## How do I begin?

God, Strangeloop is so peculiar and fantastic. The night before, I got
in bed at 10:30 but didn't fall asleep until after midnight...which
makes Strange Loop better than Christmas. I boarded the plane to St
Louis meaning to sleep but wound up reading 100 pages of Neuromancer
while listening to the [Fez](http://disasterpeace.com/album/fez) and
[Journey](http://austinwintory.bandcamp.com/album/journey)
soundtracks. By the time I hit the ground I was in no mood for a nap.

In the hotel lobby I chatted with [Jonathan
Penn](https://twitter.com/jonathanpenn) and [Sarah
Doukies](https://twitter.com/sadukie) which turned into lunch at
Pappy's Smokehouse. After lunch, I bumped into Alex Gaynor and asked
him some questions about Topaz, RPython, and the recent [One VM to
Rule Them
All](http://lafo.ssw.uni-linz.ac.at/papers/2013_JVMLanguageSummit_OneVMToRuleThemAll.pdf)
paper. Then I checked into my room, dropped my bags, and read some
more Neuromancer. When I came back down, I bumped into Jack Moffitt
and was swept away to dinner with a bunch of Mozilla Research folks
and PLT Scheme academics to be regaled with discussion of: Measure Theory,
Macro Systems, Theorem Provers, Operational Semantics, and other
abstract nonsense. Well, GC for Rust and the state of Servo came up
too. Jay and I discussed a macro system that his student [Blake is
working on](https://github.com/bdj/pldi2013), upcoming Racket VM work,
and a functional game engine on the way back.

To wrap up the night, I went to the Stripe drinkup to see my buddy
[Andreas Fuchs](https://twitter.com/antifuchs). There Peter Wang of
Continuum Analysis bumped into my table. The discussion started as
arguments about impostor syndrome and the best 5 TV shows of the last
decade. It ended as a serious discussion about the rise of
Heterogeneous Computing and how to map a data description language to
LLVM IR for optimally performant, power-efficient high performance
computing. Array languages, cache locality, verilog, NUMA
architectures, you name it, we covered it. Oh, and someone asked me at
the bar if "you're the guy who did a Nintendo Emulator in Lisp". Where
the hell else would *that* happen?

## What do you program for?

It feels dumb every time I vaguely go on about the data analytics
company I work for in a mostly business-logic, not hard-algo
capacity. It feels dumb to say my most prominent hobby project for the
last year has been a Nintendo Emulator in Common Lisp ... that isn't
finished yet. There are a decent number of smart folks doing
consulting or boring web app business logic like I am ... but you're
constantly surrounded by folks that are thinking 5 to 10 years out
from a dozen angles you haven't considered. Or academics who are
equally intimidating in a different fashion. Last year that was
thrilling but this year its a bit terrifying. I just don't have a good
answer to what my motivation in programming is right now. These are
the people I want to be around, these topics interest me a great
deal. It is hard not to feel like I've got a long, very uphill road to
get to the place I want to be technically. But I'm just not on that
road right now. One thing is for sure: If I can just manage to sleep
tonight, I'm going to learn a hell of a lot tomorrow.

## Addendum

I'd like to do a related post sometime on burnout. For me, burnout
seems very tied to the insane churn of this field. Academia does hard,
meaningful work that gets ignored. A very few brilliant folks work on
reinventing computing and never ship much (think VPRI). 1% Engineers
work at megafirms trying to change the world and stay ahead of the
train of people hoping to eat their lunch. A huge segment tunes out
and just writes whatever language they know as long as they can stay
employed. The 10-20% in the middle with their heads up? It's just
terrifying. Quoth Milosz, "I thrash in the bed of my style, searching
for a comfortable position, not too sanctimonious and not too
mundane."

I think what this all boils down to is that I've lost the ability to
convince myself that my hobby hacking, particularly the emulator, is
meaningful. I still can find some meaning in it ... but I can't help
but feel that I've been working on it a year, don't have very much
done, and haven't learned enough. If I really finish a readable pure
CL Nintendo emulator, who gives a damn? I've still failed to implement
Depth-First Search on a whiteboard in the past year. Moreover,
everybody else is out there kicking ass and solving "real
problems". Have I done hard-client side stuff? No. Played with Machine
Learning? No. Distributed Systems? No. Logic Programming or new
languages? No. I just wrote a fucking unfinished emulator that isn't
terrible to read. Whoopee.