dwheeler

Wait a sec. My understanding is that passkeys are public/private keypairs. Public is in server, so even breaking into a server does not reveal the private key.

The big risk with passkeys is storing the passkey where it will be held hostage. Don't store it in most single-ecosystem devices like Apple. Bitwarden can export, and I think 1Password can as well.

Impressive, but no castling or en passent, so it's not really chess.

Should be, but no.

Thanks! I strongly suggest copy-pasting that explanation to your web page, that's a nice summary.

I'm so glad I left LastPass a long time ago. There are better alternatives.

Perhaps, but I think volcanic eruption followed by system collapse is very compelling. Here is the story I find most convincing from the experts whose works I have read.

It likely started with a volcanic eruption, leading to widespread famine. Those in western Europe who didn't want to starve migrated en masse, as whole families, becoming the sea peoples. The powerful empires struggled to feed their people, and many were destroyed by the forced migration from the sea peoples. Egypt barely survived, but only as a shadow of itself. Many of the others were destroyed by those who had survived on marginal lands and didn't need complex societies to keep themselves fed.

Iron can't be the cause, as iron weapons pre-existed the Bronze Age collapse. I think the evidence is stronger that the collapse forced widespread adoption. The collapse devastated long-distance trade networks, which cut off the supplies of tin needed to make bronze. The scarcity pushed people to rapidly improve iron smelting.

I'm not a professional historian, but I do find the topic interesting. We should try to learn from past disasters to prevent repetition.

See Eric H. Cline's "1177 B.C.: The Year Civilization Collapsed";

Epimethius video "What was life like after the bronze age collapse (extended version)" https://www.youtube.com/watch?v=uM6JSS3l-IQ

Horrible. Brilliant. Love it.

Thanks for the credit, I appreciate it! Glad it was helpful.

I wouldn't call it a "blog post" though, it's simply an essay. I update my essay whenever I decide it needs updating, and "blog posts" are usually chronological and fixed in time once posted.

Take care!

Those interested in this might also like my "Project Hail Mary Stellar Map" at:

https://dwheeler.com/essays/project-hail-mary-map.html

My page shows a stellar map (from the viewpoint of Polaris) and a sky chart showing the key stars from Earth's viewpoint.

Source code (MIT license for my stuff) is at: https://github.com/david-a-wheeler/plot-stars

Gnome Console seems to be intended for people who don't use terminals. I quickly install GNOME terminal for real use.

No, there is no reason to believe it can be enforced.

In the US, training is considered fair use at this time, and thus DOES NOT require a license at all. That might change, but I doubt it.

I would say "limited and unreliable access". What it says is the cause might be the cause, but it's not on any way certain.

I think Inform 7 is pretty cool. I suggest checking in to that as well.

No.

And training is currently considered fair use in the US (some court cases pending).

I am not a lawyer, tho.

I suspect this is really a surveillance bill, but we won't know until the text is revealed.

I suggest this vocal performance: https://youtube.com/watch?v=GggK9SjJpuQ

I prefer the term "assistant". It can do some tasks, but today's AI often needs human guidance for good results.

I also made a list of tips on writing code with AI, with a special focus on security. Others may find the tips useful. Here they are: https://openssf.org/blog/2026/01/05/ai-software-development-...

This has many similarities to the Heartbleed vulnerability: it involves trusting lengths from an attacker, leading to unauthorized revelation of data.

Many people use Octave https://octave.org/ which is compatible (generally) with Matlab, supports this simple syntax, and is open source software. Indeed, I've taken at least one class where the instructor asked people use Octave for these kinds of calculations.

That's only true if future improvements are easy to create as past ones, that customers care as much about those improvements, and there are no other differentiators.

For example, many companies do well by selling a less capable but more affordable and available product.

I love having built-in local natural language translation implemented by AI, which Firefox provides. Local models have different properties than remote properties, and natural language translation is a useful thing. AI should be added when it solves a real need, and the risks can be minimized (or at least controlled). The goal shouldn't be to use AI, the goal should be to solve problems for humans.

The Linux Foundation's Open Source Security Foundation (OpenSSF) has released a free online course "Secure AI/ML-Driven Software Development (LFEL1012)". It discusses protecting your software development environment, creating more secure software, and reviewing changes.

Yes, you need training if you want something good instead of slop. For example, when asked to write functions that can be secure or insecure, 45% of the time they'll do it the insecure way, and this has been stable for years. We in the OpenSSF are going to release a free course "Secure AI/ML-Driven Software Development (LFEL1012)". Expected release date is October 16. It will be here: https://training.linuxfoundation.org/express-learning/secure...

Fill in this form to receive an email notification when the course is available: https://docs.google.com/forms/d/e/1FAIpQLSfWW8M6PwOM62VHgc-Y...

Summarizes what's happened in the Open Source Security Foundation (OpenSSF) since its founding five years ago.

Using AI assistants != Vibe Coding.

AI can be a helpful assistant but they are nowhere near ready for letting loose when the results matter.

Citation needed.

I'm not sure what you mean by "Unix specification". But if you mean the international standard POSIX, yes, people care. Red Hat routinely participates in POSIX spec revision.

There are a very few deviations where you have to enable "POSIXLY_CORRECT". If that's what you mean, then you can turn that on. But in every area that matters, Linux distros implement the POSIX spec by default, and you can even turn on the POSIXLY_CORRECT mode to exactly follow it. They extend beyond it, but that is allowed and expected.

The people who build the tools in Linux distros care a lot. I know the implementors of dash and GNU make routinely refer to POSIX. The Linux distros don't have to as much with POSIX because that is generally a conpleted work and it's the maintainers of the tools who must address the updates to POSIX.

Action! is a cool language.

However, by design it did not support recursion, directly or indirectly: https://atariwiki.org/wiki/Wiki.jsp?page=Action Variables were assigned constant locations, which made things easier for the cpu, but that's a harsh limitation.

It was only available on Atari and for a while you had to have a cartridge to run its programs. For many, these were deal-breakers, as you couldn't release your programs to many.

As always, trade-offs, but pretty big ones.

Phones should include an AM/FM tuner. Then you can hear local stations, and if you're out of cell range or things go badly, receive broadcasts.

I don't like the phrase "real coder". It's not clear what it means.

I really like the word "assistant" for what we have today. The AI code assistant tools available today, like Claude Code and GitHub Copilot, can't replace humans in doing software development. Not even close. But they are often useful to human developers, and today, that's the more important measure.

I've been spending time with various AI tools, especially Claude Code and GitHub Copilot. They're amazing one minute, and they make bone-headedly bad recommendations the next. It takes effort to learn how to create good prompts, and if you want the results to be good, you have to review and critique the results. I'm particularly concerned about security. They're definitely happy to write insecure code. If you know what you're doing, prompt them well, and review their results, you can get good results.

I don't know if they'll ever reach "full autonomy". They don't need to get there to be useful.