Skipping React hydration… so, only rendering on the server? BBC just re-invented server-side rendering, bravo 👏😆

I say this as an 8-year React developer. Damn, our industry really drank the kool-aid on on this one. Of course, plenty of people have been saying that React for static content like this has always been a misapplication of the tool, I’ve been reading opinions like that the entire time I’ve been working with it.

I’m glad BBC is doing this, though. Legitimate kudos to them for recognizing the issue and working towards fixing it. I actually think there are some great benefits that React has given us:

• A universal component interface for all JavaScript-targeting languages
• An enormous ecosystem of components
• Popularization of the “component model”, which has spread to basically every other language that is used to render user interfaces (the mental model is just that much better)
• A quickly-evolving (React is arguably on its third major paradigm shift) testbed for what works best for UI development

I would be happy if React was supplanted in the near future, but I also have some fondness for it. I know I’m way off topic on this post, just felt like talking about React.

Tracking scripts degrade gracefully, so you can disable JS, and the page should still work just fine.

I used to feel this way. Over the course of building out 2 calendar systems in my career (so far) and having to learn the intricacies of date and time-related data types and how they interact with time zones, I don’t have much disdain for time zones. I’d suggest for anyone who feels the same way as this meme read So You Want To Abolish Time Zones.

Also, programmers tend to get frustrated with time zones when they run into bugs around time zone conversion. This is almost always due to the code being written in a way that disregards the existence of times zones until it’s needed and then tacks on the time zone handling as an afterthought.

If any code that deals with time takes the full complexities of time zones into account from the get-go (which isn’t that hard to do), then it’s pretty straightforward to manage.

I tried Warp back when it first came out, but haven’t really considered switching to it since then. What do you really love about it?

This is why many languages have errors and warnings as separate things. Errors for things that for sure prevent the program from working, and warnings for things that are probably wrong but don’t prevent things from working. If you have a setting to then treat warnings as errors (like for CI checks), then you get all the guarantees and none of the frustration.

This was a really good talk! I’ve been using git for about a decade, but I learned several new things. Here’s a few:

• Sorting git log by committer date
• Speeding up common operations on bigger repos with git maintenance
• More useful file blame with the -C flag on git blame

Are you familiar with the concept of “atomicity” in relation to database systems? It’s actually a very appropriate term, and the article touches on its use over “immutable”.

Would Homebrew work for this? I use it in WSL for all my CLI programs.

While I agree that a lot of the hype around AI goes overboard, you should probably read this recent paper about AI classification: https://arxiv.org/abs/2311.02462

Systems like DeepMind are narrow AI, whereas LLMs are general AI.

If anyone wants a really deep dive into the diamond industry, a recent episode of Search Engine (podcast by PJ Vogt, for anyone familiar with Reply All) is all about this: https://pjvogt.substack.com/p/why-are-we-still-buying-diamonds

It’s a fantastic episode.

It weakens it a bit, but in my opinion it still has strength where it counts. If an attacker gets access to your password outside your password manager (man-in-the-middle, keylogger, phishing), then you’re still protected. Maybe it’s hubris in my own ability to keep my password manager safe, but I’ve never been worried about storing MFA in my password manager.