err4nt

Epicycles weren't untrue, they're just an unnecessarily complex model that can be used to describe what picking a different point of reference can yield as a much simpler model. Not all heliocentric models got everything right just because they picked a simpler reference point for producing a model. Judge a model based on its accuracy, not its point of reference.

Call them "Thunderwear"

I saw a great re-use of IKEA parts to build something new and spectacular when I visited an office where they desks made from table legs from IKEA, plus old wooden doors with a lot of character they found at a a home building materials salvage store, they had put a thick coat of finish on the doors and then put a sheet of glass on top of that. And the office was filled with these door-desks, each a little different yet they looked so good together. They really left an impact on me, I've been thinking about them for the last 15 years whenever someone talks about tables or desks!

In my own life right now I'm in the market for a particular kind of wooden cabinet - I need a large space with doors up top, and I don't care whether the bottom has doors or shelves. I can only find the opposite, either bookshelves with no doors, or doors top and bottom, or bottom-only doors with shelves on top. So it seems I'll have to modify something to get doors on top and shelves on the bottom.

This doesn't resonate with me at all. I tell Claude in new projects: Create a todo.md file, check for open todo items there. If we work on them, check them off and leave a note with our decision and enough context for you and me to remember what we did later. After you've done this for dozens of projects, I decided to put it in my claude.md and now I say "Make a todo file and let's use it how I like".

Or I want local git for being able to roll back edits, so I say in each new folder: Create a git repository right away. Commit often. After every round. Do not ask first, just commit all the time. And if you see me manually add or edit files, add any new files and changes and commit those too. Commit liberally.

And so I can just say: "Start using Git how I want".

Or when building iOS apps I like to do things from the API where possible instead of pushing buttons in Xcode. And I want a single CLI command that will beam a new build of the iOS app to my phone. I have preferences about icons, about pricing, about international availability, etc. So now I can say "Build the iOS app in the way I like to work".

None of these are laws written in anger, or grievances, or souvenirs of the worst 30 seconds of our 'relationship', they're just the things I keep finding myself repeating.

Maybe the problem isn't Claude.md but the way you self-control your own emotions while using it.

As a web developer who has just begun building native apps: running outside of the various sandboxes browser based apps are quarantined in. Deeper file system access, better access to hardware and sensors, notifications where helpful, deeper integration into the operating system (open with, file previews, widgets to launch the app, etc) and performance. Some things the web does well, some things it does less well than native. With a native wrapper around a web app you can leverage the best of both!

Anecdotally, I've had an iPhone for 15 years, and that whole time I've dreamed of apps - some silly, some useful, that I would tell people "Imagine if there was an app that did x…". In those 15 years I learned to program and built websites and web apps, and built very basic web apps to use on my phone but shipped zero iOS apps.

Thanks to Claude, as a hobby, I've built and shipped 4 iOS apps in the last month. None of them are world-changing, but that's 4 in the last month thanks to Claude and zero in the 15 years before it. I'm clearly getting my subscription's worth out of Claude and squeezing more juice out of my own devices. It's a fun hobby! (And I'm sure if I used Android I could do more and customize my own device even more!)

Personally, I've cracked very few programming books and I've been building websites since ~2004 back when magazines and books were the best sources for information. Some people learn skills from books well, and some people end up bewildered and confused until they can actually try it out and see how it works. I need to explore to discover and learn.

I did find "Eloquent JavaScript" very helpful, it didn't just get me started on JavaScript but actually introduced that whole computer-sciencey side of programming that I hadn't encountered in my various markup and stylesheet language coding. I never finished Eloquent JavaScript though, it was like a springboard that got me into JavaScript (which is dynamic and has an interactive runtime) and as soon as I could be having a conversation with JavaScript at runtime, from there I continued learning by talking to the REPL and seeing what happened.

I've always been told by my parents that people can still hear and understand even if they appear unconscious or non-responsive, and so I have a personal policy of speaking respectfully toward them just the same as if they were alert and responding and part of the conversation, but it has shocked and amazed me (and saddened me) by what some people say in the presence of a living person who can very possibly hear and understand but may not be able to respond. Be careful not to be cruel to people in that situation!

Seen everything on Youtube, I've ditched Youtube for Spotify/listening to music now (and I don't use AI for much besides food recipe questions)

The moment the amount of savings surpassed the annual salary of a good programmer you know you made the wrong investment.

They'll shout this message right until the day they change course. Then who will hold them to account?

When it says "arrays, which may be thought of as functions whose domains are isomorphic to contiguous subsets of the integers", is it saying that this:

    const list = ['a', 'b', 'c']

is syntactic sugar for expressing something like this:

    function list(index) {
      switch (index) {
        case 0: return 'a'
        case 1: return 'b'
        case 2: return 'c'
      }
    }

Goodbye to our thought-provoking jester. There will never be another quite like him.

Merry Christmas & Happy Holidays to all from Toronto!

I copy/pasted it into my editor to see it larger. I still used my brain. This 'test' is not a true test but a trick, and the trick doesn't have to do with programming ability.

So cool! Thanks for sharing. It reminds me of one of those very old cameras with the bellows or accordion. I wish I could look through it myself to see what you see with it!

This leaves a bad taste in my mouth. I see that the man was Israeli in the original story and I don't want to presume a religious perspective, but I can share some thoughts from my own based on his story and yours. I've read the New Testament within a Jewish framework and one of the things it says, Rabbi Shaul says in 1 Corinthians 12:23 that those people in the community who are most embarrassing or cause us to blush, like the parts of our own body who are honoured or dignified by being clothed with underwear, likewise in the community are owed a special covering and to be afforded dignity by the other parts of the same body/community. Just something to think about in light of this story!

Back about 20 years ago there was a lightweight Gecko-based web browser called 'Camino' that was a delight compared to Firefox or other 'full' browsers. It was fast, simplified, and still rendered pages well. I didn't always use it, but it was a joy to use.

I downloaded Helium and gave it a spin and my first impression was "Ah, it's like Camino all over again". I can't use this for my job where I need to use stock {Chrome, Safari, Firefox, Edge} because our customers do - but for personal browsing I'm currently using Arc which is very comfortable but heavy, I might try to use Helium for quick simple things when I want less overhead! Thanks for sharing!

The PATH network is great, especially in Toronto's freezing cold and windy winters! The beautiful 'underpass' pictured in this article, with the white marble, is on my commute to work, and it really is breathtaking when you turn the corner.

In the Financial District, the various bank towers can be told apart by the colour of the marble and other stones they build with. For an underground walkway, some parts of it are really beautiful, other parts are just what you'd expect for an underground passage in a big city (especially those parts connected to the subway transit system).

    “A webpage that counts down to a movie release.”
    Not an app. Not a startup. Just a janky react/tailwind page.

Maybe part of the problem is trying to bite off more than you can chew, spinning up React and Tailwind for what could be a single HTML element, a CSS style rule or two, and a few lines of plain JavaScript code is a symptom here. This is like taking a cruise ship on a canoe trip and then complaining that it's too time consuming and difficult. Start simple, start small!

I think the key is that dotfiles are a different genre of (code) writing than production code, with different investment, different motivations, different pain points and histories, and a sensitivity to the author that's not required when analyzing production code. You're looking into someone's daily writings, not their polished releases.

I think the fear is scrutiny, rejection, mockery for something that clearly works for you and you don't ever expect anyone else to use. But also partly that it's exposure without much reward in return. All these feelings are normal and it's fine to share or not share them. Just please honour the authors of the dotfiles you read even if you wouldn't ever think to use code in the way they do!

Counterpoint: External CSS stylesheets can be cached, saving future requests, but inlined CSS must be re-downloaded each time and you lose all the benefit of caching the styles independently from the document. Depending on the size of the CSS, there's a tradeoff.

Anybody remember Microsoft's Tay AI from 9 years ago? https://en.wikipedia.org/wiki/Tay_(chatbot)

If history repeats itself, maybe with software we can automate that whole process…

How does it know the difference?

Thanks for sharing, I'll try some of these out!

You can find their criteria for what they mean by ban on the FAQ page: https://pen.org/book-bans-frequently-asked-questions/

It's probably too long to quote here, but it's a very unique definition that seems worth taking into consideration when you read their other material.

Instead of asking me, a senior developer, how to solve a problem so I could quickly and easily explain the best approach, instead they go to ChatGPT, ask it, and then come to me to ask me to audit and prove whether ChatGPT's response was valuable or not. Often it's not, but every time it's a big time waster since the right thing to do was easily communicated without needing to vet a long generated message with probably lots of flaws or places where the details simply don't make sense.

A few years ago I ran through all the realistic emergency scenarios I could think of, living in a large city and driving in a car, and except for "being attacked with a gun" the emergency tool that seemed to be useful in basically every other scenario from a building on fire to stuck in an elevator to being in or coming across a car accident was a tomahawk or an axe with a spike/pick on the other end. They also tend to be cheaper than large axes and so it's easy to toss in the car or leave near the door, etc. It may not be the single best tool for some situations, but it seemed like the best all-around tool for almost every situation.

Does he still have a plan "to provide high-speed internet to 15,000 homes and businesses in rural, remote, northern communities by June 2025" or did that get sacrificed for the principle too?

Personally I don't think it's about the money, I think it's about what they actually want to do. I think they like their plan and are happy with it.