Remote
bazzargh
@bazzargh@hachyderm.io
mastodon
4.6.6
Baldy bespectacled beardy, beer breathed, B- bloody biker.
0
Followers
0
Following
38
Posts
Joined November 07, 2022
searchable:
searchable
github:
Posts
Replying to
@lynneguist.bsky.social@bsky.brid.gy
@lynneguist.bsky.social @lynneguist.bsky.social northern ireland and glasgow. As someone from an irish catholic family, from both those backgrounds... aimed at me. Here's an example from news reports https://www.bbc.co.uk/news/uk-scotland-glasgow-west-25672182
Replying to
@kyonshi@dice.camp
@kyonshi@dice.camp @lynneguist.bsky.social this is about a list of british english words that the nyt bee rejects-I'm aware that other words exist 😜 https://separatedbyacommonlanguage.blogspot.com/2023/04/nyt-spelling-bee-archive-of-disallowed.html?m=1
Replying to
@bazzargh@hachyderm.io
@lynneguist.bsky.social and another! (There's GITE too but already on list). Taig is a derogatory term for an irish catholic. Unfortunately heard often. https://en.wiktionary.org/wiki/taig #nytspellingbee
@lynneguist.bsky.social another BrE word today: HAEMATITE. https://en.wiktionary.org/wiki/haematite#English "chiefly British" and perfectly common: "a common iron oxide compound...widely found in rocks and soils" https://en.wikipedia.org/wiki/Haematite #nytspellingbee
Replying to
@Richard_Littler@mastodon.social
@Richard_Littler bottom of first poster is The Story of Vernon&Irene Castle; bottom left of third is Boys' Reformatory. Not sure of the other partial
Replying to
@bazzargh@hachyderm.io
@lp0_on_fire @bit101 I expect the bot to eventually post the result... but without gxr, it's more limping than stepping. Since it needs more draws the graphics are choppy and I made everything shorter, but the shades don't quite work - the yellow foot drags, but never steps in front of the blue. Could maybe recover those with user-defined characters.
Replying to
@lp0_on_fire@social.linux.pizza
@lp0_on_fire @bit101 #bbcmicrobot MODE1:VDU5,23,12,1;256;1;256;23,15,770;515;770;515;
VDU19,1,4;0;
REPEAT
FORI=-16TO160
IFI=0VDU19,0,0;0;19,3,7;0;
IFI=100VDU19,0,5;0;19,3,5;0;
GCOL0,1:MOVE8*I+128,420:PLOT0,0,16:PLOT81,4,-16:PLOT81,0,16
GCOL0,2:MOVE8*I+128,480:PLOT0,0,16:PLOT81,4,-16:PLOT81,0,16
GCOL0,7*((I DIV4)MOD2):MOVE8*I,360:PLOT0,0,200:PLOT81,4,-200:PLOT81,0,200
NEXT
UNTIL0
Replying to
@bit101@mstdn.social
@bit101 a version I made for the bbc micro a few years back. I do like that coding up optical illusions like this means you know for certain it's not what it appears to be... but you can't unsee it https://mastodon.me.uk/@bbcmicrobot/109763774911592856
Replying to
@acf@masto.alancfrancis.com
@acf on the suggestions for social meeja changes - one I wanted on the old birdsite was that if you posted the same text as someone else, that's a retweet. Some positive effect on crediting originality, plus catches out spam bots, politicians repeating the party line, joke thiefs.
Replying to
@abstractcode@eigenmagic.net
@abstractcode @NanoRaptor I always liked that about the 2001 Nero Wolfe series; actors that were cops one episode would be gangsters or butlers the next, or sometimes in the same episode https://en.wikipedia.org/wiki/Nero_Wolfe_(2001_TV_series)#Repertory
Replying to
@skjeggtroll@mastodon.online
Replying to
@cstross@wandering.shop
@cstross a relative was a student doc in Glasgow (30y ago) when a newbie turned up claiming to be a protege of a prof who was on leave. He did patient rounds for a few days then disappeared; it turned out he was an imposter.
At one point he described a liver as having a "salmon like appearance" which struck others as weird and inaccurate but somehow _still_ closer than this surgeon managed.
Replying to
@matt_zucker@mastodon.social
@matt_zucker@mastodon.social you can also get nice effects varying the parameters across the surface; here using a grayscale image to drive it https://hachyderm.io/deck/@bazzargh/116092480577520472
Replying to
@unnick@booping.synth.download
Replying to
@trebach@functional.cafe
@trebach @ZachWeinersmith I can read the code just fine; my python was this (having replaced spaces with newlines in the input); a-k appear in the cartoon, the rest I added as guesses.
import sys
alpha = {
"a": "000000000",
"b": "000010011",
"c": "001001010",
"d": "000011001",
"e": "010100110",
"n": "110111000",
"y": "100110001",
"h": "011111111",
"u": "010011010",
"k": "101000111",
"i": "100001101",
"r": "000101111",
"t": "011100101",
" ": "101111010",
"m": "111101011",
"p": "111100000",
"l": "111010100",
"s": "001110110",
"o": "111100001",
"f": "010110101",
"w": "111010010"
}
alpha = {k: int(v, 2) for k, v in alpha.items()}
for n in map(str.rstrip, sys.stdin):
m = int(n, 2)
c = None
best_d = 999
for k, v in alpha.items():
d = (v^m).bit_count()
if d < best_d:
best_d = d
c = k
elif best_d == d:
c= f"{c}{k}"
if best_d < 3:
if len(c) == 1:
print(f"{c}") #, n, (alpha[c[0]]^m).bit_count())
else:
print(f"[{c}]") #, n, (alpha[c[0]]^m).bit_count())
elif best_d == 3:
print('?', bin(m), best_d)
else:
print('?', n, best_d, "no match")
Replying to
@trebach@functional.cafe
@trebach @ZachWeinersmith what I actually got: "the beau[to]y [to]f[h ]ma[to]hema[ts]ics only[h ]shows[y ]itself[y ]tp more pa[ts]ien[rt] follo[lw]ers" ... which was close enough to guess if I squint. Some letters like r,o,l,w didn't fall out immediately so I used the corrupt character as the guess - that'll have given me some extra errors; my code is likely buggy
Replying to
@ZachWeinersmith@mastodon.social
@ZachWeinersmith "the beauty of mathematics only shows itself to more patient followers", I think.
Replying to
@tao@mathstodon.xyz
@tao "the beauty of mathematics only shows itself to more patient followers", I think? I'm probably doing it wrong as I expected 1-bit errors but thought I had more. Wrote some python finding closest hamming distances to letters in the cartoon, then added new guesses when the unknown had distance 3, repeat
Replying to
@Illuminatus@mstdn.social
Replying to
@bagder@mastodon.social
@bagder@mastodon.social you could have a line like this on your chart of charts of curl over time, showing that number being more than the number of surviving paintings by Leonardo da Vinci (about 8), of self portraits by Rembrandt (about 40), less than the number of paintings by Van Gogh (around 900) etc
Replying to
@jwz@mastodon.social
@jwz@mastodon.social any time I see youtube shorts (the horror, the horror) about half the vids have this look, and I have to wonder if this is something kids these days have convinced themselves is good?
Replying to
@maxine@hachyderm.io
@maxine@hachyderm.io about the only way to make this worse is if the website was on medium
Replying to
@cstross__dup_483@wandering.shop
@cstross@wandering.shop "The unexamined life is not worth living", said Socrates, a very modern guy
Replying to
@csk@mathstodon.xyz
@csk also, recently I said I'd been nerdsniped and one of my younger colleagues asked what I meant; I dropped a link to https://xkcd.com/356/ and then realised it's *19 years old* (so is the "I know regular expressions" one)...so came out before he could read...how did we get so old? A warning for showing this to students ;)
Replying to
@csk@mathstodon.xyz
@csk so... https://nesbitt.io/xkcd-2347/ it's slightly different in that it's not the original - though it is matter.js. It took me hitting the button twice to realise what was going on - it recreates the cartoon for random real projects. I like it all tho; I did xkcd 208 for advent of code a couple of years back https://hachyderm.io/@bazzargh/113622743090408785 (...with entirely unrealistic physics)
Replying to
@lordof1@masto.ai
@lordof1 30 years ago was walking round Xi'an and we found ourselves on a street not of hair salons but of shops that sold hood dryers for hair salons. That felt like a glitch in the matrix, why so many, so specialised.
Replying to
@tth@mastodon.tetaneutral.net
@tth@mastodon.tetaneutral.net it would have made more sense to use webgl for the floor, with the carpet being a tiled texture, but I hit limitations: I think the far end of the floor exceeds the normal coordinate limit in p5js's webgl canvas - over 10k units away - plus the clip mask I did to show Danny doesn't work in webgl, the outermost shape is the mask, when I wanted to use a winding rule to unmask the middle. I could have done it just by drawing the image twice, but in the end it was just easier for me to do the projective transform and display the 2d image...in case you hit any of that. Also, my math is a bit wrong, I think the repeat spacing? you can see it when the orange S shape rotates, the hexagons expand slightly. I couldnt be bothered fixing it!
Replying to
@bazzargh@hachyderm.io
The animation looks much better in the p5 editor; here's the code https://gist.github.com/bazzargh/1880c560707f8de8dd6328a0b3acfbb1 ... which is horrible, you can see where I added alllll those points to do the masking, and the structure kept changing as I changed my mind about animations
Floor is lava?
I intended to do this for #tilingtuesday inspired by @csk@mathstodon.xyz 's effort which made me think of this carpet. However, it took me way too long to draw the geometry, come up with some animations, calculate the homography to project the points correctly, and mask out the carpet in the original image...
#creativecoding #generative Made with #p5js ; the first lines are...
const redrum = "#8a1d1e"
const dunny = "#46281c"
const torange = "#bb6538"
Replying to
@fasterthanlime@hachyderm.io
@fasterthanlime@hachyderm.io took this yesterday, seems appropriate
Replying to
@ludicity@mastodon.sprawl.club
@ludicity@mastodon.sprawl.club @TimWardCam@c.im I've tried allowing googling in problem-solving interviews, and tell them this up front - it's more representative of what you'd do in a job. I want them to talk me through what they are thinking while doing it tho, because what interests me is can they think it through, not rote memory. (surprisingly, not many google even when allowed to!)
But we had a candidate we'd told this clearly using online chat to ask someone else the questions, and lie about it, in 2 interviews, stopped the process there.
Replying to
@grickle@mstdn.social
@grickle@mstdn.social sooo... is it possible to order prints of your doodles or just the ones we see on gricklemart? I've a goth friend this would be perfect for.