Extended bio: I’m a multidisciplinary software engineer, computational designer, artist, researcher, teacher, prolific open source tool maker w/ 25+ yrs experience. Use this thread to learn more about me & selected projects (in no particular order, adding more over time):

(Disclaimer: The results of my professional recent work (since 2016) are unfortunately under NDA, therefore this thread will only include older projects & more general info...)

1/ Since '93 I've been working hands-on (concepts ⇄ IXD/UX ⇄ datamodels ⇄ front/back/devops ⇄ fabrication ⇄ hardware): web, generative design/audio/DSP, dataviz/database/semweb, museum/art/tradeshow/light installations, architecture, branding, games, print, dozens of collabs

2/ My practice often starts deep & uses a bottom-up design philosophy (small, isolated ideas/solutions eventually forming larger units), multi-layer feedback, focus on long-term projects/goals/edu, design freedom at any stage, ensuring cohesion, yet avoiding tight coupling...

3/ Since 2006, underpinning my most successful projects (incl. those below) is self-developed polyglot open source tooling for computational design (in the widest sense), specifically http://thi.ng  & precursor http://toxiclibs.org , with (still) 1000s of users...

4/ http://thi.ng  itself a long-term meta effort consisting of 250+ sub-projects providing aforementioned bottom-up tooling for _all_ projects I’ve worked in past 10+ years. Originally aimed at #Clojure/ #ClojureScript, the focus for past 5 yrs on #TypeScript & some #C11

5/ This long-form essay from 2015 provides a deeper context, partially discussing my early years (since 1988) & general non-Silicon-Valley-tinted work philosophy/learnings from having been in this field for a while: https://medium.com/@thi.ng/the-jacob-s-ladder-of-coding-4b12477a26c1

Let’s go see some projects!

6/ DevArt Co(de)factory was a joint Google Creative Lab / Barbican commission for the Digital Revolution exhibition (2013/14). User designed 3D printed sculptures created via visual programming/modelling. Entire project open source: https://github.com/postspectacular/devart-codefactory #clojure #rpi #3dprint

7/ DevArt (above project) video feature by Google (Fun fact: I actually did work on this bench for almost 8 years - a huge productivity booster - though not just daydreaming about trees as that video seems to portray ):

8/ In 2012/13 I collaborated w/ Precedent London on an audio-responsive identity system for Leeds College of Music: Strange attractors, billions(!) of particles, desktop app, ad-hoc AWS render farm for super-hires print/video asset generation/transcoding https://vimeo.com/69896668 

9/ Some screenshots of the Leeds College of Music identity generator app, as well as a video of the associated cloud based render & asset manager (all entirely developed by myself):


#Clojure #OpenCL #OpenGL #GLSL #JavaScript

10/ In 2012 I collaborated w/ @tellart & others on another large Google project: Chrome Web Lab @ Science Museum London. My roles incl. design/dev of custom fiducial marker system (Lab Tags), their generator (350k+) for print, decoder & dispenser UI...

11/ ...the 10m wide & tall 6-screen Lab Tag Explorer wall visualized visitors as physics-based particle flocks, displayed various stats and was augmented by (+ synced with) pen drawings performed by a 2m tall industrial robot (which I consulted on too) https://vimeo.com/210692040 

12/ The wall software was split into a #Clojure/ #OpenCL based websocket server for data aggregation, robot interaction, physics/flocking sim, sending 30fps stream of particles updates to 2x JS/WebGL UIs running on Chrome (3x screens each)

13/ Gathered prev. experience w/ custom computer vision markers through work on Fid.gen, a physics-based marker generator nominated for Design of the Year 2009. This one was written w/ Processing/toxiclibs & used for Lexus @ NAIAS Detroit 2008 and...

14/ ...final year student exhibition @ London College of Fashion 2007/8 for which I too designed & dev'd the gestural table UI (whilst consulting at @movingbrands) - in both cases the custom markers were used w/ @mkalten's amazing reacTIVision software: https://vimeo.com/1393556 

15/ A 2013 collaboration w/ @ODIHQ to produce this and other London-centric ambient data visualizations for their very first summit, e.g. here showing knife crime data from 2012/13. Generated entirely from UK ONS data in #Clojure, rendered in #Luxrender on 200 AWS EC2 instances

16/ Speaking of dataviz, the http://thi.ng/geom-clj  2d/3d geometry toolkit (2011-16) was instrumental for most aforementioned projects & also includes a module for declarative dataviz. More details on Github & this workshop report about using #LinkedData: https://medium.com/@thi.ng/workshop-report-building-linked-data-heatmaps-with-clojurescript-thi-ng-102e0581225c

17/ Earlier, in 2009 & 2010 I organized & produced community showreels to document the amazing work done by Processing & http://toxiclibs.org  users at that time when the computational design crowd was still very small/young(er). Take a look!

17b/ For completeness sake, here's the 2010 toxiclibs showreel:

More to come...

18/ The volumetric/voxel tools used for some entries in above showreel were originally written in 2008, whilst working on my first 3D print, a #ReactionDiffusion #generative typographic sculpture for a special issue/interview in Print Magazine. More pics: https://flickr.com/photos/toxi/albums/72157604724789091

19/ http://thi.ng/simplecl  was one of the first http://thi.ng  #Clojure projects & instrumental for the above mentioned Leeds & Web Lab projects, but also for building this experimental volumetric data PBR renderer (w/ Clojure REPL): http://thi.ng/raymarchcl  (2012-13)

19a/ As an aside, in prep for another project, I wrote a tiny ambient occlusion renderer (w/ anti-aliasing) for isometric landscapes in <150 lines of #Java & #toxiclibs (2011):

19b/ Last tangent on this theme: These structures were part of a self-directed research project (also 2012) into Sparse Voxel Octree implementations (in Clojure) w/ aim of building a slicer for DLP-based 3D printing (resin), using Gyroid minimal surfaces for internal supports

20/ To further illustrate the bottom-up design philosophy underlying both @thing_umbrella and @toxiclibs, the latter's #VerletPhysics engine was originally written for this cute @universalevery collab for @nokia (2008), but still widely used today...

21/ Another @universalevery & @freefarm collab (2008): Forever at the @V_and_A was a fully generative audio-visual install w/ bi-directional feedback gfx⇄music and a rule-based stochastic composition system to create a fully dynamic interplay between both

22/ To promote the V&A Forever exhibit (), we decided to generate & print 5000 unique postcards, each based on a different snapshot of the physics sim driving this single line/rope and here rendered to create a cloth like appearance. Later also shown at SIAD Sheffield

23/ For the poster of that Designing Seeds show (2009) we ( @universalevery) re-purposed my earlier @Channel4 New Shoots project, using #DiffusionLimitedAggregation to grow typography from 5+ million particles. We also showed Forever, running on 6 Macbooks (operated by my son)

24/ Slight time jump: After working extensively w/ #RaspberryPI's on DevArt in 2014, I went deep into #baremetal ARM coding on STM32 dev boards, building a programmable synth (plus #Forth virtual machine) & teaching related workshops - thanks to all who were part of this journey!

24a/ Here are 2 videos live recordings of the synth in action:
-
-

Soundcloud playlist w/ 40+ example tracks: https://soundcloud.com/forthcharlie/sets/stm32f4

Project on Github: http://thi.ng/synstack 

25/ Flatworld was a never officially released library for unfolding 3D meshes for 2D sheet materials, initially created for a data viz/sculpture workshops at @HEADGENEVE (& others) in 2010, but then mainly used for this @V_and_A installation (May 2011):

25a/ This install () was related to the William Morris exhibit & made w/ Java/Clojure/Processing (runtime/server), toxiclibs (geometry, fabrication files for robo & laser cutters), JS (iPad UI). Massive thanks to @manomine, @sermad & @supernat13 for helping w/ setup!

26/ In early 2004, whilst helping on the Bagel graphics engine of early @Processing versions, I created Base26, an interactive realtime 3D visualization of all English four-letter words, interpreting letters as spatial coords (and initial letters as sets to morph between)

26a/ These are high-res versions created for various exhibitions & print media. Each image shows a set of 4-letter words starting w/ same initial. The last 3 letters of each word interpreted as XYZ positions, different colors are used for word type (noun, verb, adjective etc.)

26b/ The wireframe mesh surrounding each word/point is a 3D histogram of the overall usage frequency of each particular trigram across the entire set of words. More bulging forms = higher frequency. E.g. the trigram "ear" (fear, wear, tear) vs. "oft" (soft, loft...)

27/ Over the years I've always enjoyed working on/with datastructures/databases as much as all the more visual work you might know. Since 2004/5 I've repeatedly worked on/off with #RDF / #LinkedData / #GraphDatabases, built several related tools & semi-large related projects...

28/ http://thi.ng/fabric  was a #Clojure project centred around a general compute graph, on which sat a triple-based fact graph w/ #SPARQL-like query DSL, subscribable stored queries & #LinkedData server/endpoint, able to emit results in multiple formats (JSON-LD, CSV, NT etc)

29/ Years before that, in 2005, one of my first #semweb projects was FOAFgen, a vCard batch converter to generate @libbymiller & @danbri's Friend-of-a-Friend (FOAF) social graph format - a time before social media as we know it today: http://web.archive.org/web/20050329015144/http://toxi.co.uk/foafgen/

30/ Having implemented numerous folksonomy/tagging systems over the years (in many different databases), I'm still proud of the name-dropping in official MySQL slide decks about my earliest attempts/solutions ("getting it almost right" ) in this field:

https://www.slideshare.net/edbond/tagging-and-folksonomy-schema-design-for-scalability-and-performance

31/ As mentioned earlier on in this thread, #toxiclibs was a collection of ~300 "building blocks for computational design" (for Java/Processing) & formed basis for most of my work between 2006-2011... (These visualizations were created as exercise during a workshop at @V_and_A)

31a/ Toxiclibs included ready-to-use tools for many topics: physics engine w/ Verlet integration & behaviors, volumetric modelling, meshes, color theory/spaces, geometry/linalg, intersectors, interpolators, oscillators, reaction-diffusion, cellular automata, fluid sim, data struc

32/ Voxel modelling in action: Voxel Posse, my last project for @universalevery, a collection of ~100 3D printed, semi-random wireframe characters. Basic physics-based skeleton/joints setup & gravity sim to prune failures, final selection by my 10yr old son. Concept by @mattpyke

32a/ In part, this project was based on prior research (early 2010) about combining recursive mesh subdivisions with volumetric modelling. Also realised w/ toxiclibs, printed at @shapeways. Here a 6x subdivided cube with edges retraced in voxel space to visualize the new topology

32b/ Some further (early) examples of this subdivision research, here more with a view at biomimicry and morphogenesis, later leading to the http://thi.ng/morphogen  DSL, but also workshops w/ Architecture departments at ETH Zurich & Städelschule Frankfurt