one day I want to write a blog post with tips for writing explanations that respect the reader, like

- if something is weird, explain whether there's a good reason it’s weird or not
- spend a lot of time on making examples feel relevant
- if there's something that seems theoretically important but that you've never actually needed, leave it out
- if something is only useful in a very limited set of situations, tell them (or leave it out!)

(none of this is easy obviously)

I feel like half of programming is remembering how weird stuff works and the other half is setting things up so that you do not have to remember the weird stuff

been thinking about even though “the terminal” is technically not a single thing (it’s your shell, terminal emulator, OS, and programs) there are a ton of UI conventions that make the terminal feel like a unified environment, like

* full screen programs like top often use “q” to quit
* programs often turn off colour when writing to a pipe/file
* if a program is displaying a lot of text, it’ll often open “less” to let you scroll

a lot of these feel more like social conventions to me

spent some time today working on this diagram of how the ASCII control characters work in unix

there are a lot of mistakes/missing nuance but I think it's really interesting how little structure there is. Special codes (like `3` for SIGINT) that are handled by the OS are mixed with just regular keypresses (like `13` for "enter”) which are mixed with codes that are handled by the application (like `1` for Ctrl-A in readline)

(not looking for an ASCII history lesson right now)

quitting in the terminal

@simon ty, edited the post to mention that BLOB is probably a better choice!

do you know if `INTERSECT` needs to materialize the results of the two queries in memory?

I tried that too but the “INTERSECT USING TEMP B-TREE” made me think that it was actually making two separate queries, saving all the results, and then intersecting them.

sqlite's `.timer` seemed to suggest that the `INTERSECT` version was slower as well though I didn't benchmark it that carefully

@simon oh I thought you couldn't compare/create indexes on BLOBs so I didn't try

using less memory to lookup IP addresses in Mess With DNS https://jvns.ca/blog/2024/10/27/asn-ip-address-memory/

added a new TIL section to my blog https://jvns.ca/til/ with a separate RSS feed so that I have somewhere to put random little TILs, inspired by https://til.simonwillison.net/

man setting up a cron job to run `brew update --auto-update` every 2 hours has really fixed basically all of my problems with `brew install` being slow

(you can also set HOMEBREW_NO_AUTO_UPDATE but I decided I'd rather take the cron job approach -- `brew update` just updates the list of packages so it's not a big deal to run it every couple of hours)

@benbrown the "getting started" docs are already really helpful, I'm using Go so I can't use botkit directly but maybe I can try to implement a basic version of the API it describes. It looks like it would be nice to use. https://github.com/howdyai/botkit/blob/main/packages/docs/index.md#hearing-messages

@benbrown thank you!

lazyweb: trying to find a blog post with tips for how to design bots that interact over SMS

less "how to use the Twilio API" (that's easy), more "what state to store in the database, tips for designing state machines, how to keep it simple, etc”

@raven667 @drewdevault here's the "business card" version from How Git Works https://wizardzines.com/comics/inside-git/

(though it definitely doesn't quite fit on a business card)

@polotek ok I thought about this and I think (for me) I've logged into 1000 websites in my life and the login process always feel very similar to me as a user so it feels like implementing a login system should be easy somehow, like it should be a commodity

but today I am trying to implement login and it is not easy and it feels surprising because I've used so many of them (even though every time I have tried to implement login in the past it was also hard)

@simontatham i would love to know why this is, it feels like a Database Person would be able to explain why they're designed this way

some things that have made web dev simpler for me over the last ~20 years:

* new CSS features like flexbox / grid / etc
* "push to deploy" hosting like github pages, netlify, fly, RIP heroku, etc
* sqlite (I never learned how to use mysql or postgres)
* programming languages that include a production webserver like Go
* let's encrypt
* Caddy is SO much easier to set up than nginx
* Vue.js is much easier for me than jquery or vanilla js ever was
* `async` and `fetch` in javascript

(3/?)

I always hear that web dev is getting more complicated, but for me personally I feel like it mostly gets simpler over time because

1. all my projects are fun things I do on my own
2. all the old tech still exists and I can still use it
3. as new stuff comes out, I can pick and choose just only the things that will make my life simpler

(obviously I understand that “all my websites are fun 1 person projects” doesn’t apply to most “real” web developers, society exists, etc)

(2/?)

me, 22 years old: "idk web development seems easy and boring"

me, 36 years old: <trying to make a “simple" CRUD website and really Learning a Lot Of Things>

i have now been learning how to make websites recreationally on and off for 20 years and every year I have more and more respect for web developers

"How Git Works" has been out for 4 months now and it's SO nice to read all of the reviews from people who are learning from it

(also it really helps us if you write reviews! you can leave them here: https://store.wizardzines.com/products/how-git-works)