*this site is a work-in-progress

2022-08-15 weekly notes: jlongster.com and Actual

tags:: weekly
date:: Aug 15th, 2022
title:: 2022-08-15 weekly notes: jlongster.com and Actual

This will be short and sweet. I’m renovating my home office and the whole room is piles of plaster board and paneling that I’ve ripped down. I’m usually able to work in the middle of it, but this morning I realized that I removed all the outlets over the weekend so there’s no power. Setting up at a coffee shop forced a late start this morning.

I really want to develop a habit of writing. In that spirit, here’s some reflections of what I did last week.


I launched this new website August 4th. Last week I cleaned up a lot of stuff and streamlined my publishing process further.

I added more styles and chose a font. I’ll probably switch the styles quite frequently to explore design.

I’m using logseq to write content. Sadly, I’ve noticed several bugs that make me lose confidence in the app. The biggest problem is I’m not confident that when I publish my content, they actually appear on the website. This is not due to bugs in my workflow; it’s bugs in how logseq persists data and inconsistencies in the data.

Luckily, they aren’t fundamental to how the app works and I’m optimistic those bugs will get fixed. I’m still getting enough value out of it, but there’s a few critical bugs that would make my publishing process a lot smoother.


Overall, last week I was more active in the community and caught up with the current status. I helped people on discord and thought about what process we need on the repos.

I was going to work on processes (giving people access to help triage issues, enabling bots to help with PR reviews, etc) but I decided the biggest thing we need to do right now is get through the backlog of PRs. My goal is to get the number of open PRs on both actual and actual-server down to low single digits in the next 2 weeks.

I will work on process and structuring work better soon, but right now it’s more important that I help get over the PR backlog.

Last week:

I also met with @trevdor on Wednesday about his work on the mobile app. Looking through the code, it’s a good approach for an initial read-only version. We talked about simplifying it further to make it merge-able; I’d like to avoid too much duplication and transient code, and we need a better plan long-term for making a single app responsive. But overall I’m bullish on his approach and after I work on the current PR queue we should be able to get it in.

This week

I stopped on the Migrate to yarn v3 PR last week. This is one of the big PRs that will need some verification and research. I briefly asked Tom about it and he said the main thing this gets us is more deterministic dependencies and better workspace support, but overall it should work the same.

I’ve ignore yarn updates for a long time because I didn’t like yarn v2, but I’ve known we’d need to address this at point. Thanks @TomAFrench for leading this work; this week I’ll pull down the PR locally and try it out. I’ll also read up on yarn v3 and make sure nothing looks scary.

Once we get this in, it’ll unlock work to migrate to webpack 5, swc, and more importantly TypeScript.

I’ll merge that PR this week and continue going through PRs, hopefully closing a good number of them.

I’m sure I’ll do other things this week; I’m letting the PRs guide my priorities so it’s hard to foresee exactly what it’ll be like. We might be able to migrate to webpack 5 and swc this week as well if that goes smoothly.

I might also think about the release process. We definitely need a more solid plan for cutting releases and communicating them, but it might not be until next week when I can work on that.