Can you imagine the sort of worldwide panic that would ensue if everyone’s smart phone suddenly stopped working? Something analogous happened thirty years ago when, on January 15, 1990, the entire US long-distance telephone network crashed for nine hours.

This was a strange, dire, huge event. During the nine long hours of frantic effort that it took to restore service, some seventy million telephone calls went uncompleted.

I was working at Bell Labs at the time, which was a subsidiary of AT&T. Limited and isoated failures, akin to modern outages in your ISP’s internet service, were relatively common in those…

There seems to be a widely held opinion that this election was close.

I think it’s due to the fact that several states, like Georgia, Arizona, and Wisconsin, were indeed decided by less than 1% of votes cast. But when looking at the aggregate outcome, it’s hard to characterize the result as anything but historically decisive.

This article from The Nation notes that, despite Trump’s claim to have “WON THIS ELECTION, BY A LOT!”, you’d have to go back 88 years to find an incumbent loss by this margin — Biden won a higher precentage of the vote than any incumbent challenger since FDR defeated Herbert Hoover in 1932.

The president may deny reality, but he cannot change the fact that he is now losing the popular vote and the Electoral College by momentous margins. @TheNation https://www.thenation.com/article/politics/trump-lost-denial-pompeo/

Originally published at https://mco.dev on November 17, 2020.

Nevada has 86% of its votes reported and won’t announce addtional results until later today. But it’s pretty easy to crunch the numbers because it has only 22 counties.

Nearly all of those counties are red except for Clark and Washoe, which are by far the largest in population. All you have to do is extrapolate the remaining votes, on a per county basis, to break the same way the existing votes have gone.

One pitfall here is timing bias, i.e., the later counted votes within a county may be distributed differently than those counted earlier. That’s quite possible if…

(Visit mco.dev or subscribe to my newsletter for more stories like this.)

America’s rail infrastructure has fallen far behind the rest of the world’s. Here’s how to fix that, and a few other problems along the way.

Under FDR’s leadership, the federal government pulled the United States out of the Great Depression, in part, by funding ambitious public works projects, like building our vast interstate highway system. Now we’re in the midst of another severe depression, calling for another round of big thinking and big investment.

What better priority to focus on than building a network of high speed, clean…

Originally published at https://mco.dev on July 24, 2020.

Dear Smartphone,

We’ve had a great relationship over the years. You bring the world right to my fingertips. And you’re always there for me. Unfortunately, I’m a little too into you.

I found myself spending more and more of my time focusing on you rather than the world around me. It got to the point where I would put you down and then seconds later find you were back in my hand, without even realizing I’d picked you up again.

Lately, most of the time we spend together is not quality time…

Data Visualization

This is the second of a two part series in which we focus on interesting queries and visualizations using the data pipeline we created in part one.

In part one, we covered the problem statement and the data engineering solution, including all the code needed to build a reliable, robust big data processing pipeline. In this part two, we’ll cover some of the fun things we can do once we have our data pipeline running: interesting queries and data visualizations and a Data Studio dashboard you can try yourself.

What can we do with this data?

For starters, let’s find out the most popular Wikipedia article so…

Data Engineering

What’s bigger than Wikipedia? Spoiler: Wikipedia page views. This is the first of a two part series in which we’ll explore how to build a data engineering solution to process all 10TB of published wikipedia page views and entity data.

In part two, we’ll cover some of the fun things we can do once we have our data pipeline running: interesting queries and data visualizations and a Data Studio dashboard you can try yourself.

Introduction

I often give talks to students and I like to ask if anyone knows what’s represented in the cover photo of this article. I find it…

17 Dec 2017 · edu

I spend most of my professional time helping developers understand how to harness Google Cloud technology to build great applications. But a recent engagement brought me somewhere I’d never been before, somewhere, I suspect, no other Google Developer Advocate has gone.

My home country, the United States, has the largest prison population in the world, and the second-highest per-capita incarceration rate. …