!

danderson, to random
@danderson@hachyderm.io avatar

Okay well first thing I've learned with SELinux: it's kinda broken on Universal Blue atomic distros right now, due to a quirk of how OCI-based ostree images get built. Binaries in such images don't get the right SELinux labels applied to them, and so installing anything privileged like libvirtd or incus results in broken programs, unless you turn off selinux enforcement :/

iiuc, Silverblue is not affected by this since it doesn't (yet) use the OCI based ostree mechanism.

4raylee,
@4raylee@mathstodon.xyz avatar

@danderson now I want a distribution that’s an overlayfs of merkle trees in a trench coat.

johncarlosbaez, (edited ) to random
@johncarlosbaez@mathstodon.xyz avatar

It's time to have that talk with your kid. About quantum mechanics:

https://www.smbc-comics.com/comic/the-talk-3

4raylee,
@4raylee@mathstodon.xyz avatar

@lutzray indeed yes! https://physicsworld.com/a/the-double-slit-experiment/

@johncarlosbaez

ColinTheMathmo, to random
@ColinTheMathmo@mathstodon.xyz avatar

Is this deliberate? The "mistake" means it's getting a lot of free repeats.

Is it AI generated?

So many questions ...

https://universeodon.com/@georgetakei/112360777460218175

4raylee,
@4raylee@mathstodon.xyz avatar

@ColinTheMathmo it exists. Different label has the same mistake, in a different problem. Perhaps they aren’t into carrying ones.

https://itasia.wine/product/cabernet-sauvignon-correcto

danderson, to random
@danderson@hachyderm.io avatar

dangit, my inner data structure has a structural fault because of rust ownership semantics.

Conceptually, the inner struct is a binary tree where inner nodes can carry a value, and leaves can carry a value or a child tree. If you need a leaf to hold both a child and a value, you store the child and move the value to the child's root node.

Conceptually again, lookups walk down this tree-of-trees looking for the node representing the lookup key, and nearest self-or-parent value is the result.

4raylee,
@4raylee@mathstodon.xyz avatar

@danderson another approach I’ve used is to give up on using pointers for representing edges in graph algorithms, and using indices into a backing array instead. This can work well as long as you never need to return a language-native pointer to an element. No idea of that’s an easy option here — I haven’t read the algorithm.

christianp, to random
@christianp@mathstodon.xyz avatar

Who can find a street with a number in its name that's bigger than the number of any building on that street?

4raylee, (edited )
@4raylee@mathstodon.xyz avatar

@christianp “Infinite Loop”, Cupertino, CA

4raylee,
@4raylee@mathstodon.xyz avatar

@christianp Fair enough :-) How about Billion Rd, Memphis, TN, USA?

BartoszMilewski, to random
@BartoszMilewski@mathstodon.xyz avatar

I have a confession to make: When I was studying physics I flunked thermodynamics, and it's still a bunch of meaningless formulas to me. I'm watching Susskinds lectures now, and I'm none the wiser.

4raylee,
@4raylee@mathstodon.xyz avatar

@BartoszMilewski I found the arguments and derivations more enlightening than the final formulas. RIP Herb Kroemer who just passed, whose book with Kittel I still have on a shelf.

astro_jcm, to random
@astro_jcm@mastodon.online avatar

Happy #PiDay to those of you who don't follow ISO standards or who live in non-euclidean spaces with negative curvature.

(Yes, YES, I know that the value of #pi itself doesn't change with geometry, don't ruin my joke)

4raylee,
@4raylee@mathstodon.xyz avatar

@astro_jcm ISO 8601 says today is 2024-03-14 so live it up and have some pie.

ColinTheMathmo, to random
@ColinTheMathmo@mathstodon.xyz avatar

Oh ... crap.

Another friend lost to cancer.

RIP Richard A Parker.

We'd fallen out of contact so I didn't know. I'll remember working with him on a project about parallel computers, writing code with him to communicate over a mesh of low powered machines, and playing backgammon.

Some time ago I wrote a blog post about a lesson I learned from him, and which I've used many times over the past 30 years.

I'll remember Richard for a long time.

4raylee,
@4raylee@mathstodon.xyz avatar

@ColinTheMathmo back-pressure in distributed systems is pretty important, as is graceful degradation. I’m sorry for your loss.

johncarlosbaez, (edited ) to random
@johncarlosbaez@mathstodon.xyz avatar

Here Ethan Siegel (@startswithabang) gives a clear explanation of the 'multiverse' and why a lot of physicists think it exists.

Actually 'multiverse' means multiple different things. Siegel focuses on just one — which is fine. But beware, there are others, and also lots of people who cluelessly confuse them.

Siegel focuses on the inflationary cosmology, why it could be correct, and how it could create many 'bubbles': regions of the universe that can't communicate with each other. I have my doubts about inflation, more than Siegel, but this scenario seems possible.

There's another meaning of the multiverse, which is that we can think of the universe as a quantum superposition of different approximately classical worlds in which different events occurred. This scenario seems almost unavoidable to me, and it's not in conflict with the first.

Then there's a more speculative version of the multiverse, where these events in the very early history of the universe include the universe settling down to have different approximate laws of physics. For example maybe it's a superposition of states where the Standard Model works the way we see, and states where there are other numbers of forces and other kinds of particles. This seems possible too, but I'd only get interested if someone 1) makes up a precise, consistent theory of physics where this happens, and 2) convinces me that this theory is likely to be correct. So far I don't think even 1) has happened.

An interesting thing about these multiverses is how little they matter. They might someday, but they're not something I spend much time on. There's so much math and physics that's more interesting and practical!

https://bigthink.com/starts-with-a-bang/scientists-think-multiverse-fiction/

4raylee,
@4raylee@mathstodon.xyz avatar

@johncarlosbaez I had already been writing code for years before I started my physics degree. That grounding in computability made me wary of any use of the infinite in physical arguments.

I think the inflationary universe scenario matches our observations pretty well. We seem to have gone through it once, anyway. I'm not sure about extrapolating that to twice or more. And while I understand (statistically speaking) that if the universe were infinite, 10^120 meters that-a-way is another copy of me having a slightly different conversation, I find the arguments for that a bit hollow. (Despite Max Tegmark's best efforts in one of his books.) Our universe appears unbounded but that's a far cry from infinite.

As for the many-worlds interpretation, the quantum world of superpositions sure seems like it's there, and we only experience a roughly classical facade due to our scales of experience, and decoherence.

The multiverse theories that simply declare our constants and laws of physics to be random fluctuations seem like cop-outs.

Dunno. Maybe I'm exposing my experimentalist upbringings here. It makes for fun fiction, though.

4raylee, (edited )
@4raylee@mathstodon.xyz avatar

@johncarlosbaez "The arguments are quite solid if the universe is infinite and homogeneous." Agreed.

"And it's simpler to assume the universe is infinite than to assume it goes on for a huge finite distance..." eh...

I like the infinite in a few different ways it gets used. Infinite as "all the basis vectors", sure (cf. Fourier basis). (Even if our universe appears to only be able to represent frequencies between 1/(the Planck time) and the Hubble constant.)

Infinite as a limit to extract asymptotic behavior, great.

Infinite as a completion, such as the point at infinity in complex analysis, love it.

Infinite processes, such as moving through a continuous path between two points, I start getting twitchy. (cf. your own series on struggles with the continuum.)

Infinite as an infinite amount of stuff? That seems a big assumption. I see approaches to make toy models of an unbounded but finite universe, with a finite amount of stuff as the result of a process through time. But models that simply assume an infinite amount of stuff? I don't even see how to approach writing something precise down about that, other than starting with it as an axiom. That's where I feel like we're heading beyond physical theories and into pure speculation.

"I think some people find the possibility of distant near-copies of themselves upsetting, or at least interesting - to which I'm tempted say, "get over yourself!""

Actual lol. Getting over myself has also been a process through time 😉

"In general the right response to questions about an [...infinite vs finite] universe seems to be letting other people think about them while I do more productive things."

Fair enough. Besides, I have a quaternion paper to read 🙂

4raylee,
@4raylee@mathstodon.xyz avatar

@johncarlosbaez "it's more convenient to assume the universe goes on forever than to assume that extends beyond what we can observe, but not forever - since in the latter case we need to make up a theory for how and where it ends, which is a lot of work for no apparent gain in predictive power."

Good point. The only way I see some believable theory for an in-between size (beyond observable, less than infinite), is if it falls out as a side effect of some other, more fundamental theory which gives greater predictive power about other things we can measure. Not that I have one of those handy.

johncarlosbaez, (edited ) to random
@johncarlosbaez@mathstodon.xyz avatar

GALOIS' LAST LETTER

Galois invented group theory - and in his last letter, written to a friend the night before he died in a duel, he had some interesting things to say about it.

"Möbius transformations" are functions

f(z) = (az + b)/(cz + d)

where z,a,b,c,d are complex numbers. But it's better to think of them as maps from the Riemann sphere to itself. The Riemann sphere is the complex plane together with a point called ∞, which lets us know what to do when we divide by zero.

Möbius transformations are precisely all the transformations of the sphere that preserve angles! That's why they're important, geometrically. We say they form a "group" because doing two angle-preserving transformations gives another one, and for any angle-preserving transformation there's another one that undoes it.

We can copy this story with other number systems replacing the complex numbers. And in his last letter, Galois considered the integers mod p where is a prime. He defined a group of Möbius transformations

f(z) = (az + b)/(cz + d)

where now z,b,c,d are integers mod p. This group acts on a finite version of the Riemann sphere that consists of the numbers 0,1,2,3,...,p-1 together with ∞.

This baby Riemann sphere has p+1 points. But Galois showed that the group of mod p Möbius transformations has a special property when p = 2, 3, 5, 7 or 11, but no higher prime! Namely, it also acts transitively on a set with p points. "Transitively" means these transformations can map any point to any other point.

For p = 5 our group is the symmetries of a dodecahedron, which acts on the set of 5 tetrahedra shown here!

(1/n)

4raylee,
@4raylee@mathstodon.xyz avatar

@johncarlosbaez “We can copy this story with other number systems replacing the complex numbers.”

A feature of conformal mappings that’s kinda nifty is preservation of solutions for potentials. Does this property extend to the quaternions? Assuming I ever find myself needing to find the field for an oddly shaped 4 dimensional E&M problem, can one solve it for the easy geometry and transform the solution?

I have no cases where I need to do this, I’m only asking out of pure curiosity.

4raylee,
@4raylee@mathstodon.xyz avatar

@johncarlosbaez ah, neat! Thank you for the explanation, and the link. Between this and the Pais book I have plenty to keep me busy for a bit.

ColinTheMathmo, (edited ) to random
@ColinTheMathmo@mathstodon.xyz avatar

This must be a solved problem ...

A group I'm in is reviewing and revising a document. It's a word document, because some of them are seriously, seriously technically challenged, and that seems to be the only thing they can cope with.

But making comments on and passing copies around of a Word document is just a completely nightmare, and it ends up ... as you will all know ... a mess of formatting, wrong versions, just ... urgh.

If there is a website with the document visible and little boxes spread throughout into which people can type comments to be collected and collated, that would be much easier.

Is there such a thing?

4raylee,
@4raylee@mathstodon.xyz avatar

@ColinTheMathmo if they already understand Word then it might be a small jump to get them to use Office 365’s online version. Looks like collaboration features are free, and online versions are always in sync: https://www.microsoft.com/en-us/microsoft-365/free-office-online-for-the-web

4raylee,
@4raylee@mathstodon.xyz avatar

@ColinTheMathmo The collaboration tools in word allow for attaching sticky notes style comments to things. Does that help?

4raylee,
@4raylee@mathstodon.xyz avatar

@ColinTheMathmo I worked my way through school doing tech support. At some point there’s no substitute for sitting with a person and doing the workflow with them a few times until it sinks in. (Sometimes a few times across multiple days.)

4raylee,
@4raylee@mathstodon.xyz avatar

@ColinTheMathmo agreed, it’s hard. Google docs is simpler than Office 365 so that seems a better option. Perhaps you can have them watch this 25 second video ten times in a row. https://youtu.be/mnRS-m3pqmg Good luck!

4raylee, (edited )
@4raylee@mathstodon.xyz avatar

@johncarlosbaez @dougmerritt I never did buy Einstein’s whole 1% inspiration 99% perspiration tough talk. What reasonable person can mange to get through the work that quickly? I bet his perspiration factor was measured on a negative log scale like the rest of us mere mortals. [ -log(1-p) ]

(A perspiration factor of 2? C’mon, get real!)

4raylee,
@4raylee@mathstodon.xyz avatar

@johncarlosbaez Agreed, I was trying to be funny. Also agreed that ruling out false paths quickly does help quite a bit to cut down effort. I’m still deeply impressed, even if I think his glib quotes may undersell the effort.

@dougmerritt

4raylee, (edited )
@4raylee@mathstodon.xyz avatar

@johncarlosbaez while my physics education was usually pretty good with the historical aspects, I do have a gap around the earlier paths he took to get to GR. (Say, 1905 through ‘15, maybe.) Pais’ book sounds like something I’d enjoy. Luckily I have free time coming up, too.

@dougmerritt

noneuclideandreamer, to physics German
@noneuclideandreamer@mathstodon.xyz avatar

So I'm trying to solve the heat equation numerically. (Here with 2 sources)
But why doesn't it spread in a circle?

I basically do

f(x,y,t+1) = f(x,y,t)+c(f(x+1,y,t)+f(x-1,y,t)-2f(x,y,t))
+c(f(x,y+1,t)+f(x,y-1,t)-2f(x,y,t))

which is the numerical heat equation, right? Do I just need to finetune the c? Or maybe it's not actually a problem and my coloring takes small changes too seriously, after a while it looks nice...
(I already fixed the central horizontal distortion?)

video/mp4

4raylee, (edited )
@4raylee@mathstodon.xyz avatar

@noneuclideandreamer not exactly. Compare with the Laplacian in polar coordinates. Or written in a more coordinate-free way as the divergence of the gradient. Mathematically the same, but a way different naive conversion to code. You left out influences from the diagonal neighbors so of course you get a diamond pattern.

4raylee,
@4raylee@mathstodon.xyz avatar

@noneuclideandreamer ah, oops. Now I have to reset the sign: “[0] days since I’ve given unsolicited advice“

BTW thanks for adding a bit of art into my day!

  • All
  • Subscribed
  • Moderated
  • Favorites
  • •
  • Leos
  • ngwrru68w68
  • Youngstown
  • thenastyranch
  • InstantRegret
  • slotface
  • PowerRangers
  • Durango
  • khanakhh
  • kavyap
  • tsrsr
  • everett
  • DreamBathrooms
  • mdbf
  • cisconetworking
  • magazineikmin
  • rosin
  • osvaldo12
  • hgfsjryuu7
  • cubers
  • ethstaker
  • tester
  • vwfavf
  • GTA5RPClips
  • tacticalgear
  • modclub
  • normalnudes
  • anitta
  • All magazines
    •