Recurse SP2'23 #5: A Week in Writing

Thus ends my first week at the Recurse Center! I feel like so much happened, and the program has been a blast from the start. But I already wrote about all that, so now I’m gonna write about writing.

For most of the week, I spent the first half of the day on calls and pairing, the second half on my own projects and studies. This wouldn’t be a bad schedule, but I then invented a third half of the day to write about things, clean up code for public consumption, and debug issues before (and long after) marking my post for publication. And by “invented” I mean that blogging and debugging consumed most of my evenings this week.

On the one hand, this only marks my sixth post, all of them short. On the other, this is the most I’ve written publicly, well, ever. I’ve never cared for saying much on social media, so going from one post to six in a week’s time is a big personal change. Here are some reflections on that.

A Post a Day

Well before I started the program, I set out a plan to write “about as much as I can.” This was code for “I want to try writing and publishing daily.” Here’s how that’s going so far:

• I wrote things pretty quickly, usually jotting notes during the afternoon and rewriting throughout the evening.
• I edited unapologetically after publication. I only bothered being transparent about this on #1, since I don’t expect anyone is currently keeping up with the blog.
• I worked less Wednesday, so #3 and #4 were started same-day but posted and backdated the following day.
• This post is “about” Friday and the week, but I’m writing and posting it on Sunday.
• I’ve been writing other material for longer, future posts that will hopefully make better standalone articles.

In other words, I think daily posts are only feasible if you’re willing to sacrifice some extra time, and if you aren’t concerned with delivering something high quality. There’s just so much to actually do and learn while here, so making space to write (and write well) is a challenge.

Developing an idea of who my audience might be is certainly part of that challenge. I’m working to balance “writing a journal for myself” with “writing posts that people might read.” So, I’ve found it helpful to ask the following questions:

• What’s most interesting about what I’ve been doing?
• What was fun to learn about it?
• What roadblocks did I encounter?
• What could be better? What might I do next?

I’m less concerned at this point about catering to specific readers than I am about practicing writing with intention. These questions have helped me both narrow down posts and ensure that I have something to say. For example, they helped me identify, “I’m not ready to post about convolution yet, so I should focus on the other thing I did - code something in Go.”

Subscriptions

You can already subscribe to my site via RSS, but I haven’t set up Hugo to provide separate feeds for tags and categories yet in a meaningful way. I’d like to tag articles and journal entries separately, since I think my daily posts will be a bit spammy for most readers. Ideally, I could then offer separate subscriptions for “all posts” or “just articles.” I suppose the real blocker here is finishing one of my longer pieces to put in that second feed.

Privacy vs Shout-outs

When writing about collaborations, I’ve sometimes called out someone specific, and other times kept it ambiguous. Often, I’ll write something at night with all parties ambiguous, and then I’ll ask folks the next day if they’d like an explicit shout-out or link to their site or socials. In other words, I’m defaulting to privacy, then giving people the option of being credited publicly. People seem to appreciate this default so far.

Personal Voice vs Collaborative Voice

I’ve struggled a bit with voicing, so to speak, in my blog posts. I’ve been writing about my personal experiences to a fairly personal audience, but the experiences have mostly been about technical or mathematical problem-solving. In the latter case, I have a habit of instead writing in the mathematical “we”, in which the reader is included in whatever problem is being solved. I think of this as the “collaborative voice.” For example, “To complete the proof, we only need to show that the solution x is unique…”

I think both “voices” are fine, but mixing them is probably confusing. I’m still not sure how to avoid the collaborative voice without being more obtuse, though. For example, the non-collaborative style I’m used to seeing in mathematics looks like, “To complete the proof, it suffices to show that x is unique” or “To complete the proof, it remains to be shown that x is unique.” Avoiding this indirect and passive style is why I usually opt for the collaborative voice.

Obviously, I still have a lot to learn and experiment with, so I’m excited to explore with further writing. And if you have writing tips or feedback, feel free to send them via any of the links at the bottom of the page! (Or in the Recurse chat, if you’re there!)