You have probably heard about how the language of the bees was discovered in the beginning of the XX century bit.ly/2DPqwt3. Roughly speaking, scout bees, circling around, transmit information about where they saw food (flowers, nectar, etc.) to bees-workers. At the same time, their movements encode the direction to the food source, the distance to it, and an estimate of the quantity of food. The direction of the sun is used as the main reference point. The Austrian zoologist Karl von Frisch deconstructed this signaling system back in the 1920s, and in 1973 he received the Nobel Prize for this work.

Later, in the early 1990s, an international team of scientists constructed a mechanical cyborg bee to test a number of subtle points in this theory. Then they, controlling its flight, forced the worker bees to fly to a given goal. Here you can read a popular article about this research in Scientific American from June 1994 bit.ly/2NkmZCN.

I decided to wrote about it because recently I came across an article “Automatic detection and decoding of honey bee waggle dances” bit.ly/2IwXfSz, where some German scientists made automatic recognition for the language of the bees with a neural network and delivered it all down into production on a live hive. Software and hardware specifications have also been published.

Today just a mechanical 3-bit binary counter, via twitter: bit.ly/2PbS4dA

There is a magazine LOW←TECH MAGAZINE, dedicated mainly to the practice of modern high technologies denial. Guided by the well-known "eat your dog food" principle, authors set their website up on a solar-powered home server somewhere in Barcelona. The site had to be made static, without cookies and js, with terribly squeezed pictures, but there is a very nice indication of current power amount, as well as a weather forecast and site availability statistics.

Here is a description of the site and server: bit.ly/2Pi3gVX, here is an blog article about building low-tech websites bit.ly/2IEnoiE and
weather/uptime statistics bit.ly / 2zXdSo3.

I hope this post is not for DDoS-it.

Look what I've found on Twitter, an author called it snetris (snake + tetris) bit.ly/2OSlC2G

In my childhood all the telephone exchangeы in my city were stepping switched, and the end devices were mostly electro-mechanical. Thus, we had to read rare copies of the 2600 and phrack magazines. Those days I had some strange portable DTMF-beeper like this bit.ly/2QKFjqE. From that time, by the way, I still have the skill to dial the number with the reset lever on the impulse dial phone and read the dialed number by ear -- both on the impulse and on the tone dialing. Then we came up with picking up the melodies on the telephone's tone dialing, something like this: bit.ly/2OiCJeN. Nowadays, it turns out, there are entire sites with tablatures for different melodies. For example, bit.ly/2NBBcLB or bit.ly/2OlNtcp.

And on the similar topic: recently I discovered there are church midi-carrilons bit.ly/2IQAedH. You can upload melodies, set alarms and manage them remotely from Android / iOS app.

I've already posted here some Japanese snake-robot bit.ly/2OUEgHu. Here's another one for you, from Matsuno Lab bit.ly/2CGFYGE.

Recently Strange Loop 2018 passed, it's a pretty good hardcore conference bit.ly/2yjUYX2, I usually watch some talks from there on Youtube. This time I liked the talk by some guy from MIT, who broke (literally and figuratively) a vintage Western Electric telephone switchboard from 1927 and made it run a game bit.ly/2IVxkUN. Some of other talks are also pretty interesting.

Recently I visited another fireworks festival, the one where they shoot rockets while playing classical music. There I realized I have a few questions: first, how do they design fireworks shows? Testing such things in production should by bloody expensive, thus, there has to be some special environment for modeling, taking, probably, into the account of a meteorological situation, an arrangement of a scene, sound characteristics, with calculation of a zone of falling of garbage, and effects of difference between speeds of light and sound. Also, the simulation of the projectile flight and the subsequent explosion of single fireworks rocket itself is not a trivial problem. Secondly, something tells me that the colors of fireworks can not just be changed to arbitrary -- the color is determined by the working substance, the substance in turn affects the type of explosive reaction and the mass of the projectile. In short, I decided to search and read what smart people write.

The second question is more or less simple one. Indeed, there is a set of working substances that determine the chemistry of the reaction - the color and type of explosion. Here bit.ly/2NvapEJ someone assembled a good summary -- what and how it works in terms of design of explosions, but without much technical details.

As for the first question, I discovered some amazing stuff -- there are special expensive design packages, for example, Final Fireworks bit.ly/1CpkicF for $700, so you can design a scene with dynamics and music in a 3D-editor and then get the location, angles and timings for launchers. Talking about the free solutions there is a simplified version of ShellCalc bit.ly/2PAFW67 -- a plug-in for Excel, designed primarily for calculating garbage drop zones.

Also check this article from the Proceedings of the Royal society A: bit.ly/2IWGWPr about flight and explosion modeling of a fireworks rocket.

There is an annual event NaNoGenMo bit.ly/2EEHefo and even in my opinion it is extremely strange. This is a competition for writing code that generates literary texts (at least 50K words are required). The artistry of most of the resulting texts is under the question for me, but there are some interesting instances. I'll tell you about two of them.

American artist Austin Kleon bit.ly/2yLU8BL came up with the Newspaper Blackout genre bit.ly/2PH6O4m: one takes a sheet of newspaper and blacks out all but a few words that themselves form a new sentence or a short poem. Once upon a time I did something similar with the Moscow metro scheme bit.ly/2S0zKWG. So, the former CTO Safari Liza Daly bit.ly/2CSuOyH automated this process for NaNoGenMo 2016 bit.ly/2NNd0WM. You can use an arbitrary image with the text as an input, and possible sentences are sampled from some kind of probabilistic grammar.

In 2015 Kevan Davis bit.ly/2AgzJqH came up with the another great idea bit.ly/2NQAGcC: let's take each sentence of the input text and replace it with the closest Levenshtein sentence from the key text. I should tell you, Shakespeare's text transcribed by International Code of Maritime Signals, 1969 edition is amazing bit.ly/2OwVlHP

Came across a nice automata/fractal blog bit.ly/2S4uMrI
Also, a harsh article about a sphere coverage with the flowsnake fractal using the Gosper-islands division bit.ly/2R9VE8D

Interactive map of buildings in Amsterdam according to the constuction date bit.ly/2yUrFdf + animation based on this data bit.ly/2CVbHEh.

Yearly frequency of colons in titles of academic articles within literary studies. Source: bit.ly/2O6mexl

As a responce to my yesterday's post Borislav sent me a good article with a colon in the title: "PianoText: Redesigning the Piano Keyboard for Text Entry" bit.ly/2qef4hd

Crazy geeky pumpkin carving from Alex Wer, check out his site bit.ly/2yyEXwA for more amazing works.

There is an interesting effect of self-organization in well-distinguishable macro-structures in large bird swarms bit.ly/QLGwia, which is often referred to as flocking or herding.
It also occurs sometimes in schools of fish bit.ly/1vx4n6R or, for example, in flocks of sheep bit.ly/2i7DVkQ. Such complex behavior always gave rise to a lot of speculation, especially until Ilya Prigogine developed a theory about self-organization in dissipative structures. In 1986, computer graphics specialist Craig Reynolds bit.ly/2AzpAp7 (who participated in making of "Tron" movie) came up with a very simple algorithm called boids (= bird + droids) bit.ly/2Og1d3o. The algorithm has only three basic rules of behavior for each agent and does not require communication between agents: avoid collisions (move away from the local neighbors if they are closer than a certain threshold), move in the same direction as the local neighbors (on average), move towards the center of mass of the local neighbors. These three vectors are averaged and give a working vector for each agent, then we iterate.

The result was so good that it became actively used in computer graphics for the cinema. For example, in the movie "Batman returns" in 1992, the behavior of flocks of bats was calculated using this model bit.ly/2zbl7qS. Similar models are used in different institutions and for more serious purposes, not only for simulating birds bit.ly/2CNHNAK, but also, for example, for modeling the behavior of people in the subway bit.ly/2F2BagX, at stadiums and crowded spaces bit.ly/2yEj3rZ and during fire evacuation bit.ly/2z7Mdik.

Here bit.ly/18Mk2QF you can read the ancient post of Craig Reynolds himself with explanations of how he invented this model. And I myself learned about it from Philip Ball's book The Critical Mass, bit.ly/2ERMcW2, devoted to the usage of physical models for describing various, including social phenomena.

Recently learned about a classification of car parking structures with names for different topologies. Source: 125 pages handbook of recommended practices for the parking construction bit.ly/2Q9OAZm

A.T. Fomenko. Gradient descent. 1976.

Anatoly Timofeyevich Fomenko, famous for his, ahem, extravagant "New chronology" books series, is a good mathematician, for many years works at the Mechanics and Mathematics Faculty of Moscow State University as head of the Department of differential geometry and applications. In his spare time, he has been drawing quite strange pictures since the seventies. I remember the first time I saw his illustrations in his textbook on topology. The picture above is called "Gradient descent", and more works can be viewed on its website: bit.ly/2Qopq9m.