8,155 karma · joined April 1, 2013
x should be designed not by piling feature on top of feature, but by removing the weaknesses and restrictions that make additional features appear necessary. - R^nRS
Nothing is more important than good abstractions.
Haskell is practical; it pays my bills. (https://www.obsidian.systems/ is where I work.)
> Nine coal power plants that were set for retirement last year have had their operating lives extended, including five in response to emergency orders from the Department of Energy.
Maybe the other 4 still stay open without the bullshit DoE order keeping the 5 open, but who knows.
Everything fundamental that makes good easier for humans to maintain also makes it easier for LLMs to maintain. Full stop.
CC: Claude Code
TC: total comp(ensation)
Hell no. I, a craftsman, was going out of my way to use things like Haskell. I was very aware of the divide the entire time. The present is a relief.
- give up capitalism for information, and rely on UBI and gov grants for art and media
- make the market great again with micropayments and subscriptions
Both of these have problems, but also both are better than ads, which have been an unmitigated disaster.
The US having a dogshit healthcare delivery system but so much research means that good vertical integration is not possible.
Conversely a more integrated EU — continent scale welfare state — could do really interesting "integrated OpEx and CapEx" medical research in ways that are simply impossible in the US.
Remember the Danes making Ozempic is making something that is fundamentally far more useful for Americans than Danes (of course the money is good for Danes). Most non-American drug research today probably chases the lucrative American market, but ideally that would change.
To me, "begging the question" doesn't mean assuming the conclusion in particular, it just means that some of the premises used are less obvious than they are being passed off as. Assuming the conclusion is merely an especially egregious form of that.
Multiple kernels, one filesystem is the dream. We'll restore competition between kernels by lowering switching costs to zilch, finally solving the problem in the infamous Rob Pike paper (which is that the switching costs make the path dependency of "just keep working on the same legacy-constrained OS" too powerful).
FreeBSD is just the first non-Linux kernel to be done this way --- OpenBSD has been NixOS-ified too (a bit bitrotted atm though) too. Your favorite kernel can be NixOS-ified too!
I'm dead serious, we should be in a golden age of "programming in the large" formal protocols.
These bootstraps essentially speedrun software history, and so they tell us a lot about how we got here, and why we write the things we write. But they also create perfect game to weite greenfield alternative bootstraps. The shortest, most readable bootstrap, is proof of the best abstractions, the best way of doing things.
It's a chance to finally put the sort of software stack / tech tree stuff on a more apples-to-apples basis.
I mean, my company buys them, I know at least one other that does too. Both are too technical I think to bother with a hardware support contract. But others might!
I'm not kidding :)
(Of course, they were not buying it all.)
I'm banking on a future that if users feel they can (perhaps vibe) code their own solutions, they are far less likely to open their wallets for our bloatware solutions. Why pay exorbitant rents for shitty SaaS if you can make your own thing ad-free, exactly to your own mental spec?
I want the "computers are new, programmers are in short supply, customer is desperate" era we've had in my lifetime so far to come to a close.