My first and last Strange Loop

Strange Loop 2023 (Sep 21-22) was the final Strange Loop. It was also the first Strange Loop that I was able to attend; thanks to my employer Sourcegraph.

Below are some of my unfiltered thoughts on the event, and a bit of my personal history relating to Strange Loop.

I’ve tried to do as little editing as possible after my first pass, which is not generally how I write blog posts. So if it seems a bit rambly…

Anyways, without further ado.

Some time around 2016-2018

Even though I joined a physics PhD program excited to do research, I became disillusioned with it in not too much time.

One was applications, where they just felt impossibly far away, if at all possible. The other thing was, most of my day-to-day work was writing code. However, my research code felt unpredictable, brittle, undebuggable, unstructured, and I didn’t feel like I was learning how to write better code (despite the amount of time I was spending on it!) by talking to my PI or the postdoc who was advising me (I don’t fault them for it; I’m sympathize with why they spent most of their time on research questions rather than software engineering concerns).

During this time, I would look online for various resources related to programming. And one such resource that I stumbled into was Strange Loop.

Strange Loop didn’t really teach me how to write better code, at least not directly. But it gave me something very important. It showed me that writing code could be fun, and it didn’t have to be detached from people’s everyday lives, and even debugging could be an adventure rather than a slog.

It gave me a new source of joy, and excitement, and wonder. Wow, there’s so many people doing so many cool things with code! And even though I had much less training in how-to-code compared to how-to-do-physics, I was still somehow able to kinda’ sorta’ understand what these people were saying, moreso than the physics seminars that I attended. (This was both encouraging and depressing to some extent.)

Presentations didn’t have to be dry or devoid of emotion, they could come alive and have strong narrative arcs. You didn’t have to make hand-wavy justifications for why you were doing something, you could just do things because you thought they were cool.

Strange Loop reminded me of the fun of exploring problems that I used to have earlier when I was learning and enjoying physics.

March-Apr 2023

I’d seen the announcement for Strange Loop 2023, and how it was going to be the last Strange Loop. I was sad about it, but I decided that I was going to attend one way or another.

Over the past few years, as I’d continued following Strange Loop online, I’d grown to wish that some day, maybe I’ll make something awesome that would be worthy of presenting at Strange Loop.

So I drafted a talk proposal and an outline for one of the projects at work that I thought would be interesting to folks, especially PL nerds/dev tools folks.

But the outline was clocking in around 20-23 minutes, whereas the talk proposals being solicited were only for 35-40 minute slots. I personally really dislike talks which seem to have fluff or padding, so I didn’t want to just add filler material that wasn’t exciting. And overall, the content in the outline I’d prepared just seemed “interesting,” it didn’t meet my internal bar for “wow, it would really be cool to share this with people.” And I didn’t have any “oomph” ideas that would make it better as a talk, compared to a different format like a blog post.

So I scrapped that.

It’s fine, I told myself, I was going to just attend as a regular attendee. Based on my previous experience at RubyConf 2022, I further rationalized this as a good thing “hey, I will be able to socialize with people more freely, and won’t be constantly stressed/miss out on things because of my desire to do final practice runs and last-minute polish.”

The day before: Sep 20 2023

For a couple of months, I’d been thinking about getting some cute nail art right before Strange Loop. Maybe different kinds of knots, a trefoil knot, the unknot, maybe a Moebius strip, maybe show some Reidemeister moves somehow. The threads would be green, and the background would be black.

I have zero nail art skills.

I got to my hotel in St. Louis around 6pm the day before. Most nail salons in St. Louis seemed to be closing around 5pm-6pm.

I knew there was going to be one nail salon open till 7pm based on Google maps. so I walked there.

That nail salon seemed to have gone out-of-business.

Oh well. Insofar as this was a start, it wasn’t a good start.

The conference: Sep 21-22 2023

Overall, I really had a lot of fun at the conference. Many of the talks were quite cool. Out of the regular track talks that I attended, the one by Ben Titzer about the Wasm engine Wizard was probably my favorite. Obviously, there’s an element of bias here, I’m a big fan of designing systems so that debuggability and observability are affordable from the get-go and in the default configuration.

With conferences, the people are really what make it special, and Strange Loop’s talks always attract a lot of interesting people.

I got to chat with researchers working on papers I’ve found interesting, PhD students developing cool ideas related to education, people hacking on next-generation text editors and diagramming tools, folks driving the development and adoption of languages new and old, people hacking on fun and whimsical ideas, people who’d done cool talks after digging into the history of certain ideas in programming, and everyone else around and in-between.

As a PL nerd, I also got to hear many different but well-thought-out opinions on programming languages compared to, say, the average online discussion on large forums.

For some time slots, when I was helping out at the Sourcegraph booth, it was fun to talk about some of the stuff we’ve been working on, and have some of our approaches and tooling resonate with other devs.

I’m very grateful of having had the chance to meet so many cool people working on very many cool things, with a perceptible sense of hope and optimism, despite the challenges and uncertainties involved. It made me feel recharged.

I hope that I was able to share some of the same energy in my interactions with others.

Of course, not everything was perfect.

I was quite tired after the first day, and didn’t have enough energy for the party at City Museum on the evening of Day 1.

Then on the morning of Day 2, I ended up going to two FedEx offices, Walgreens, Chase Bank and USPS for submitting a Singapore visa application, and I missed out the talk I was looking forward to the most, Evan Czaplicki’s The Economics of Programming Languages.

Also, there were some people I wanted to meet and chat with, but the sheer number of people and things going on made it hard to find some people. That’s one of the downsides of a conference of this size.

Oh well. It’s OK.

The closing keynotes

The two closing keynotes were really the highlight of Strange Loop for me; I don’t think there could’ve been a better end to the conference.

The first one was by Julia Evans, on why certain things are hard to learn and what we can potentially do to help other people learn them better. I still remember watching one of Julia’s earlier presentations at Strange Loop online, and how it made me feel empowered despite not knowing or doing any kernel hacking.

Julia ended her keynote with one of her recent comics around how there are many different roles that people in a programming community play, and how they help other folks. I’d seen it before on Twitter, but seeing it again, at the end of her talk, served as a good reminder of “hey, you know, we’re all in this together, and we can figure things out together.”

The second closing keynote was by Randall Munroe, the creator of XKCD and author of the What If? books. Randall traced his journey from the inception of XKCD, to working at NASA, to working on XKCD full-time, shared some fun interactions with people across the world while and after working on What If, and finally sharing how his perspective on online discussions had changed over time.

I was first introduced to XKCD by an MIT admissions blog post by rfong, around 2011. It was a fun distraction at the time. Then, when I joined IIT Bombay to study physics, I found myself in like-minded company, where we would regularly and eagerly discuss the newest XKCD that had came out, and occasionally just respond with XKCD numbers as a shorthand.

I also learned at the time that Randall had started out studying physics.

During the keynote, when Randall talked about wonder, and the beauty of asking questions, and trying to answer them, I recalled the time where I got nerd-sniped into calculating orbits for a satellite around a ringworld by a question on StackExchange. I’m still a little bit embarrased to admit it, but those 2 days I spent doing those calculations and plots were some of the most fun days during my PhD.

When Randall brought up the ‘Someone is wrong on the internet’ XKCD, I remembered the time I used to be an enthusiastic youngster around programming language discussions, where I would try to show-off my knowledge about the large variety of PLs with “hey, this thing you think is impossible was actually done by this academic language.”

When Randall brought up the ‘Lucky 10000’ XKCD near the end, I remembered how 2-3 years ago I’d started trying to actively lean into the same feeling, after repeatedly seeing Julia’s various zines and tweets around her excitement to share what she’d learned with others.

Depending on the situation, Randall might be described as a cartoonist and an author, but I feel like a better description of who he is, is that he is really a philosopher, because via his comics, he has given us shared vocabulary that distill common experiences and feelings into something that can be talked about easily. There are many feelings which are hard-to-describe, but for some of them, the best description is really an XKCD.

I think, at this point, Randall’s keynote is probably my 3rd favorite Strange Loop talk of all time. Sorry Randall, for me, Mouse Reeve’s Mapping Imaginary Cities is in number 1 position, and Emery Berger’s Performance Matters is number 2.

What next

Even though there won’t be any more iterations of Strange Loop, I feel that the spirit of Strange Loop will continue to live on, be it in smaller conferences or un-conferences, or meetups or virtual spaces.

As more people continue to enjoy the archives, I hope that conference organizers and attendees use Strange Loop as an adjective that’s in the present, rather than a noun that was in the past.

Thank you Strange Loop for everything. I love you. ❤️