#c0ffee is the colo[u]r

This is fun:

What real words are actually valid CSS HEX colors? Parsing an English dictionary for entries containing only the letters ABCDEF and limiting the result to words of exactly 6 or 3 letters length (#FFFFFF or #FFF) gives us some interesting results.

Although sadly:

#faeces refered a tad too pale

The neural network will name your next band

Janelle Shane:

An important part of starting a new band is choosing an appropriate name. It is crucial that the name be unique, or you could risk at best confusion, and at worst an expensive lawsuit.

The neural network is here to help.

Prof. Mark Riedl of Georgia Tech, who recently provided the world a dataset of all the stories with plot summaries on Wikipedia, (enabling this post on neural net story names) now used his Wikipedia-extraction skills to produce a list of all the bands with listed discographies – about 84,000 in all.

I gave the list to the Char-rnn neural network framework, and it was soon producing unique band names for a variety of genres. Below are examples of its output at various temperature (i.e. creativity) settings.

Come for the funny names, stay for the bizarre shark influence.

Yoda’s syntax in other languages

All Things Linguistic:

What does Yoda’s syntax look like in non-English versions of Star Wars? For those who aren’t familiar with Star Wars (all two of you), Yoda is an alien who, when speaking English, uses what seems to be an OSV syntax instead of the traditional SVO syntax.

So how do foreign translations of the script handle this? I am particularly interested in what it looks like in non-SVO languages. Are there any translations where Yoda’s incorrect syntax is emulated by using an English-like syntax? Or are other languages’ syntax so free that mistakes in the use of case or verb conjugations must instead be used to emulate Yoda’s “alien” speech?

Some interesting answers:

Estonian: Free word order language. Yoda retains the English OSV order. This is grammatical in Estonian, but does make it seem as though Yoda is constantly stressing the object phrase as the main point of his statements. This gives his speech an unusual quality.

Japanese: An SOV language. Yoda seems to use a more or less correct syntax, with a more archaic vocabulary.

Romanian: An SVO language. Yoda speaks in OSV. He also places adjectives before the noun instead of after the noun, and uses an archaic form of the future tense.

Turkish: An SOV language. Yoda speaks in OSV. Note: This order is also used in classical Ottoman poetry, so the syntax may have been chosen in order to emphasize Yoda’s wisdom or age.

Atomic cookery

Joe Pinsker writes for The Atlantic about metacookbooks:

It’s a shame that the standard way of learning how to cook is by following recipes. To be sure, they are a wonderfully effective way to approximate a dish as it appeared in a test kitchen, at a star chef’s restaurant, or on TV. And they can be an excellent inspiration for even the least ambitious home cooks to liven up a weeknight dinner. But recipes, for all their precision and completeness, are poor teachers. They tell you what to do, but they rarely tell you why to do it.

This means that for most novice cooks, kitchen wisdom—a unified understanding of how cooking works, as distinct from the notes grandma lovingly scrawled on index-card recipes passed down through the generations—comes piecemeal. Take, for instance, the basic skill of thickening a sauce. Maybe one recipe for marinara advises reserving some of the starchy pasta water, for adding later in case the sauce is looking a little thin. Another might recommend rescuing a too-watery sauce with some flour, and still another might suggest a handful of parmesan. Any one of these recipes offers a fix under specific conditions, but after cooking through enough of them, those isolated recommendations can congeal into a realization: There are many clever ways to thicken a sauce, and picking an appropriate one depends on whether there’s some leeway for the flavor to change and how much time there is until dinner needs to be on the table.

The downside of learning to cook primarily through recipes, then, is that these small eurekas—which, once hit upon, are instantly applicable to nearly any other dish one prepares—are most often arrived at via triangulation. It’s like trying to learn a language only by copying down others’ sentences, instead of learning the grammar and vocabulary needed to put to paper lines of one’s own.

In our house, I am the one inclined to create a meal out of ingredients we already have. For me, this reduces waste, it encourages the use of seasonal ingredients, and it can result in novel discoveries and surprises which I think more than make up for any occasional disappointments.

My girlfriend, in one of the few areas of life that she is predominantly risk-averse, prefers the more consistent results produced by meticulously following recipes, and is not interested in such outlandish or risky approaches as ingredient substitution or generally making it up as you go along.

We’re both keen cooks with different approaches. I was working on a web design project today, using a pattern library, when it occurred to me: Pinkser is talking about breaking recipes into component methods. It’s an atomic approach.

Atomic design

Brad Frost’s 2013 post Atomic Design explores the building blocks of web design and how they combine:

A lot has been said about creating design systems, and much of it focuses on establishing foundations for color, typography, grids, texture and the like. This type of thinking is certainly important, but I’m slightly less interested in these aspects of design because ultimately they are and will always be subjective. Lately I’ve been more interested in what our interfaces are comprised of and how we can construct design systems in a more methodical way.

In searching for inspiration and parallels, I kept coming back to chemistry. The thought is that all matter (whether solid, liquid, gas, simple, complex, etc) is comprised of atoms. Those atomic units bond together to form molecules, which in turn combine into more complex organisms to ultimately create all matter in our universe.

Similarly, interfaces are made up of smaller components. This means we can break entire interfaces down into fundamental building blocks and work up from there. That’s the basic gist of atomic design.

Frost’s approach consists of five distinct ‘levels’ or ‘stages’ (he uses both terms interchangeably) of atomic design:

  1. Atoms. The core building blocks, e.g. HTML tags, form labels, inputs or buttons.
  2. Molecules. Combinations of atoms, e.g. a working form comprised of a label, input and button.
  3. Organisms. “groups of molecules joined together to form a relatively complex, distinct section of an interface,” e.g. a masthead consisting of a logo, primary navigation, search form and social media links.
  4. Templates. These are groups of organisms which together form a layout. Think of wireframes and mockups.
  5. Pages, which are specific instances of templates. All placeholder information is replaced by real content.

This is useful in many ways. People can focus on things in detail as well as having a broader view. Faced with problems, you can take a particular stage and investigate if the issue relates to how it combines with others, or if the problem is with what the item itself is comprised. You can look both ways for a solution.

Reverse-engineering cookery

Pinsker praises people like J. Kenji López-Alt for their reverse-engineering approach to cookery. Kenji’s The Food Lab column looks at a specific dish, e.g. katsu curry, by tweaking ingredients and methods and deconstructing other cooks’ approaches. He’s looking at the concept of katsu curry and seeing what approaches do and don’t work, relaying this to the reader, who then has all of the following:

  • A solid katsu curry recipe
  • An understanding of why that combination of ingredients and methods works
  • How to adapt the recipe according to what’s available
  • A set of transferable skills, e.g. making cutlets, brining, breading and frying

It offers more information than a typical recipe. It’s recipe and method together. Another proponent of this is Felicity Cloake, who’s How to cook the perfect… column for The Guardian breaks apart cooks’ approaches in search of a single foolproof recipe. There are others, too. Food52 have a Not recipes section which focuses more on a toolkit approach to a dish rather than a single recipe.

Atomic cookery

I hope it is obvious that ‘atomic’ here is unrelated to molecular gastronomy, however that is being defined at the moment. It is also unlikely to be a novel thought. Scientific or linguistic metaphors for cookery are already well-used. But it has given me a new way to think about how we approach these things in our particular house and with our particular approaches. If it’s easy to break apart a set of core recipes, and we understand how best to do this and why, we can reassemble them in new and (hopefully) successful ways.

By drawing a parallel between atomic design and cookery, we have something like:

  1. Ingredient/method/technique, e.g. eggs, deep-frying, julienning
  2. Preparation, e.g. making a soffrito
  3. Component, e.g. ragu sauce
  4. The conceptual ‘dish’, e.g. the idea of bolognese
  5. Specific recipe, e.g. a specific cook’s version of a dish, made up of components

This isn’t quite an exact parallel with the atomic design approach. There isn’t the same progression of increasing nested complexity through the list. A bolognese recipe isn’t more complex than the ‘concept’ of bolognese, for example; instead, it will use a subset of all the possible components.

Pinsker praises Salt, Fat, Acid, Heat: Mastering the Elements of Good Cooking by Samin Nosrat as the best metacookbook along these lines. I’m certainly going to buy the book based on his description of the ‘checklist’ approach it instils in the cook:

Has it been sufficiently salted? How was fat used to inflect its flavor and texture? Is there acid in there to balance out the overall flavor? And should it have been exposed to a different type or amount of heat? This is the book of cooking grammar that so many novices would benefit from.

And experienced home cooks, I bet. I can also recommend An Everlasting Meal: Cooking with Economy and Grace by Tamar Adler, which concerns itself with the grammar of good, economical cookery. I particularly like the section on soaking and cooking dried beans and pulses, and think often of ‘the fat boy in his prime’:

(Some caution to end. I’ve invoked a scientific metaphor taken from web design and also mentioned grammar while discussing food preparation and recipes. There are pitfalls of jumping too wholeheartedly into using these abstractions and metaphors. They might inadvertently impose a way of thinking that is more restrictive rather than freeing the cook to think about things in new ways. In a recent post for the New York Times, John Herrman cautions against the widespread use of the technology ‘stack’ as a metaphor, ending with a quote from computer scientist John Daugmann: “We should remember that the enthusiastically embraced metaphors of each ‘new era’ can become, like their predecessors, as much the prison house of thought as they first appeared to represent its liberation.”)

Emoji as modern gargoyles

James Vincent for The Verge:

Emoji are going to be some of the most recognizable icons of the 21st century, says architect Changiz Tehrani, which is why he decided to cast 22 of them in concrete and use them as decoration for a building in the Dutch city of Amersfoort.

“In classical architecture they used heads of the king or whatever, and they put that on the façade,” Tehrani told The Verge. “So we were thinking, what can we use as an ornament so when you look at this building in 10 or 20 years you can say ‘hey this is from that year!’” The answer was obvious: emoji.

Notes on a case of Nicholas Gurewitch

In December 2014 I backed a Kickstarter project by one of my favourite artists, Nick Gurewitch, creator of the Perry Bible Fellowship:

The project is called Notes on a Case of Melancholia, Or: A Little Death.

This is book about Death’s despair regarding his kid- an affectionate “Little Death” who simply doesn’t have what it takes to carry on the family business.

Dr. Edgar O. Wye is a psychoanalyst who takes Death on as a patient. The book’s rhyming narration will be taken from his case notes.

The book will run about 42-50 pages long, and will be completely illustrated. Graphic novel “frames” will be used on occasion, but this will really be more of a picture book – deliberately similar to the short books of Edward Gorey, but with a character-driven plot. Though it has a pretty high body count, it is in essence a family story.

It’s running slightly late. (About 18 months.) This is mostly due to the painstaking subtractive work required to produce each page: they are created by painting a board with black ink then ‘drawn’ by scratching millions of tiny lines with a scalpel.

Nick and his project were the subjects of a short documentary:

Nick’s just found out that his publisher has folded. It’s not too late to support the project to ensure it appears in a (somewhat) timely manner.

Smash Mouth’s ‘All Star’ as a musical fractal

Interesting bit of music theory/mathematics by Adam Neely:

Neely takes the notes of Smash Mouth’s ‘All Star’ and speeds them up around 1,000 times according to their relative pitch frequency. In short, each note in the original melody is made up of all the notes within itself, just sped up by different amounts. The result is self-similarity: a musical fractal.

‘All Star’ has been elevated from dorky throwaway tune into first a similarly dorky meme and then a core component of several interesting works. Neil Cicierega’s mashup work is a prime example, to the point where I can’t hear Modest Mouse’s ‘Float On’ without internally singing “Somebody once told me the world is gonna roll me…”:

Games that might have been

Babble Imperium:

A couple of years ago someone attempted to make a list of every video game ever made, and put it in a 6.5MB flat file. Like any sensible person, I used it to train a recurrent neural network.

Selected examples:

  • Metal Cat (2001, Sega) (Windows)
  • Spork Demo (?, ?) (VIC-20)
  • Black Mario (1983, Softsice) (Linux/Unix)
  • Soccer Dragon (1987, Ange Software) (Amstrad CPC)
  • Mutant Tycoon (2000, Konami) (GBC)
  • Dick of the King (2007, Activision) (PC-9801)
  • Spork Race (Universe) (1990, Atlus) (Arcade)

The ‘Spork’ franchise sounds like something I’d play, and ‘Black Mario’ seems sufficiently inclusive.

See also these wonderful recipes generated using a predictive text interface:


And Friends episodes: