jbranchaud

This largely depends on the kind of software system you are working on. These sorts of DDL migrations are common on the kinds of Rails apps I’ve worked on over my career. Size of the table, traffic patterns, who uses associated features, tolerance for small downtime windows, or orchestrating a multi-phase zero-downtime migration are all ways to justify these migrations, and is preferable to alternatives that would be comparably over-engineered for that app’s business and technical context.

The best upgrade to these kinds of git branch scripts is piping them to fzf. I’ve been using this fbr one from the fzf repo. https://github.com/jbranchaud/dotfiles/blob/main/zshrc.local...

One way that I leave a trace, which is a practice I've adopted from Simon Willison, is to record blogmarks for interesting and useful things I come across. Sometimes I just leave a pull quote, but usually I try to add my own thoughts or make a connection to something else.

https://still.visualmode.dev/blogmarks

I don't remember my first steps with subversion, mercurial, or git too well, but I don't think any of them were any more intuitive. My understanding is that jj is supposed to be easier to use for day-to-day version control. I'm very comfortable with git, but that was hard earned and I don't see that level of confidence with most devs that I've worked with over the years. Hoping jj can be more accessible to the average working software engineer than git has been.

The core advice of the article seems relevant and useful to both of those types of design.

Great article -- I'm going set aside some time with a notebook and pencil today and give this a try.

I'm not sure I agree with the left brain/right brain, analytical and logical vs creative and emotional dichotomy. I think good design work can be just as analytical as it is creative. I think it has to be when it comes to UI and UX.

I outsource most of my esoteric `jq` syntax questions to ChatGPT. It does really well with them, usually turning up solutions that I'd struggle to munge together from several different google search results.

I wonder how ChatGPT would do with focussed pandoc requests.

Moveable looks like a super full-featured library for this kind of thing https://daybrush.com/moveable/

Maybe this: https://interactjs.io/

Hey, I'm Josh, the creator of this TIL repo. I've started it back in 2015 and still contribute to it a couple times a week. I reference TILs all the time to remember how to do something. It has been a great practice and I'd recommend to anyone to maintain their own TIL repo.

AMA!

I tend to use one tmux session per project. I usually have 2-3 open for various client projects, one open for my TIL repository, and then I go from there. Within those sessions I tend to have separate windows and panes as a way of organizing any (web) servers and other long-running processes. The 'first' window of each session is usually a split between Vim (where I'm working) and zsh (where I'm executing git commands, running tests, etc.).

> I refer back to mine a LOT. The principal audience for them is future me - if anyone else finds them useful too that's just a bonus.

100% this. Several times a week I'll be trying to do something, remember I've written about, and then find a TIL my past-self wrote that explains how to do the thing. It's really satisfying each time it happens.

I wrote about my thinking and process behind this TIL repo here: https://dev.to/jbranchaud/how-i-built-a-learning-machine-45k...

Because all these posts are 200 words or less, they tend to be the kind of thing you can read in ~90 seconds. I like sharing these smaller learnings to twitter. For a venue like HN though, I agree they don't meet quite the effort threshold.

Postgres, though not new, has some incredibly powerful features these days that make it worth investing in.

- JSONB column type, get NoSQL-esque document storage when you need it.

- Generated columns, computed from other columns without the need to maintain triggers or syncing app code

- Extensions, all kinds of OSS and SaaS-based extensions that add powerful facets to your DB

I became aware of both of these articles through Tom Critchlow's writing on Digital Gardens.

In this piece, he references the Garden and Stream concepts, while also adding in his own, Campfires. https://tomcritchlow.com/2018/10/10/of-gardens-and-wikis/

Then in this follow on piece, he touches on Stock and Flow referencing the original article. https://tomcritchlow.com/2019/02/17/building-digital-garden/

Analogous terminology for Stock and Flow are the Garden and Stream metaphors explored in Mike Caulfield's _The Garden and The Stream: A Technopastoral_. https://hapgood.us/2015/10/17/the-garden-and-the-stream-a-te...

I usually remember when I've written on something even if I've forgotten the details. I tend to be able to do a Cmd-f search of the README for a keyword to turn up what I'm looking for.

All that said, I've been contemplating moving all of this into a Gatsby-powered site with either a custom search backend or Algolia to make it all a bit more accessible.

No better time than today to start building your own learning machine

https://dev.to/jbranchaud/how-i-built-a-learning-machine-45k...

Vim is by far my most TIL-ed topic. I recently started gathering a lot of that learning into a screencast series: https://www.youtube.com/playlist?list=PL46-cKSxMYYCMpzXo6p0C...