twitter and politics

You might have heard that there will be a presidential election in the United States in 2016.

There’s always some form of political background noise on Twitter. Politics come with the news, and Twitter excels at news. In 2012, Twitter was hitting critical mass with some subset of the general population just in time for that presidential election. However, it seemed more-or-less clear that Barack Obama would be the Democratic nominee and almost as clear that he would be re-elected. Only the Republican half of the equation needed a candidate, but it provided an incredible amount of fodder thanks to the rise of Tea Party politicians within Republican ranks and the continuing trend (in both parties) towards epistemic closure.

This is the first election where Twitter has such prolific reach and both parties must nominate a non-incumbent candidate. I feel like the result is a lot of Twitter-based posturing and direct attacks by candidates. The number of candidates seeking nomination would logically mean an exponential growth in the number of such exchanges.

I find policy discourse extraordinarily interesting. Unfortunately, the mass appeal of actual discourse is limited, and this problem is only worsened by the constraint of 140 characters and an audience with a limited attention span. As a result, I get pretty frustrated with politics generally – but particularly pandering by politicians – on Twitter.

As a result, I try to limit expressing my support for any particular politician – or even political party – on Twitter. Oddly, I also try to suppress any desire to engage in policy discourse there – I do change my mind on things every so often, and unfortunately the medium doesn’t lend itself to noting such a change. I have fantastically smart and interesting followers, but no one is immune to the emotion-removing powers of text.

I do occasionally end up leaking my opinions through what news I relay – recently, things like Ferguson and police brutality, electronic surveillance, and this year’s Supreme Court rulings.

I live in San Francisco, but I never really let go of Missouri – a swingy-split state where the 2008 presidential election was decided by a margin of 4,000 votes, or 0.1%. As described by Wikipedia:

Missouri has been known for its population’s generally “stalwart, conservative, noncredulous” attitude toward regulatory regimes, which is one of the origins of the state’s unofficial nickname, the “Show-Me State.”

The past two presidential elections have marginally gone to Republicans, but the Democratic governor was elected by a substantial gap in 2008 and was re-elected 2012 in spite of a stronger Republican presidential victory.

That anti-regulatory attitude flies in the face of an apparent escalation of authoritarianism in both parties.

In other words: I expect I’m a bit politically ambiguous to the world, and that’s half-intentional and half-incidental. I still haven’t found a great way to map my beliefs onto political candidates, but I tried ISideWith to get a starting point. The results were pretty surprising, but I’ve still avoided professing any allegiance so far.

That led me to a question: just how much information about my politics do I leak through Twitter?

The Survey

I asked my followers to fill a survey about what they thought my candidate possibilities and favorite would be. This was posed as the same question in multiple choice and single-choice formats. The idea here is that usually a person has an ideal candidate (the single-choice) and viable competitors in some kind of stack rank in case the ideal candidate is not nominated (the multiple-choice).

I’m also interested in the echo-chambery nature of Twitter. I expected that more than half of my followers would think support the same candidate as they do, so I included the single-choice question but for the follower’s preference. (Not that they would be wrong, necessarily…)

I wanted to have an equal number of candidates from either party, but the Republican race is vastly more fragmented this early in the process. I based my inclusions on the top competitors in several nomination polls aggregated by RealClearPolitics. I included only the top 6 Republicans. The Democratic race seems clearer, with Hillary Clinton and Bernie Sanders appearing as the only viable contenders. (I have to say, I didn’t even realize Joe Biden was even in the possibility set until seeing him in the polls, and apparently he still hasn’t professed a bid. The fact that he leads other bidding candidates is probably a sign of how secure the race is between the two Democratic front-runners.)


I was sort of bummed that I could only convince a dozen people to complete the survey. I’m unsure if that biases these results at all – it may be that people who know they disagree with me on some thing or another believed I was doing this for some nefarious purpose. One person abstained from trying to guess my preferences, and I’ll respect their choice. :)


Screen Shot 2015-08-12 at 7.19.08 PM

As expected, the results leaned Democrat. Also not surprisingly, Walker/Cruz/Huckabee were correctly excluded completely. The tie of Clinton/Sanders/Biden was a bit surprising to me only because of its exactness.

The interesting results are in Bush and Carson.

Bush is currently the Republican frontrunner. I initially wondered if maybe people thought it was a feasible hedge against the possibility of Sanders as the Democratic nominee – that maybe it looked like I couldn’t handle going that far left, that I couldn’t handle voting that far left. However, the preference results show, I think clearly, that that wasn’t the case.

Carson’s inclusion could be for any number of reasons. I do think he has a few interesting positive traits among the Republican candidates – he is not a career politician, he once espoused reasonable views on end-of-life care, and it sounds like he once supported a single-payer healthcare system. Unfortunately, he seems to have rescinded those opinions.

That brings us to one more surprise: Trump is the front runner, and any kind of feasibility hedge mentioned above would have to include Trump as a possibility. I gave him some very very faint praise during the debate because – like Carson – he once supported a single-payer healthcare system.


Screen Shot 2015-08-12 at 7.43.23 PM

This is much more interesting to me.

Clearly I’m appearing solidly Democrat. But the selection of Sanders as my first choice is really interesting.

The entire Sanders platform seems to be based on regulation and expanding the size of the federal government – precisely what I expect, as a Missourian, to be diametrically opposed to.

There’s something more there, but I’ll come back to it in a moment.

Follower Preferred

Screen Shot 2015-08-12 at 7.55.17 PM

No real surprises here. The Youths™ love some Sanders.

Where I may or may not have screwed up

I mentioned that my Missouri ties give me an incredulous attitude towards government regulation. That sounds a hell of a lot like libertarianism, though lately I find that its philosophy and pragmatism are conflicting for me. I’m also a fan of ending warrantless surveillance on American citizens and reversing police militarization. So where is Rand Paul in this equation?

“Not polling well” is the answer. I thought including him in spite of this would be a bit of a “tell”.

What’s the truth?

I’m deeply conflicted admitting anything in any public way whatsoever, but given that it will likely come out over the next year or so anyways, here’s what I feel comfortable talking about on the two candidates who topped my ISideWith and, methodological screw-ups aside, this survey process.

Bernie Sanders

The first four issues listed on Sanders’ campaign website – income and wealth inequality, political fundraising reform, job creation, and racial justice – are very important to me. He also recognizes our underdeveloped infrastructure as a priority.

On the other hand, I’m not sure I agree with his proposals on how to make those things happen.

One other bothersome issue is his rhetoric – the pandering I spoke of earlier.

Screen Shot 2015-08-12 at 8.13.31 PM

I can’t decide which would be the worse scenario – he actually doesn’t understand, or he is pretending he can’t understand to pose for a sound bite.

If a car loan defaults, you can repossess the car and get your money back. If you finance someone’s degree in Typewriter Engineering and they are unable to pay you back, you unfortunately are unable to repossess the knowledge imparted to them, and if you could, it would be as worthless as it was when it was imparted. (None of this violates the theory of Endogenous growth. It would be irrelevant to know this “why” if we individuals didn’t directly pay for college, but that’s not the question he’s posing.)

Bernie Sanders, nonetheless, topped my ISideWith survey when I performed it exhaustively.

Rand Paul

To start, I don’t think there’s a chance in hell Rand Paul gets the Republican nomination. So maybe I was right to leave him out of the survey.

Rand Paul shares my views on ending government surveillance and enforcing the Fourth Amendment, demilitarizing the police, reforming criminal justice, term limits, and avoiding military conflict to the greatest extent possible.

Unfortunately, he’s not a slam-dunk either. It sounds from his platform that he would be in favor of rapidly dismantling as many federal programs as possible; when it comes to these things, I prefer a measured and consensus-based approach to avoid unanticipated second-order effects. (edit: In case it’s not clear, there are also some things that should probably be left alone.) There are some other wacky, more personal things that I don’t feel entirely comfortable sharing publicly.


Lack of inclusion of other candidates doesn’t mean I wouldn’t vote for them – it just means it’s a much more difficult choice among them, particularly since at this point the platform differentiation is minimal within parties. I expect that will change over time.

So who’s it gonna be?

I have to admit that I haven’t studied most of the candidates in-depth beyond the few I’ve mentioned above. But I’m not ashamed of that: I don’t want to make the policy-to-politician mapping sooner than I need to, because platforms evolve and I want to make policy the first-class citizen.

I’m still not committed to a candidate, and if I were, I’m not sure I’d want to admit it right now. Candidate-based politics are almost entirely liturgical and are divisive. Eventually, the process is necessarily divisive in the absence of runoff-based voting, and I’m okay with it at that point. Until then, maybe we can get together and bend some platforms.

In spite of my mentioned disdain of politics above, I’m looking forward to discussing this on Twitter!

WWDC 2015: Session Video Guide

(UPDATED: 14:05 PDT, 12 June 2015)

If you’re looking for a starting point to traverse the immense number of session videos now available from WWDC 2015, I’ve listed my favorites below in an order I think makes most sense if you have no ordering constraints and you’re starting fresh. I’ll be adding notes throughout the weekend, but I wanted to get a quick list up for anyone who wants to get started right away. (Scroll to the bottom for more info on the underlying rationale.)

Phase 1: Prepare yourself

  • Keynote – duh.
  • Platform State of the Union – Get the global view of what are thought to be the most important features of the new releases. (I found that this year it skewed pretty far from what I thought was important, but it’s good to know what’s on their mind.)
  • What’s New in Xcode – No matter what, you’ll be using Xcode 7 soon.
  • What’s New in Swift – If you’re going to be using Xcode 7 soon, and you use Swift, you must migrate your Swift code to Swift 2. Error handling and method availability APIs are all-new and very useful.
  • Introducing WatchKit for watchOS 2 – Big changes are afoot for watchOS 2. If you haven’t already built a WatchKit app, this is more compelling – WatchKit 1 apps will continue to work on watchOS 2, albeit with their characteristic slower UIs.
  • What’s New in Cocoa (Mac dev only)
  • What’s New in Cocoa Touch – Hope you read between the lines last year and migrated to size classes and Auto Layout! Multitasking relies on these features to make your app look good. I think this is also the start of a thematic thread for iOS this year: UIStackView.
  • What’s New in Storyboards – I’m a fan of Storyboards, and this year we got a few key feature updates, but one stands out to me: Storyboard References. You can break your massive MainStoryboard.storyboard into smaller functional segments but retain the benefits of other Storyboard features.
  • UI Testing in Xcode – MY PERSONAL FAVORITE NEW FEATURE. Similar to KIF, but with a VERY fancy test recorder. Writing new UI tests looks like a breeze with the recording feature. You can also take screenshots during UI tests – perhaps we can use this to automate taking App Store screenshots.
  • Continuous Integration and Code Coverage in Xcode – A close second for my personal favorite new feature. The code coverage is based on llvm-cov, which as been around for a while now. The combination of CI, Xcode’s new UI testing, XCTest unit testing, and code coverage data is going to be amazing.

Phase 2: Swift updates and depth

  • Swift and Objective-C Interoperability
  • Improving Your Existing Apps with Swift
  • Protocol-Oriented Programming in Swift – If you’re finding yourself force-casting and force-unwrapping everywhere in Swift, or just perplexed, this can reveal important features of the language that will explain your problems and even give you architectural enhancements. (Some found the pedagogy of this session confusing, but the underlying concepts are very important and useful.)
  • Swift in Practice
  • Optimizing Swift Performance

Phase 3: Important iOS 9 Features

  • App Thinning in Xcode
  • Getting Started with Multitasking on iPad in iOS 9
  • Introducing Search APIs
  • Seamless Linking to Your App – This is an extraordinarily important feature for anyone whose iOS app mirrors content on the web. You can now “hijack” http:// links to your site to open natively in your app, saving a ton of JavaScript and redirection hooey. The user experience will also be vastly improved.
  • Introducing On-Demand Resources (for apps with assets) – Very useful for the class of apps it targets like games and media; not so much for most utility-based apps.

Phase 4: Important WatchKit Features & Info


  • WatchKit In-Depth, Part 1 – Covers migrating your watch extension from living on an iPhone under watchOS 1 to living on the watch under watchOS 2.
  • WatchKit In-Depth, Part 2
  • Creating Complications with ClockKit – A highly-anticipated feature for watchOS 2, but it’s best applicable to a relatively narrow set of apps.
  • Designing for Apple Watch – tl;dr: Do fewer things, and be careful about your presentation format.

Phase 5: Layout Improvements & Techniques


  • Mysteries of Auto Layout, Part 1 – Here we continue the theme of “adopt UIStackView and stop being clever”.
  • Mysteries of Auto Layout, Part 2

Phase 6: Other iOS 9 Improvements


  • What’s New in Core Data – A new batch deletion method is stunningly useful. Scott Perry gives us a great Core Data performance debugging example.

Everything Else


  • Cocoa Touch Best Practices – Revisit core parts of UIKit.
  • Introducing the new System Fonts – San Francisco!


We will be able to use Swift 2.0 features in shipping apps as soon as Xcode 7 is GM, but iOS 9-only features are almost… advisory… unless you’re going to a 9.0 deployment target. There are also a lot of high-detail talks that will be great references after implementing these new features, but that are less useful before then.

blog reincarnation exists – again