Metacircular

In response to Jason's comment (reproduced below), I've decided to write up a quick first-impressions of the Pixelbook.

@matigo
> November 3rd 2017 9:35:23 pm
> @kyle nice. I’ve heard a lot of positive things about the PixelBook. Well … aside from the price. A lot of people seem to expect anything from Google to be “free” ?

Funny story: I almost never buy brand new hardware (as in, I rarely pre-order things or want new hardware before it comes out). As I was figuring out that my current setup (an N22 running Debian sid) wasn't quite what I needed, the pixelbook turned out to be the exact machine I (thought I) wanted. I basically followed Kenn White's post "My $169 development Chromebook", though at a cool grand, this doesn't exactly fit the "cheap enough that should it get lost/stolen/damaged, I wouldn't lose too much sleep" description in his post.

So yeah, it's expensive: about 150 USD more than an equivalent XPS-13 (I have one of those, too, that is mostly unused), plus another 100 USD for the stylus. But, my XPS-13 isn't convertible and I use it in that mode about half the time. One of the things that concerns me the most with new laptops is the keyboard, and while this isn't the best, it's certainly good. Plus, I don't have to dick around with Linux, I get a lot of security benefits, and I can still be productive.

It's not perfect; there's still a few things I wish I had on here that I haven't gotten working yet: Racket and/or Common Lisp (okay, I'd even settle for Clojure early on), Tensorflow, and Processing. I do have Python 3, Go 1.9.2, and Rust 1.20.0. I also wish I could get emacs as a first class citizen at the top level. Some form of virtualisation or containerisation would ameliorate pretty much all of these: a single VM or container with an Ubuntu VM that I can SSH into from the Secure Shell with a proper development environment would be acceptable. It looks like some kind of containers or VM support is coming, so there's time.

One of the first questions that I got was "why ChromeOS and why not a macbook?" It comes down to a few things, most notably missing is price (the equivalent macbook air seems to be the same price).

1. I wanted a convertible (the so-called "2-in-1" form factor).
2. From what I can tell on my Macbook Air and Pro, I feel like there's a declining quality and lack of QA in Apple's desktop releases lately. Hot take, maybe, but it's my wallet.
3. I'm currently in the middle of the Udacity AI nanodegree, and having the Udacity app so I can download the lessons and quizzes for my (often) craptastic network connection. Bonus: it turns out the Android version of the Udacity app is significantly better than the iOS version, which came as no small surprise to me.
4. Stylus support in tablet mode is great.
5. I like having an escape key.
6. I'll take Linux over OS X most days.

Currently, I'm able to do my class work; I edit code in Caret, I can run my homework and labs in an SSH session to Termux, I can watch videos offline with the Udacity app, and I can read relevant papers and AIMA. That's what I need (and what I need to focus on) for the near future, which gives time for improvements to come. I haven't even really touched on the Google Assistant integration, mostly due to a slight sense of being creeped out by giving Google that much data.

I started a subsite for keeping my technical posts separate from my non-technical posts, so I figured I should open this up with something and not just sit on the site name.

It's probably no secret, if you know me at all, that I'm more or less burned out on tech. It seems to me (and while this may be unfair, this is my current perspective) that it's all a dumpster fire of poor life choices motivated by what sells the best story. Case in point, this last quarter I had one project I wanted to work on, but that was pushed aside for a PR project that our CEO wanted and for an effort to repair part of our system that sort of caved in, but that we were able to prop back up with chicken wire and duct tape.

When I first moved into this place, I didn't have furniture or internet; I jury rigged a chair out of a pillow and a wall, and I cached the Racket docs. I read interesting books like Paradigms of Artificial Intelligence Programming and Automated Planning, Cities and Complexity, and all manner of random Dover maths books to pick up the bits and pieces of graph theory, lattice theory, or abstract algebra that I needed to understand things. I was excited and motivated to work on hard problems.

Three years later, and I've built at least one interesting thing (the CA behind LetsEncrypt, for which we got a shout-out at 32C3), learned a lot about UIs and security (from Red October), and for the precursor to this latest PR project, read a lot of papers and research material on PRNGs and entropy and the like (only to have our infosec team try to come back and explain these things to me, for which I could only bite my tongue). Lately, though, I've been bored out of mind and not able to work on things I want to work on. It's mostly navigating the labyrinth of deployment pipeline, with all its concomitant obstacles and gotchas.

It's the eternal problem --- build things people will actually use and let your brain rot; or do intelligent, creative, and/or thoughtful things but have no one use your system. Unfortunately, it now feels like my brain is rotting and I'm forgetting how to program; I spend more time linting YAML than I do actually writing software, and the software I write is rather dumb. I suppose dumb is good from a certain perspective, but in this case, we could and I'd argue should be more intelligent about what we're doing.

Not to mention three years of basically being socially isolated in the Bay Area has kind of worn on me, and that enthusiasm hasn't really gotten me anywhere. This has led to not wanting to be on the computer when I'm not at work, and not wanting to go out of my way for stuff at work. I've been doing anything but tech, avoiding meetups like the plague (except Systems We Love, which might be the only tech shirt I own).

Obviously the best possible outcome is either to quit tech and do something else, or stay here and find a happy middle ground. The latter is difficult when you feel almost allergic to tech, but I think I've been getting back to a better place; someplace back to the exploratory, intellectual side of computing. This is starting with working through the Haskell book; Haskell is the language I have the most fun programming in, besides maybe Racket, and I like how it forces me to think. I think afterwards, I'll poke at Racket again, and then start working back through Programming Distributed Computing Systems, which has one of the most accessible introductions to the λ-calculus that I've seen. I have the Kindle version, and started reading it two or three years ago, but never made it far, though my office whiteboard wallpaper strip is littered with λ-calculus and π-calculus expressions from working through the book. It'd be nice to actually get through it, and to maybe start actually building things with Haskell. Hopefully, this rekindles some of the joy that I'd felt when exploring computers in the past. Maybe one day I'll even get through GEB, SICP, and the Architecture of Symbolic Computers. One day.