mastodon 4.7.3Author, teacher, programmer, outdoor guy · ehmatthes.com · Python Crash Course · django-simple-deploy
Calculator manufacturers (afaik) never said "You'll never have to do math again!" They just said "You don't have to do rote calculation all the time."
Today I got another message from someone along the lines of "I built something that worked great for a while, but I can't go any further. Do I really have to learn programming to do something with this project?"
There aren't any easy answers to these questions.
It's the season for beautiful morning walks in the woods. :)
I finally got back out to Looking Glass yesterday, for the first time since January I think. It's such a grounding place to return to. :)
Happy transition back to hot cereal for breakfast season, for all those who celebrate. :)
@AlSweigart@mastodon.social 42, of course.
In all seriousness, sometimes I use a random seed that generates appealing output. Otherwise I choose a low integer that I like.
I stopped by a NAPA auto parts store this morning for wiper blades. Every auto parts store I've ever been in has had a staunch right-wing vibe.
OH: "I heard someone say the other day if Trump was a Democrat he'd still be an idiot."
OH: "They just keep blaming everything on Biden, but it ain't Biden anymore."
I don't know that any of these people would actually vote differently, but it's sure interesting to hear the vibe shift in some of these places.
When I get a PR these days, I often look at the user's profile before deciding how much time to invest in interacting with them.
I'm playing around with a tool that gives you a snapshot of information about a GitHub user, to help make these kinds of decisions. I don't want to generate any kind of score; I just want to surface the kind of information I look for in these situations.
Does a tool like this already exist?
Documentation habit: When I'm finished a PR where I've opened a bunch of tabs while working, I usually add a comment with the header "Resources" on the related issue. Then I add links to the tabs that would be most helpful if I come back to similar work in the future.
It's way more useful than just dumping tabs into a bookmarks directory.
Here's today's example: https://github.com/ehmatthes/django-simple-deploy/issues/188#issuecomment-1320806026
The code is not the valuable artifact, though.
I get the point there, but I think it's a little more nuanced. I think there's often a promise we make when building something that we'll be able to maintain it, and add on to it over time.
If an AI-assisted codebase can be maintained over time, that's great. I think we're going to see some companies fail or struggle really hard because they solved something quickly, but couldn't iterate beyond that initial success.