Ex-goat herder. Living on a planet in space. Once and future thing. And this is my blog. It has my name on it, and there are dates, the blog-spoor are everywhere.

Leprechaun dust from DuckDuckGo

I made a heart locket animation for a Valentine’s day project years back, and stuck a generator for it online that recently took on a life of its own when it got adopted as a meme I think because of a cat called Bingus??

bingus my beloved

My dumb little site had a memory leak for years, that I “solved” with a script to restart it every night. After traffic blew up I finally fixed this… by restarting every few minutes hash tag should be ashamed of myself. But traffic continued to grow so I finally read the ancient code and soon had improved server stability by orders of magnitude by cleverly NOT IMMEDIATELY LEAKING MEMORY WITH EVERY UPLOAD.

shameful close

And that was that, people were happily making their memes and my server was quietly serving them. Then hubris hit. I said to myself, St Patrick’s day is coming up, why not add … a green heart locket? That will blow these memers minds. Born and raised on the red heart locket these young folks will say “what is this” and “how can this be” and “all we have known is wrong.” So I fired up Blender 3D and made a tasteful draft of a luminescent green heart with shamrocks plastered all over it:

green draft

Great! To your eyes, this may look grungy and low quality. Well, it is, but that’s not necessarily a bad thing, there’s a whole scene out there with that aesthetic. Or at least, that’s what I use as an excuse for my many technical and artistic deficiencies.

Next, I applied my secret sauce (carefully explained in this pdf, I’m not good at secrets) to convert the Blender design into a precompiled map so users’ uploads can be quickly substituted in for images on surfaces without needing to re-render. That results in a sequence of templates like this:

map files

Along with some other images, these summarize what coordinates of the users’ inputs to lay out, and where. Once that was all done, I went to my site, visited my new generator, stuck in some images and text and voilà:


Irsih is a misspelling of Irish for humorous effect you see, slyly referencing the Fnich meme and oh nevermind. But hang on, what are the funky little white speckles in there? Leprechaun dust I suggested on twitter. I stared at my maps, looking for a rendering problem. Nothing. I tried making animations in different browsers, and found them only happening in Chrome. Betrayal! I dug in to see if there were new canvas options I needed these days (in olden times I’d been bitten by context.imageSmoothingEnabled). Nothing. Finally, I turned off all my browser extensions… and the problem went away. Turning the extensions back on one by one, the speckles turned out to be from the DuckDuckGo privacy extension. Digging into its source code, I found a modifyPixelData reaching into the canvas and randomly twiddling least significant bits to disrupt canvas fingerprinting.

A somewhat helpful, somewhat mischievous entity sprinkling a bit of magic grit into my life. So yes, basically, leprechaun dust.

English to SQL: progess updates

English to SQL translation is getting better! Can you guess which of these queries to find the actor appearing in the most movies (in a fictional dataset) was written by a human, and which by machine?

a human query, or is it

a machine query, or is it

Well ok yes you probably can, but the point is they are both pretty good!

In early 2019, when I was looking for published models for text to SQL translation, I had the most joy with one called SQLova (Hacker News discussion from that time). SQLova was trained on the WikiSQL dataset, a collection of natural language questions and SQL queries for tables drawn from Wikipedia. Part of what made SQLova work well was using BERT to transform its input. BERT is a model released in 2018 for generic NLP tasks, trained on BooksCorpus and English Wikipedia, that led to a wave of state-of-the-art improvements in all sorts of language problems as people bolted it on to their models. That wave continues to this day, and BERT’s descendants get better and better.

SQLova was limited to queries on a single table, which is not the most interesting use to make of relational databases. In 2018 the Spider dataset had been released, with questions and queries involving multiple tables, so I was looking forward to models showing up trained on that. IRNet come out with a pretrained model in late 2019, and I started playing with it. It was great to see joins and more complex clauses coming into play. But a downside of work benchmarked on the Spider dataset was that the benchmark didn’t include what is called “value” prediction (such as figuring out parameters in a query), and research code developed with Spider in mind often skipped that. This makes sense from a research perspective, but generates incomplete queries.

I noticed a network called ValueNet released recently. Building on IRNet, it adds on some machinery for guessing values, and generally makes more of an effort to produce fully executable queries. I appreciate that. Let’s try it out!

As a concrete example, let’s use the Sakila database, as introduced on the jook website. This is a (fictional) database about movie rentals from physical stores back when that was a thing. It has 16 tables, with a rich relational structure:

a database schema

On the jook website, they give an example of a (human-written) query that “finds the actor with most films”:

human query for acting-est actor

Let’s see what ValueNet can do:

machine query for acting-est actor

(I wrote movies rather than films because I’ve lived in the United States too long, but it works either way). The SQL looks plausible! It picked the right tables, relations, and aggregations. But the answer is just an ID. Let’s be more explicit that we want a name:

give me a name

Gina! Looks like the result matches what the human got. ValueNet gave us a name, but in a begrudging way, just the first name. Let’s get more specific again and ask for a full name:

a full name

Neat, it gave us all the name parts in the table. It didn’t know how to construct a full name but who does really. To match the human query, let’s also ask how many films this Gina person acted in:

how many movies

That was a pleasant experience. There was no cherry-picking here, I downloaded the database and tried it on ValueNet and this is what I got. It’d be nice if the network could better guess what full set of information is useful to give, but it was definitely helpful. Also, compared to single-table operation in Sqlova, I am excited to see the network picking good tables to look at from the soup available, and relating them appropriately.

The jook website goes on to give another example (human-written) query to “calculate the cumulative revenue of all stores”:

human query for revenue

That’s a mouthful. Let’s see what ValueNet does:

machine query for revenue

That’s not bad! It figured out the relevant tables and columns. I couldn’t nudge ValueNet to break this down by day though. Understandable, since there is nothing structured around days in the database. I’d expect the next generation of models to be able to get that kind of thing though, given that GPT-3 could no doubt write a detailed essay about it.

As a consolation prize, I asked for a break-down by movie, which I got without fuss:

revenue per movie

Neat! And the great thing is, you don’t need to take my word about any of this, this is not some massive network behind an invite-only API like some networks I could mention 😉. Just go to https://github.com/brunnurs/valuenet and follow their instructions.

My experience with research code in this area is that it is often difficult to run on data outside of the dataset it is benchmarked on, since the pipeline for pre-processing that data can be hard to nail down. I’ve made a docker image specifically for running ValueNet on fresh data, and without a GPU (since you don’t need it for inference really). Also, to make the whole thing self-contained, I ripped out a use of Google’s Named Entity Recognition API and replaced it with something from spacy. If you have docker, and it is configured so it can use lots of RAM (3GB or so?), then you can do:

$ wget https://paulfitz.github.io/files/sakila.db
$ docker run --name valuenet -d -p 5050:5050 paulfitz/valuenet
# wait a while
$ curl -F "sqlite=@sakila.db" -F "q=what is the title of the longest film" http://localhost:5050
# {"result":{"answer":[["CHICAGO NORTH"]],"sql":"SELECT T1.title FROM film AS T1    ORDER BY T1.length DESC LIMIT 1"},"split":"case_1db37f79-bbfe-4d70-a367-13e785f5e180"}
$ curl -F "sqlite=@sakila.db" -F "q=what is the full name of the actress whose last name is 'WITHERSPOON'" http://localhost:5050
# {"result":{"answer":[["ANGELA","WITHERSPOON"]],"sql":"SELECT T1.first_name, T1.last_name FROM actor AS T1 WHERE T1.last_name = 'WITHERSPOON'"},"split":"case_2cbddbed-2704-4037-8a4a-82e9a8a0f3a2"}

There’s more information at https://github.com/paulfitz/mlsql.

How we query databases is ripe for a transformation. Companies have been developing and releasing natural language interfaces for decades. But outside of niche uses they’ve so far been more trouble than they are worth - brittle, inflexible, lacking common sense. But the field of natural language processing has evolved a lot, to the point where we now have access to representations that are far more robust, flexible, and - if not exactly having common sense - encode a lot of knowledge about relationships between currencies, names, chronology, phone numbers, movies, discounts, street names, airports, email addresses, menus, pokemon, and so on. I’m looking forward to database query engines that know know more of that!

Spreadsheets with python formulas

When doing spreadsheet work, I like Grist (disclaimer, I work at the company that makes it), because it lets me use python formulas. Once I know how to access cells and tables, I don’t have to remember weird spreadsheet-specific commands - I can just use python list comprehensions.

For example, when I wanted to wish someone a happy birthday every day for a week, it was super easy to import the latest exoplanet data and do some orbital math in python to see which exoplanets are back in about the same point in their orbit as they were when that special someone was born.

exoplanet birthdays

It is a simple difference but a real pleasure.

By the way no, I wasn’t born in January 8, 1935, that was someone else.

Every row has its thorns



Last year we all got excited about 💩 separated values (.psv) for our data interchange needs. But that is so 2017. This year, it is all about þ separated values. @pobocks posted the complete motivation and spec:

PSV Spec

Technically the spec has a small glitch. Every spec has its glitch. Just like every rose has its thorn.

What even is a thorn

What are the earthworms, field-mice, and platypuses of the economy?

Venture-backed start-ups and public companies get a lot of attention. They are the peacocks and elephants of the business ecosystem. But what are the earthworms, the field-mice, the platypuses? There are lot of interesting businesses out there built with very different DNA. One of them is the Data Commons Co-op, and more of them are members of members of that co-op. Let me break that down.

In the eyes of the law, the DCC is a Mercantile Cooperative, incorporated in the Commonwealth of Massachusetts (opencorporates listing). It is organized as a shared services cooperative, where members band together to lower operating costs for services that they all need.

The Data Commons Co-op greases the flow of data between communities in the cooperative, solidarity, new, call-it-what-you-will economy. The co-op not only serves these communities, it is owned by them.

(from datacommons.coop)

Some of the members of the DCC are individuals such as academics, but most are umbrella or support organizations – nationwide federations (e.g. the US Federation of Worker Co-ops), state-level development organizations (e.g. Cooperative Maine), sector-level organizations (e.g. the Cooperative Grocer Network), and others. So the members of the members of the co-op are thousands of fun places like Local Sprouts, a worker-owned café and bakery in downtown Portland Maine.

Financially, the Data Commons Co-op is engineered by its bylaws to operate at cost. It has a tiny budget, funded mostly from membership dues. In 2017 it has no staff; most of the work today is done by a board of directors, and occasional volunteers (I’m one). There are always new and interesting people joining and leaving.

The main niche the co-op has found for itself is backing member-created maps and directories. For example, the U.S. Solidarity Economy Mapping Platform, solidarityeconomy.us, builds on DCC software, and relies on DCC hosting. The co-op is becoming a pool for census projects and member directories, reducing costs and clearing away grit. It is a place where people go to ask questions about the weird and wonderful organizations shaping the cooperative/solidarity/new economy.

The DCC can’t answer a lot of those questions yet. The find.coop directory was an attempt by the co-op to make a “stone soup” map of all the organizations it (and its members) knew about. The idea was that members would chip lists of organizations they could vouch for, and the directory would pull them all together.

What we’ve learned since making this prototype:

  • Most fresh, reliable information comes from organizations making a determined survey or census of a part of the economy they care about and have expertise in.
  • Drive-by individual contributions can happen, but are not a great starting point for the DCC at this stage of its development. Getting moderation and trust right at this level of granularity seems expensive.

I’m now working on a new map (codenamed “stone souper”) incorporating these lessons. The main change is an extreme streamlining of how member-contributed listings are added.

  • Every listing in the database can be tracked to a contributing member. All listings from a given contributor can be removed or replaced without trouble.
  • That means contributors will feel less pressure to have their listings in perfect shape before chipping in, and can be secure in knowing they can pull the listings if something unexpected comes up.

The main technical change in the map has been to move more intelligence into search and display, so that related listings are discovered, grouped, and summarized appropriately. With the existing directory, it was practically impossible with the volunteer time available to import data that overlapped with existing listings, including previous versions of the same listings (d’oh!)

Excited to be finally building this, and thanks to all the DCC members who’ve been so patient and helpful along the way!

Editing SQL databases in vi/emacs/nano

Often enough, I want to make a small change in a SQL database. Reset a user. Delete a row in a request tracker. And so on. In code, using my favorite ORM, I’ve become used to not worrying about which flavor of SQL database I’m working with. I’d like the same freedom on the command line. So I made catsql for cating slices of databases, and visql / emacsql / nanosql for making quick edits. Just specify a table and any filters you want to apply, and the table will show up in vi (or emacs, or nano) in csv format. Any edits you make will be applied back to the original source.

editing a SQL database with vi

This works fine on large tables. Just specify a filter for the part of the table you want to work on. Only that part will be requested and shown. Under the hood, daff is used to figure out what changes you want and how to apply them with a SQL update.

detecting a pattern with a convolutional network and no data

Imagine a buddy asks you to write a program to detect a simple visual pattern in a video stream. They plan to print out the pattern, hang it on the wall in a gallery, then replace it digitally with artworks to see how they’ll look in the space.

a pattern in a gallery

“Sure,” you say. This is a totally doable problem, since the pattern is so distinctive.

In classic computer vision, a young researcher’s mind would swiftly turn to edge detection, hough transforms, morphological operations and the like. You’d take a few example pictures, start coding something, tweak it, and eventually get something that works more or less on those pictures. Then you’d take a few more and realize it didn’t generalize very well. You’d agonize over the effects of bad lighting, occlusion, scale, perspective, and so on, until finally running out of time and just telling your buddy all the limitations of your detector. And sometimes that would be OK, sometimes not.

Things are different now. Here’s how I’d tackle this problem today, with convnets.

First question: can I get labelled data?

This means, can I get examples of inputs and the corresponding desired outputs? In large quantities? Convnets are thirsty for training data. If I can get enough examples, they stand a good shot at learning what I used to program, and better.

I can certainly take pictures and videos of me walking around putting this pattern in different places, but I’d have to label where the pattern is in each frame, which would be a pain. I might be able to do a dozen or so examples, but not thousands or millions. That’s not something I’d do for this alleged “buddy.” You’d have to pay me, and even then I’d just take your money and pay someone else to do it.

Second question: can I generate labelled data?

Can I write a program to generate examples of the pattern in different scenes? Sure, that is easy enough.

Relevance is the snag here. Say I write a quick program to “photoshop” the pattern in anywhere and everywhere over a large set of scenes. The distribution of such pictures will in general be sampled from a very different space to real pictures. So we’ll be teaching the network one thing in class and then giving it a final exam on something entirely different.

There are a few cases where this can work out anyway. If we can make the example space we select from varied enough that, along the dimensions relevant to the problem, the real distribution is contained within it, then we can win.

Sounds like a real stretch, but it’ll save hand-labelling data so let’s just go for it! For this problem, we might generate input/output pairs like this:

some generated samples

Here’s how these particular works of art were created:

  • Grab a few million random images from anywhere to use as backgrounds. This is a crude simulation of different scenes. In fact I got really lazy here and just used pictures of dogs and cats I happened to have lying around. I should have included more variety. pixplz is handy for this.
  • Shade the pattern with random gradients. This is a crude simulation of lighting effects.
  • Overlay the pattern on a background, with a random affine transformation. This is a crude simulation of perspective.
  • Save a “mask” verson that is all black everywhere but where we just put the pattern. This is the label we need for training.
  • Draw random blobs here and there across the image. This is a crude simulation of occlusion.
  • Overlay another random image on the result so far, with a random level of transparency. This is a crude simulation of reflections, shadows, more lighting effects.
  • Distort the image (and, in lockstep, the mask) randomly. This is a crude simulation of camera projection effects.
  • Apply a random directional blur to the image. This is a crude simulation of motion blur.
  • Sometimes, just leave the pattern out of all this, and leave the mask empty, if you want to let the network know the pattern may not always be present.
  • Randomize the overall lighting.

The precise details aren’t important, the key is to leave nothing reliable except the pattern you want learned. To classic computer vision eyes, this all looks crazy. There’s occlusion! Sometimes the pattern is only partially in view! Sometimes its edges are all smeared out! Relax about that. Here’s a hacked together network that should be able to do this. Let’s assume we get images at 256x256 resolution for now.

the model

A summary:

  • Take in the image, 256x256x3.
  • Batch normalize it - because life is too short to be fiddling around with mean image subtraction.
  • Reduce it to grayscale. This is me choosing not to let the network use color in its decisions, mainly so that I don’t have to worry too much about second-guessing what it might pick up on.
  • Apply a series of 2D convolutions and downsampling via max-pooling. This is pretty much what a simple classification network would do. I don’t try anything clever here, and the filter counts are quite random.
  • Once we’re down at low resolution, go fully connected for a layer or two, again just like a classification network. At this point we’ve distilled down our qualitative analysis of the image, at coarse spatial resolution.
  • Now we start upsampling, to get back towards a mask image of the same resolution as our input.
  • Each time we upsample we apply some 2D convolutions and blend in data from filters at the same scale during downsampling. This is a neat trick used in some segmentation networks that lets you successively refine a segmentation using a blend of lower resolution context with higher resolution feature maps. (Random example: https://arxiv.org/abs/1703.00551).

Tada! We’re done. All we need now is to pick approximately 4 million parameters to realize all those filters. That’s just a call to model.fit or the equivalent in your favorite deep learning framework, and a very very long cup of coffee while training runs. In the end you get a network that can do this on freshly generated images it has never seen:

testing on fresh generated images

I switched to a different set of backgrounds for testing. Not perfect by any means, I didn’t spend much time tweaking, but already to classical eyes this is clearly a different kind of beast - no sweat about the pattern being out of view, partial occlusion also no biggie, etc.

Now the big question - does it generalize? Time to take some pictures of the pattern, first on my screen, and then a print-out when I finally walked all the way upstairs to dust off the printer and bring the pattern to life:

real life testing

Yay! That’s way better than I’d have done by hand. Not perfect but pretty magical and definitely a handy tool! Easy to clean up to do the job at hand.

checking on the gallery image

One very easy way to improve this a lot would be to work at higher resolutions than 256x256 :-)


There’s a somewhat cleaned up version of this code at github.com/paulfitz/segmenty

the year of poop on the desktop



So last month an idea surfaced from @j4mie for an alternative data format: poop separated values (PSV). Here’s the complete spec.

PSV Spec

At first, I laughed. Then, maybe I cried a little – 2016 went in a kind of 💩y direction at the end there. Finally I started thinking. I realized that PSV is a brilliant idea, and here’s why.

What we’re talking about here is the :poop: emoji, standarized as U+1F4A9 PILE OF POO in Unicode 6.0. Look at that unicode for a minute: U+1F4A9. That’s a big number. That’s outside the Basic Multilingual Planes my friends, and into the Astral Planes of the unicode standard. You have to have your act together to deal with this. For example, Mathias Bynens covers all the muddles javascript gets into with poop in Javascript has a Unicode Problem. This is not the 128-odd ASCII characters your grandmother grew up with.

So isn’t that an argument not to use an astral symbol as a separator? Let’s reflect on the failings of CSV. Jesse Donat has a great list in a piece called Falsehoods-Programmers-Believe-About-CSVs, mirroring similiar lists about names, geography and so on. The first 8 falsehoods are all encoding related. With CSV, if your data is a table of numbers, you don’t really have to think about encoding at all. That’s nice right up until the moment a non-ASCII character sneaks in there it all goes pear shaped. But if we lead with a mandatory poop symbol, from an astral plane no less, no-one is going to be able to punt on the encoding issue. You have to get it right up front.

The logical conclusion of this idea would be to borrow a string like Iñtërnâtiônàlizætiøn☃💩 from unit tests and use that as the delimiter. But 💩 alone gets us a good way there, and looks cleaner. Err.

Here’s an example of PSV in action. I’m editing a PSV file called checklist.psv that lists my current goals in life. I’m using emacs to edit the file, and git to view differences against a previous version.

PSV and daff

I’ve configured git here to use daff to view tabular differences cleanly. I’m doing this on my phone because phones currently excel at showing emoji – the same thing on a laptop also works fine but the poop is less cheerful looking.

One danger with PSV is that people could get sloppy about quoting rules. With CSV, you’ve a good chance of seeing a comma in your data, so you deal with quoting sooner rather than later to disambiguate that. The business data I’ve seen in CSV form has never had 💩 in it, so I could imagine someone skimping on quoting. One solution for that is for more of us to put poop in our names and transactions, Little Bobby Tables style (xkcd, @mopman’s company).

There aren’t a lot of programs supporting PSV yet. So far as I know, daff is the first. The purpose of daff is making tabular diffs and helping with version control of data, but until format converters crop up you can use it to convert to and from psv as follows:

pip install daff           # or npm install daff -g
daff copy foo.csv foo.psv  # convert csv -> psv
daff copy foo.psv foo.csv  # convert psv -> csv

Or you can write into the text boxes at the start of this post :-).

releasing a library for many languages using Haxe

npm/gem/pypi/php packages

Every programming language is a special snowflake with its own idiosyncratic beauty. Porting code from one language to another is an art, requiring dodging and weaving to give idiomatic results. If a library you’d like to use hasn’t been ported to your language, one option is to use a foreign function interface (FFI). A lot of reference implementations get written in C for this reason. The result is definitely not a thing of beauty, but it works.

Haxe gives another option to a library writer who needs to support communities using different languages. We can write the bulk of the library in Haxe, have that automatically transpiled to the languages we care about, and maybe add a little hand-written code in each language to make the API feel comfortable. This scales the effort involved way down.

For example, I wrote the daff library in Haxe and publish it to:

It turns out that a bunch of PHP users showed up, giving great feedback. That’s the target I personally know least about and would never have gotten around to supporting without Haxe.

The Ruby language was the one I personally cared most about at the time I started this. Ruby isn’t supported by Haxe, but it turns out to be surprisingly easy to add a target to Haxe that is “good enough” at least to translate code that is just a bunch of logic and algorithms (I did this for ruby here).

There’s an important downside to this approach though: you may not get as many pull requests. Users are not likely to be familiar with Haxe (yet), so working with the source will be a challenge for them. Haxe is a very straightforward, “common denominator” language to read and write – but it is a new language.

How square am I

Let me count the ways:

  • The icon my RSS client uses is of a newspaper printed on paper. RSS, RSS client, newspaper - that counts for three ways right there.
  • My t-shirts say nothing witty. I’m currently wearing one emblazoned with an enigmatic logo for a local bike/walk event and more importantly (if typography is any guide) THE LOGOS AND NAMES OF ALL ITS SPONSORS. There’s a white stain from some goop my daughter spilled on me that I half-heartedly and ineffectually tried to rub off with my nails. Update: I changed t-shirt.
  • My laptop has no funky stickers from events or projects. All there is, is a fading Intel Centrino Inside sticker it came with, that has been gradually rotating under my hand. It is now at a 30 degree angle, with traces of exposed glue on the upper side. I don’t plan to take any action about this anytime soon. Update: the sticker fell off.
  • I don’t have an interesting phone. My phone is as dumb as they come, and not in a hip way, more in a total-neglect and out-of-touchy kind of way. 2016 Update: gave in, now editing this in emacs on an new phone so I don’t even have the retro thing going for me any more. argh how do i type ctrl x ctrl s oh right volume down.
  • I’ve never changed my name on twitter. Why are people doing that? Update: I get this now, thanks to pikesley Update: I mean Only Zuul Update: I mean Galactus of the Left BIG PIPES Nun Of The Above Taylor's Wift Human Blockchain Unicode Batman Lord of the Flues Head of Snorkelling 6MillionBitcoinMan Sharing Economist Lol Cream Peter Gunn Jolly Local Swagman Dandy Highwayman Cubic's Rube Cyber with Rosie Cyber with Roadies Institute of Codeine Santos L. Halper Hugh Jif True Bill Clay Jeremy Kylo Ren New Year's Steve Triangle Man Cognitive Sourpuss Manic Bitcoin Miner Null of Kintyre Safety Third Phineas Gage Thousand-yard Stairs London Supercloud register.register WinstonNilesRumfoord Sabre Wulf Cable Select ISA Bus Sides of March Click Dummy Fatuous Pauper Ringlefinch Alonzo Mosely Ian Bothan R Dweeno Horse it's June When Devs Cry Roko's Obelisk Riemann's Zero Penn Rose-Tile Belouis None Rush Goalie 2-Tractor Auth All Is Aflame Jason Bourne Shell SCSI Terminator X bonsaikitten Kid Charlemagne Fronk-en-shteen Metric Martyr Billy Yum-Yum 2x2 Full-Sack Developer Konix Speedking Metropolitan Liberal Glidd of Glood Sexy Cartoon Fox Elite: Liberal Activist Grudge Gee Suswept Big Dope Shill Britain’s Best Tree Rogue Nun Deus X-Wing Ice-Cream Fan Man of Leisure SCMODS R Dweeno Alternative Fax Auntie Kithera Straight Banana iso8601 or GTFO Fake Gnus Two 9s Uptime Vex Machina several people Proof of Kraftwerk UNNECESSARY HASHTAGS Mutant Tycoon LXC Sale Seb O'Teur Spinach Recall Bactrian YAML Big Daddy Smalls Roy Neary Freeze Peach Stanky Bean Special Order 937 HTTP RANGE 14 Monk and beefhead Chareth Cutestory Me an intellectual 🇫🇷 is 🥓 🇫🇷 is 🥓 🔷 Content Negotiator Gallant and Goofuth Buddhachain Harlan Pepper Kid Charlemagne Heated Gaming Moment Chuckie Egg Stalked by Fitzy Banana Force Kale Ban Anna Force Kale Coughin' Nail Sugar Bear MC Ren & Stimpy Prefab Doubt Graham Snumber PDF;dr 1 like = 1 respect 31 for 2 Brian Force Plus Blue Stratos Hans Gruber TulipCoin No Bitcoin only Zuul Snoke is Jar-Jar Core Stickwit Application Jason Voodoo Cray-1 Onceable Spoon Ball is in parking lot Emotional Support Peacock Cheddar George George Brontosoros Graham Snumber UVB 76 Paywall Fade Ferretocracy Tauntaun Club Han shot first Zoe Trope 12-inch Dance Mix Paperclip Minimizer Delete You're A Count Ball Tamperer Satoshi Station Agnes DiPesto Left-Wing Rabble Rouser Ambitions not Targets Bolshevik Bin Man Bitcoin Binman Dot Matrix with Stereo Sound Clinical Psychologist Corridor of Uncertainty Principle GitHub for Workgroups Tomato Farmer Jewels Remain Gareth's Waistcoat Self-made Rent-a-Lefty Daze of Fuchsia Parsed The Third Summers Brother M'Kraan Crystal Gerald of Galactus Commit Bit Short Leader EMF Camper Techno-Organic Virus Nose For R2 Countless Screaming Argonauts Maribou Stacy Pariah Carey Trente Percenter Fink Angel Bette Davisize 2-Bar Phrase Terry Jen-Mist Don Newkirk Mule With A Spinning-Wheel NATHAN WIND as COCHESE Contempt of Parliament/Funkadelic Hot Dog Jumping Frog Albuquerque C. Bourne-Frait Corvus Glaive Bolivar Trask Former user (Inactive) John Cougar Melonfarm Ice Hockey Hair NICAM Stereo Many Bothans Daryl Canvas Obvious Exemptions For Fishing The π Approximator nineteencharacters Peter Hitchens Boycott NPM Sbeve Rogers Invisible Border a² + b² = c² Pharell Hogg Pearsquoosher 200 OK BOOMER Jeff Porcaro fan account PUT Malone egg on branch 5G REZOTONE SHIELD Impatient Zero Grudge Unit Social Distance Warrior Love In A Time Of Covid Nerd Immunity Staying In For The Summer Happy Maundys The Longest Good Friday Guns don't kill people Tories do Covid Rules Everything Around Me Mars Bar Paywall Unbeaten Covid Quiz Participant Some sort of database in the sky Edison's Boyhood Giftshop Join A Union Today Severe Childcare Issues The Rules Are Optional Guns Don't Kill People Tories Do Join A Union Today Gil O'Teen Harold Hill of Harold Hill Did I learn nothing from Richie Aprile? PUT ON THE F***ING MASK William Butler Yeets :(){ :|: & };: Sneedle Flipsock Mute Ant Al Go Rhythm Wakanda Forever Jen d'Reveal Technical Co-Flounder Mo Sizely Senior Mouse Pusher copy of_copy of COVID (1).Oct20.xls(1) FINAL..xls Tier-2 of a Clown Tier-1++ Hello John Got A New Motor? Sabre Wulf 16 Czechoslovakians Freeman Hardy & Willis On The Land The Dukes of Hazzard in the Classroom I was brainwashed Feigning Concern For Fishermen .
  • The things I’m putting on this list and not putting on it no doubt reveal assumptions I’m making about what is hip that confirm further my squaritude in ways I can’t even imagine.
  • I totally get grumpy about the dumb things that kids are doing these days.
  • I totally get grumpy about how much sense the dumb things that kids are doing these days ends up making in the end.

Diff and merge CSV files in your git client

Sometimes, you want to version-control your data. As a programmer, many of us are used to putting everything in git. For large datasets, that is currently a recipe for sadness, but smaller ones can work just fine.

There are a few hurdles. CSV files have a special tabular structure that git knows nothing about. This means that diffs will be noisier than they need be, and that git may see conflicts when merging where there are none.

For diffs, James Smith has a great explanation and a good start at a solution. In the git client, he proposes a custom git diff driver that understands CSV structure. On the server side, he shows how to tweak gitlab or (via a plugin called CSVHub) github to get pretty diffs using the daff library.

For merges, on the client side, the coopy library has for some time provided a similar merge driver to let git understand and use CSV structure. As of today, daff can do the same, and it is much easier to install.

$ npm install daff -g
$ daff git csv

Once that is installed, you’ll get nice diffs produced by the same library James used for his github plugin:

Random diff

And you’ll get nice merges too. Let’s look at an example.

Suppose we have this table stored in digi.csv:


And in one branch we correct thre to three, and in another branch we correct 33 to 3:

one,1	   	  one,1	   
two,2	   	  two,2	   
three,33   	  thre,3	   
four,4	   	  four,4	   
five,5       	  five,5     

If we try to merge these files in vanilla git, we’ll get an ugly conflict:

Auto-merging digi.csv
CONFLICT (content): Merge conflict in digi.csv
Automatic merge failed; fix conflicts and then commit the result.

And the CSV file is no longer a valid CSV file, which is unfortunate if we’re using a CSV-aware editor for it:

<<<<<<< HEAD
>>>>>>> 634275495ecd86c287e292e2719e89a9c1188ed1

With a CSV-aware merge driver, we get:

Auto-merging digi.csv
Merge made by recursive.
 digi.csv |   12 ++++++------
 1 files changed, 6 insertions(+), 6 deletions(-)

and the changes are correctly merged:


What if there was a real conflict? Suppose we replaced thre with three in one branch but thirty-three in another? We are told:

1 conflict
Auto-merging digi.csv
CONFLICT (content): Merge conflict in digi.csv
Automatic merge failed; fix conflicts and then commit the result.

And the conflicted file looks like this:

"((( thre ))) thirty three /// three",33

This remains a valid CSV file, and so can be edited in a CSV-aware editor - we aren’t suddenly kicked out into needing a text editor.


Many years ago, I had a list of hobby projects I worked on from time to time, each with a little summary that began: “Until Google solves this problem nicely, …” Most of these problems have now been solved, except this one:

I’d like to be able to communicate with aliens over great distances. Until Google solves this problem nicely, I’m working on a cosmic OS.

So Google hasn’t yet sorted this out, but Hans Freudenthal made a great start back in 1960 with Lincos, a “Language for Cosmic Intercourse.” Lincos starts out in a by-now conventional way (though it was inventing the conventions) with 35 pages describing a message for teaching basic math from first principles. Then it moves on to 12 pages on time. Then (and this is where things get very interesting) a whopping 79 pages on behavior, with imaginary conversations between imaginary personalities called Ha and Hb.

Chapter III: Behavior

Ha and Hb discuss mathematics, since that’s about the only topic for conversation, but that is arbitrary. In their discussions, they introduce useful ideas such as good and bad (in the sense of constructive versus non-constructive). Now we are getting somewhere.

I was struck by the value Freudenthal was able to get from descriptions of extremely basic conversations, and wondered, what could we communicate through richer interactions? What if we described simulated environments that could actually be evaluated, and played forward or reversed, to see full simulated encounters take place? That was the seed for CosmicOS.

The idea with CosmicOS is to start with math, as Freudenthal did, and then build from there to a basic programming language, and then from there to programs and simulations. CosmicOS compiles down to a series of four arbitrary symbols that could be encoded and transmitted any way we like:

CosmicOS as digits

In human-readable form, it looks kind of Lisp-y, since that happened to be the syntax that introduced least complications. CosmicOS is communicated as a long series of definitions and demonstrations:

Human-readable CosmicOS

The initial language isn’t super important, because we quickly bootstrap to any language we want. At the time I was writing this part, I was keen on Java, so I wrote a translator for it, targeting what has to be the least efficient JVM ever written.

   (field q ((int) new))
   (method <init>-V
     (lambda () /
      let ((vars / cell new / make-hash / vector
                    (pair 0 (self)))
           (stack / cell new / vector)) /
      state-machine (vars) (stack) / ? jvm / ? x / cond
         ((= (x) 0) (jvm aload 0))
         ((= (x) 1) (jvm invokespecial <init>-V 0 0))
         ((= (x) 2) (jvm aload 0))
         ((= (x) 3) (jvm iconst 0))
         ((= (x) 4) (jvm putfield q (int)))
         ((= (x) 5) (jvm return))
         (jvm return))

Then I wrote a little maze game in Java, shoved it in the message, and promptly dropped the whole project for several years :-). But now I’m back and fiddling with it again, mostly at my son’s goading. I’ve brought the project up-to-date enough to be able to get pull requests. You should contribute! You know you want to.

And just so its clear: I don’t have any particular belief in extraterrestrials or any special reason to be interested in contacting them. It is an interesting puzzle though, figuring out all the different ways we might try to do so. You may also want to check out the recent Archaeology, Anthropology, and Interstellar Communication book.

The Data Commons Co-op

The Data Commons Co-op is a quirky start-up that I’ve been helping out with. Its job is to maximize the impact of the data held by its members, and reduce costs in managing it. Its members are “alternative economy” organizations of all types. Dan Nordley calls it “perhaps the geekiest of all cooperative organizations on the planet!”

DCC Retreat

The infrastructure for collaborative data projects could be a lot more fun than it is now. Open Data initiatives are pushing things forward quite a bit, primarily with government data in mind. That is sort of a top-down direction of data flow. We’re looking at bottom-up, grass-roots economic organizing. Worker co-ops, buying clubs, community gardens, time banks, and so on. There’s a lot of overlap in communities, and potential for network effects. The Data Commons Co-op is a way to pay for the infrastructure that every one needs and no-one can make happen alone. So far we’ve produced a simple diff format for tables documented on the Data Protocols site (some background in an Open Knowledge Labs post), along with two programs called daff and coopy for comparing and merging table versions. Beyond the technology, we’re also figuring out how to a culture of sharing can work in the economy. There’s a lot of reflexive data-hoarding and hiding that goes on, which is totally understandable. For individual organizations, the cost of thinking about all the issues around sharing data can outweigh by far any potential benefit. Hopefully the DCC can tilt that equation!


I wrote daff to better visualize diffs between tables (daff = data diff). You don’t need this if you work with append-only data, for example a stream of events churned out by a sensor or bureaucracy. But if you have a collection of assertions that can change with time or need correcting, then data diffs are handy.

bridge diff

daff can be used from that command line, as a library, or on github, using James Smith’s CSVHub. CSVHub can convert a diff like this:

bus diff line based

to something like this:

bus diff

Fragments of identity

Hello good evening and welcome. I'm so totally going to make a website one of these days. In the meantime, here are some fragments of my identity.