The Official Status Thread
-
@topspin As you say, the first one looks good, maybe? I don't really know if there should be any backwards pressure like that. The second one just reminds me of a cellular automaton (with one of those stupid space-filling rules; I will only tolerate replicator).
After doing that I am interested in maybe doing a wave simulation thing myself sometime. I think to actually simulate it, you just change the amplitude of each point based on the relative amplitudes of those around it. So 1 surrounded by 1s won't change at all, surrounded by 0.9 will come down faster than surrounded by 0.95. Maybe extra push towards rest? And to not have reflection, I would guess you just ignore the edges rather than imagining them as being at rest.
It was just a quick bodge to get an image out and I don't have much in the way of graphics libraries, so I just used an HTML canvas. You see it if you're interested. !!Warning!!: won't display in Internet Explorer
Maybe some sort of wave tank is a suitable task for a Coder Challenge?
-
-
@Tsaukpaetra said in The Official Status Thread:
you feel like an absolute retard?
I'm not sure why I thought this would work. It compiles fine though. Just doesnt' work.
Edit: oh. It does work. Just, I didn't tell it to do what I said I did. Go me...
-
@TimeBandit said in The Official Status Thread:
@Luhmann said in The Official Status Thread:
As long as Access is around ...
I prefer it to the alternatives. Our accountants love using Excel as a database, it's like a terrible disease. Also their macro-laden disasters cause huge performance problems, as the ODBC thing they're generally querying is crap at the best of times.
-
@Tsaukpaetra hey, I just did a trading system myself! Fun.
-
@pie_flavor said in The Official Status Thread:
Instruction: Print a line of your favorite character before the menu.
Obvious response:print_line('\07', 10);
If
\07
is a zero-width character, wouldn't a line of them be infinitely many?
-
Status: Darn...
-
Status: It's about 30 minutes into the day and I swear I'm going to go postal.
My phone isn't logging into the supervisor's wifi!
OK, we can get get that fixed right up.
I would do it myself but the password posted in the canteen has disapeared.
...
What password would that be?Unauthorised use of IT is a major disciplinary offence, whether I can get our management to actually do anything is
Global password reset for all APs has just been pressed.
Fucks sake.
Edit: I'm not going to be asking to roll out WPA2-Enterprise any longer. Pre-shared keys is a shit solution for an organisation.
-
@Cursorkeys said in The Official Status Thread:
What password would that be?
I don't remember
Who posted it there?
I don't rememberAre you 12? 'I don't recall' is such a facile response.
-
@Cursorkeys said in The Official Status Thread:
@Cursorkeys said in The Official Status Thread:
What password would that be?
I don't remember
Who posted it there?
I don't rememberAre you 12? 'I don't recall' is such a facile response.
Maybe she's future president
-
@Tsaukpaetra said in The Official Status Thread:
@Tsaukpaetra said in The Official Status Thread:
you feel like an absolute retard?
I'm not sure why I thought this would work. It compiles fine though. Just doesnt' work.
Edit: oh. It does work. Just, I didn't tell it to do what I said I did. Go me...
Only on C/C++ I think? But I'd expect anything that doesn't support it to give some form of error message.
-
@PleegWat I mean, if it didn't work, you could make it work with some
&
s and a*
.
-
status: happy birthday to me
-
@pie_flavor Happy 20th tropical cycle of our common mass of iron and silicate!
-
@pie_flavor happy anniversary of your mother screaming in pain trying to get rid of 20 inch parasite!
-
@kazitor said in The Official Status Thread:
@topspin As you say, the first one looks good, maybe? I don't really know if there should be any backwards pressure like that. The second one just reminds me of a cellular automaton (with one of those stupid space-filling rules; I will only tolerate replicator).
After doing that I am interested in maybe doing a wave simulation thing myself sometime. I think to actually simulate it, you just change the amplitude of each point based on the relative amplitudes of those around it. So 1 surrounded by 1s won't change at all, surrounded by 0.9 will come down faster than surrounded by 0.95. Maybe extra push towards rest? And to not have reflection, I would guess you just ignore the edges rather than imagining them as being at rest.
What I did was go to Wikipedia and check the general equation:
Then I thought "sure, looks simple enough" and tried to implement a simple time integration on a rectangular grid. The two things I've overlooked:
- The simple finite differences implementation of ∇2 for a NxM grid only outputs an (N-2)x(M-2) grid, so I need some boundary conditions. I've played with repating the boundary or just setting ∇2 to zero there, but I really don't know what's right. Both reflect (which is also cool, I guess).
- Just like you described above with checking the neighbors, this means that information only propagates from a pixel to its direct neighbors. Now, if you choose the speed c in such a way that the wave would travel farther than one pixel in one time step, it obviously falls apart, because the information can't propagate that far. So I guess that's already a limit to numerical stability, but it's probably more complicated than that and I'd have to ask someone who actually knows this stuff.
It was just a quick bodge to get an image out and I don't have much in the way of graphics libraries, so I just used an HTML canvas. You see it if you're interested. !!Warning!!: won't display in Internet Explorer
Or Firefox, or Chrome.
Maybe some sort of wave tank is a suitable task for a Coder Challenge?
And get yelled at by Blakey because there's math in the challenge?! Oh wait...
-
@topspin said in The Official Status Thread:
What I did was go to Wikipedia and check the general equation:
Then I thought "sure, looks simple enough" and tried to implement a simple time integration on a rectangular grid. The two things I've overlooked:
- The simple finite differences implementation of ∇2 for a NxM grid only outputs an (N-2)x(M-2) grid, so I need some boundary conditions. I've played with repating the boundary or just setting ∇2 to zero there, but I really don't know what's right. Both reflect (which is also cool, I guess).
Options for boundary conditions:
- Periodic boundary conditions (ie the Nth pixel and the 0th pixel are forced into equality). This gets...interesting in 2 or more dimensions.
- Fixed-end boundary conditions (boundary pixels set to an arbitrary value). This will cause phase-inverted reflections, possibly leading to standing waves.
- Free-end boundary conditions (usually a zero-gradient condition, but you can set the gradient to an arbitrary value). These will cause non-inverted reflections.
- Damping boundary conditions--here you "cheat" and set the actual grid into a much larger grid and apply an exponential damping condition to the "outside" pixels. If done right, this simulates a finite window into an infinite space. Information propagates outward and doesn't come back.
- I'm sure there are many others.
Another fun option is to have different values for
c
--a medium change. You can get all sorts of fun reflection effects.
-
@Gąska said in The Official Status Thread:
screaming in pain trying to get rid of 20 inch parasite
Could have been a C-section :>
-
@dkf said in The Official Status Thread:
@Cursorkeys said in The Official Status Thread:
We have a
randon
in one of our code bases that I can't refactor to be spelt correctly.For ages, we had
baord
baked into one of our repositories…We've got "downloaf" and "uninclused" but I think those are both intentional.
-
@Tsaukpaetra said in The Official Status Thread:
Status: Watched a short clip on progress with a robot:
https://www.youtube.com/watch?v=ESC9gu0NAak
YouTube decided to show this as suggestions:
WAT.
A very similar thing happened to https://www.youtube.com/watch?v=8BxdMLS5wO8 when it first went up.
-
-
@pie_flavor said in The Official Status Thread:
happy birthday to me
-
@topspin said in The Official Status Thread:
Now, if you choose the speed c in such a way that the wave would travel farther than one pixel in one time step, it obviously falls apart, because the information can't propagate that far. So I guess that's already a limit to numerical stability, but it's probably more complicated than that and I'd have to ask someone who actually knows this stuff.
The usual fix is to shorten the timestep until the problem goes away, and then sample the model to give the impression that the original timestep was in use. It's a hack, but it works pretty well here as you're not using any event-based parts. (When you've got a hybrid system, it gets trickier.)
-
Status: Damn my C# is rusty.
How do I make a method/function I can call without creating an object again?
Oh...
static
.
-
-
@Captain public static ReturnType FunctionName()
-
@Zerosquare It's more Haskell-y than Rust-y. I know what's possible, but how to do it? :-)
-
@Captain said in The Official Status Thread:
@Zerosquare It's more Haskell-y than Rust-y.
True, but languages are
-
@Benjamin-Hall As I've said, I've tried periodic and fixed-end boundary conditions. Both seemed to reflect / cause standing waves, but there's always the chance I messed something up.
Adding a medium change sounds like a fun idea, should cause refraction, right?
-
@dkf said in The Official Status Thread:
@topspin said in The Official Status Thread:
Now, if you choose the speed c in such a way that the wave would travel farther than one pixel in one time step, it obviously falls apart, because the information can't propagate that far. So I guess that's already a limit to numerical stability, but it's probably more complicated than that and I'd have to ask someone who actually knows this stuff.
The usual fix is to shorten the timestep until the problem goes away
That is of course what I did. "Huh? This looks wrong. Let's try a smaller dt".
Normally I'd try an implicit time integration scheme, too, but I don't know if that helps here (propagation speed is definitely a limiting either way) and that wouldn't be simple to throw together anymore.
-
@Captain said in The Official Status Thread:
@Zerosquare It's more Haskell-y than Rust-y. I know what's possible, but how to do it? :-)
As in "you can compute everything, but there's no side effect to observe the results"
-
Does C# really have no syntactic sugar for lists?
-
@Captain Of course it does.
var list = new List<string>{"A", "B", "C"};
-
@topspin No, as in using type classes and generic transformations to structure an algorithm/program.
-
@pie_flavor Ugghliest sugar ever, but it should work.
-
@Captain said in The Official Status Thread:
@topspin No, as in using type classes and generic transformations to structure an algorithm/program.
I see you did not appreciate my joke.
-
@Captain That's because it's not really list sugar. Anything implementing
IEnumerable
and that has anAdd
method taking one parameter can be initialized in that way.
-
-
@topspin said in The Official Status Thread:
@Benjamin-Hall As I've said, I've tried periodic and fixed-end boundary conditions. Both seemed to reflect / cause standing waves, but there's always the chance I messed something up.
Adding a medium change sounds like a fun idea, should cause refraction, right?Doing periodic boundary conditions right is hard. You have to equalize both the value and the gradient of the function. Otherwise it acts like an impedance-mismatch boundary (like a medium change or an end that moves with damping) which will cause both refraction (changes in the transmitted waveform) and reflection.
-
@topspin I still have nightmares about the course where they made us use C++ to do this stuff:
-
@topspin said in The Official Status Thread:
Adding a medium change sounds like a fun idea, should cause refraction, right?
@Benjamin-Hall Pretty cool! I see both refraction as well as reflection (stronger than I expected). I know a real wave does both, but does it follow from the wave equation? Apparently it does.
-
@topspin said in The Official Status Thread:
@topspin said in The Official Status Thread:
Adding a medium change sounds like a fun idea, should cause refraction, right?
@Benjamin-Hall Pretty cool! I see both refraction as well as reflection (stronger than I expected). I know a real wave does both, but does it follow from the wave equation? Apparently it does.
Yeah, it's from the wave equation. Easiest to see in 1D with say two strings of different unit mass densities. Since waves propagate at different speeds in the two strings, you get refraction. But since momentum depends on wavelength, and the wavelength of the transmitted wave is different (because the speed is different), you also need a reflected wave to conserve momentum. I ended up having to solve this problem for the quantum wave equation (where some components are complex) for my thesis. I was moving a wave packet between an initial potential surface and another. You get nice back-scatter. The transmitted packet is related to the transition probability between the new and old electronic state, while the reflected part is the unmodified state.
For "thick" glass in air, there's about a 4% minimum reflection for light.
In physics, just about everything is some combination of conservation laws. Energy + momentum conservation explains the vast majority of phenomena out there, even if the connection is obscure.
-
@topspin said in The Official Status Thread:
@kazitor said in The Official Status Thread:
It was just a quick bodge to get an image out and I don't have much in the way of graphics libraries, so I just used an HTML canvas. You see it if you're interested. !!Warning!!: won't display in Internet Explorer
Or Firefox, or Chrome.
I had to refresh after uploading, but you could always just download it first. I'm not going to say that not arbitraily executing uploaded javascript on this domain is a …
-
@kazitor Refreshing works.
-
Status: Why does Dapper.NET keep telling me it can't find a table in my database? I even set the initial catalog on my stupid connection string. :'(
Oh, I get it now, and my frist test passed!
-
@Applied-Mediocrity said in The Official Status Thread:
@topspin
Must be from the code of conduct WTDWTF doesn't have.18 WTF § 4(b)(15)(iii) Only @Tsaukpaetra is allowed to post ponies. No exceptions.
-
Status: Awake super early so I can be ready to catch my shuttle back to the airport to go back home. Hate these work trips.
-
Status: Looking at the React 'getting started' tutorial example:
Seems about right for webdev....
-
@Cursorkeys could be worse - when I was starting out, I couldn't read the documentation because of fucked up CORS (that was last month - they seem to have fixed that now).
-
@e4tmyl33t said in The Official Status Thread:
Status: Awake super early so I can be ready to catch my shuttle back to the airport to go back home. Hate these work trips.
Status: awake super late because fuck my mind.