The Aggregate Weekly Newsletter🔬September 16th, 2019
Hello! I am Lars E. Schonander, a writer for MediaFile and a blogger on international affairs, tech, and general wonkery. Happy Monday! Here is my weekly newsletter with a weekly analysis with interesting data, along with links related to things I found particularly interesting that week. Any Questions? Send me a message or just respond to this email!
The Weekly Data:
So for my school, I am starting to write my economics thesis on open source technology. I find it strange that while covered by economic departments in the late 1990’s, open source technology has not been an issue covered as much by economics as much after that period.
The main thrust of my analysis has been using data collected from GitHub and Google Trends. The Google Trends data is pretty simple to collect, but the data from GitHub is fascinating, mainly because of how surprisingly difficult it was initially to work with, along with rate limit issues.
For reference, the GitHub API contains lots of nested JSON, so it requires a lot usage of purr
and anonymous functions to grab the desired data. Anonymous functions and the HTTR package go surprisingly well together when you need to make mass API calls.
The three parts of the API I am collecting are stars, contributors, and commits, with the Google Trends data being what I am trying to predict, and what I am using as a rough approximate for “popularity”. To begin, as I said before, the trouble with the API is your roughly rate limited to 1334 pages, which seems like a lot, but you only get 30 stars per page, and some packages, like React, Git, TensorFlow, Keras, and others have far more stars then the 40020 stars one is granted. For reference below, a interesting illustration of this issue when looking at the big three JavaScript frameworks, React, Angular, and Vue, and how many stars they got per year.
Angular lasts the longest, but that’s actually because it has the least amount of stars out of the three major JavaScript frameworks with a comparatively low 51,562 stars. I’m currently chatting with someone at GitHub about this issue, but if you know any workarounds, I’d much appreciate them. In fairness to GitHub, this method is cheaper and less tedious then doing all of this in Google BigQuery.
While the contributor API masks emails after a certain point, the thing that really matters is A, the distribution, and B, the total number of contributors. Like the star API, it has a hard limit, in this case at page 14. As a example, here are the top ten contributors for the TensorFlow repository, along withe distribution of how many commits they actually did.
As one would suspect, it follows a power law distribution, with tensorflower-gardener being a way for commits appearing from Google’s codebase to appear on GitHub.
Finally, for completions sake, below is a example of what graphing repository commits overtime looks like.
Most commits to angular it seemed happen in later years, with commits rocketing to 3,000, and staying their more a less.
Now, some links…
Natalia Mehlman Petrzela (Logic Magazine): Remote Work
The future of high-end fitness may look like a mirror.
Peter Schwartzstein (The Center for Climate and Security): How Jordan’s Climate and Water Crisis Threatens its Fragile Peace
For years, security service recruitment has masked climate instability in rural Jordan. Now that strategy is breaking down and no one knows what will take its place.
In the desert villages of south Jordan, the security services dominate. They run many of the schools. They maintain the roads, water infrastructure, and bridges. Crucially, they also employ most of the men.
Roughly 70% of those in full time employment in rural stretches of the southern governorates are in the army, civil defense, or intelligence corps, according to CCS research conducted in about 20 villages, a figure that rises to around 90% in some of the most distant, isolated communities. Most of the other residents are dependent on soldiers’ spending. Such is the security services’ outsized role that many districts have practically been emptied of young and middle-aged men. “It’s only when the soldiers are back home that this feels anything like a village,” said one farmer in the far southern Aqaba governorate.
Cliff Kuang (Eye on Design): How a Band of Design Misfits Brought Anti-aesthetics to Bloomberg Businessweek
There was a time, around 2013–2016, when the most experimental magazine in the world wasn’t some Berlin fashion zine that doused its models in crude oil but Bloomberg Businessweek, a once-dowdy battleship of American journalism. This was a time when you’d see a new BBW cover about an airline merger that showed one airplane humping another or one that looked like it had been assembled from PowerPoint clip art by some sales manager on a cocaine bender and think, “How the hell are they getting away with that?” And the answer happened to be some strange wrinkle in the universe that aligned certain people with one another, at just the right time, under rare circumstances. “Richard Turley, BBW’s creative director at the time, and I had done conventional work. We had our shot and we didn’t want to waste it,” says Josh Tyrangiel, who was hired as BBW’s editor-in-chief at just 37 years old.
John Moy (Highsnobiety): Fast Food is the New Streetwear, But Is It a Finger-Lickin’ Good Thing?
Streetwear has long relied on a scarcity and exclusivity model, with surprise drops, limited releases, and special collaborations creating a landscape for consumers that is familiar enough to navigate, but still peppered with enough surprises to hold their attention. The voracity and enthusiasm with which streetwear’s consumer base purchases their wares has given notice to seemingly unrelated industries that this model is incredibly successful in generating not only sales, but name recognition and the holy grail for all marketers: organic social media growth.
Lately fast food companies have adapted this model — and it’s working. It makes sense: fast food is based on basic, consistent options, and if you think about it, so is streetwear. Often the only difference between the kinds of products offered by fast food restaurants and streetwear labels is the logo on the box. And similar to streetwear, there’s a wild amount of brand loyalty in the fast food world.
Jon Baskin (The Chronicle Review): Academia’s Holy Warriors
It was a Saturday night in May 2018 in Fox Hill, N.Y., and the Notre Dame political-science professor Patrick Deneen was talking about his garden. The occasion was a conference called “Beyond Liberalism,” organized around Deneen’s unlikely bestseller, Why Liberalism Failed (Yale University Press, 2018). The setting was the mess hall of a “Bruderhof” intentional community, patterned on those established by Christian pacifists in the wake of World War I. Partly, Deneen talked about his garden because he had been on the road promoting his book and was looking forward to spending more time at home. But the garden was significant for other reasons. Every summer in his compost, Deneen said, he made something new out of something old. This was the essence of his ideal of “culture.” Culture was not, as many in liberal America assumed, about liberating ourselves from nature and convention. It consisted, rather, in “responsible stewardship.” Culture was “thick, inherited, and connected to a place,” Deneen said. “It renews itself like soil.”
What I’m Reading
I recently bought and rapidly read NYT tech report Mike Isaac’s new book Super Pumped The Battle for Uber. I thought he treated Travis Kalanick quite fairly in the end. While Travis did do many unethical acts, there is something tragic about how Uber ended up, in how Uber did not have to end the way it ended up. It is also a very intense book, with their being far to many interesting subplots to count, from Benchmark Capital to the variety of internal tools Uber used to against various groups that were hostile to the company.
What might be interesting to watch to as many of the e-mobility startups were founded by Uber alum’s, how much as Uber’s culture spread? Considering Travis VanderZanden, the founder of Bird, worked for Lyft and Uber.
I’m very excited to meet him in person in October when he is coming to D.C. to chat about his book. I’m curious what next big tech company he will tackle next.
What I’m Working On
As seen above, I’m working on my thesis, looking at open source software from a economist’s perspective.
I am also attending the Life Capital Conference in D.C. If you are attending it’d be nice to say hello in person.
Thanks!
Thanks for taking the time to read this, I will be back next Monday. In the meantime, you can follow me on Twitter or reach out via email.