@sejo @compudanzas #madrid #solarpunk #circulobellasartes
Faltó @miel por allí
@sejo @compudanzas #madrid #solarpunk #circulobellasartes
Faltó @miel por allí
gta vi
gta emacs
gta wordpad
some photos from #FrugalCoding #LiveCoding performances at #SolarPunk #Festival #Madrid tonight. festival continues at the weekend.
setup and sound+video check done: @mathr and @compudanzas are ready to play.
solarpunk festival opens tonight 18:00 with conference (first talk 18:30), frugal coding live show 21:00 (first set 21:30, our set 22:30).
free entry until capacity filled.
https://www.circulobellasartes.com/en/humanidades/frugal-coding/
https://www.circulobellasartes.com/en/humanidades/festival-solarpunk-abundancia-o-barbarie-2026/
I'm really excited to announce that our second ever From Scratch event will happen on the 20th December! RSVP here!
https://luma.com/5j4m4xdm
As a reminder, this is an experimental event where all the night's electronic music and visuals are made LIVE ... entirely FROM SCRATCH 🫥🎀
Everything you see and hear will be entirely improvised from a blank screen. Experience music being created from nothing! Watch visuals being crafted out of thin air!! ☁️💓☁️
Everyone is welcome to perform!
(There's a sign up sheet on the event page)
Made a page about the PolyBLEP algorithm for bandlimiting waveforms with step discontinuities like sawtooth:
https://mathr.co.uk/web/polyblep.html
approach: integrate a cubic spline approximating raised cosine windowed sinc function to get a bandlimited step function which replaces each step discontinuity
pulse waveform can be achieved by difference of two out of phase saws, triangle can be achieved by integrating pulse; doing it for hard sync is possible too but I haven't worked out the details yet.
FRUGAL CODING
Círculo de Bellas Artes (Madrid)
20 Nov, 21h.
Códigos sonoros desde la intimidad de la máquina, instrumentos híbridos de interfaces recicladas, arqueologías alienígenas y danzas pixeladas… Una fiesta para hibridarnos con microorganismos digitales, biológicos, lumínicos y electromagnéticos y habitar futuros tecnológicos simbióticos desde la pista de baile.
Con @compudanzas @mathr @loopier @alien @alt234
#livecoding #frugalcomputing #algorave #Madrid #music #visuals #instruments #solarpunk
https://www.circulobellasartes.com/humanidades/frugal-coding/
#noisevember 13th more WIP "inch" drone, this time with a pinker spectrum (fewer obnoxious high frequencies). I did a 6dB/octave high pass filter at 20Hz in Audacity in post to avoid mush, will see how to integrate that into my code.
https://media.mathr.co.uk/mathr/2026-toot-media/mathr%20-%202026-11-13%20-%20inch%20demo%20-%20pinker.mp3 24MB, 10mins
porting to JACK seems to have eliminated the xruns (apart from at program exit - which Audacity now reports correctly), so it seems SDL3 (or my use of it?) is to blame and not Pipewire...
#noisevember 12th stereo spectral drone with WIP inch version
https://media.mathr.co.uk/mathr/2026-toot-media/mathr%20-%202026-11-12%20-%20inch%20demo%20-%20stereo%20drone.mp3 24MB, 10mins
lots of xruns I edited out in this one too. I'll try porting to JACK to see if the (unreported) xruns are from SDL3 or Pipewire....
#noisevember 11th another session with my spectral instrument (which is code-named "inch"), this time a dual mono* drone with occasional plips
https://media.mathr.co.uk/mathr/2026-toot-media/mathr%20-%202026-11-11%20-%20inch%20demo%20-%20dual-mono%20drone.mp3 24MB, 10mins
I edited out 3 or 4 xruns (short silent gaps in audacity's recording) by hand, hopefully I didn't miss any...
*I've since coded up a stereo version, but not had time to test it yet....
SDL3 audio seems fine, better than SDL2 where I could never figure out duplex operation. I created 3 stereo audio streams (1 input, 2 output) and connected them to 3 different devices (physical sound cards), and it seemed to work mostly ok using Pipewire.
Occasional clicks though, which might be xruns from the unsynchronized clocks, not sure how to diagnose (does Pipewire even report xruns?). Maybe tuning SDL3 environment variables could help, not sure which backend is default.
#noisevember 1st-9th, no noise. 10th: playing with a spectral instrument I started working on in 2023, ported to SDL3 the last couple of days.
woe, spotted a slop commit in recent MicroHs changes (making Text wrap ByteString instead of String, or so) :(
@mathr plays clive alongside @compudanzas with Jarotsim and the dancing Turing machines, at Frugal Coding, Solarpunk Festival, Madrid, Spain.
Our set is at 22:30 on Friday 20 November 2026, Columns, Círculo de Bellas Artes, Alcalá, 42, 28014 Madrid, Spain.
Festival opens 18:00, show starts 21:30, free entry until capacity is reached.
Festival continues 20th, 21st and 22nd November 2026.
- https://www.circulobellasartes.com/en/humanidades/frugal-coding/
- https://www.circulobellasartes.com/en/humanidades/festival-solarpunk-abundancia-o-barbarie-2026/
- https://mathr.co.uk/web/clive.html
- https://compudanzas.net/jarotsim.html
#LiveCoding #Madrid #FrugalCoding #C #Jarotsim #TuringMachine
Hey everyone
Please find attached a new EP for your delectation <3
https://syntel8.bandcamp.com/album/alien-ep
bisous x
I have 3 points in 3D space (a, b, c), and a 4th point p. Any quick trick for finding the distance d between point p and the line equidistant from points a, b, c (which is perpendicular to the plane containing a, b, c, through their circumcenter)?
So far I have tried calculating the equation of the line m+tn where m is the circumcenter of a,b,c and n is normalized (a-c)×(b-c), then computing
q = m + ((p-m).n)n
d = |p-q|
but I don't need q, just d...
a couple of interesting things I've found in the #Miranda programming language today:
diagonalized list comprehension syntax [//]:
> take 10 [ (a, b) // a <- [1..] ; b <- [a..] ]
[(1,1),(1,2),(2,2),(1,3),(2,3),(3,3),(1,4),(2,4),(3,4),(4,4)]
parameterized scripts: script declares %free identifiers with types, %include binds them to identifiers defined in including script (the manual uses an example of element-type-agnostic matrix calculation library, that gets number type passed in from outside).
The Haskell Foundation's latest board member's spotlight Linked In post :(
https://haskell.foundation/news/2026-11-02/haskell-foundation-update-october-2026.html
Search is on to find a new favourite non-strict pure functional language, not run by arms dealers.
supposing I have a volume defined by a signed distance field which isn't necessarily aligned to my discrete mesh:
(i-1) X ( i ) (i+1)
if sdf(i-1) = u > 0 and sdf(i) = v < 0, then sdf(i-x)=0 at X roughly when x=(u + (1-v)) / 2, justified by mean of distance to boundary at either side, or something
then want to approximate a smooth function P defined on the mesh inside the sdf region by a quadratic, such that passes through i,P(i) and (i+1),P(i+1) with gradient 0 at (i-x),(P'(i-x)), to find the reflected function value outside the sdf region at unknown (i-1),P(i-1).
finally integrate the pde with forward euler:
dP(i)/dt += dt * c*c * (2*P(i)-P(i-1)-P(i+1))
P(i) += dt * dP(i)/dt
that's what I think might work for 1D, will see what goes wrong when implementing it and extending to higher dimensions
partial differential equation for waves is c^2 laplacian(p) = d^2p/dt^2, where p is (small) deviation from ambient pressure and c is the speed of sound (~343m/s)
boundary conditions need more thought... want to have emergent frequency-dependent reflection/absorption behaviour...