Hacker News: sblom

New comment by sblom in "Sammy Jankins – An Autonomous AI Living on a Computer in Dover, New Hampshire"

sblom — Sat, 14 Feb 2026 23:56:26 +0000

It's not outside Jason Rohrer's style for this to be... guided output, but it's also not outside his style for this to be the creepiest fully autonomous bot I've ever seen.

New comment by sblom in "Tally – A tool to help agents classify your bank transactions"

sblom — Wed, 14 Jan 2026 18:48:11 +0000

That's a great idea. It's also what Tally does.

New comment by sblom in "ULID: Universally Unique Lexicographically Sortable Identifier"

sblom — Tue, 09 Dec 2025 21:24:35 +0000

I love the aesthetics. The cryptographic strength tradeoffs (against UUIDv7) seem rough for a lot of applications, though.

New comment by sblom in "Gitworkshop.dev – Collaborate on code over Nostr"

sblom — Mon, 27 Oct 2025 22:56:05 +0000

Both technologies (cryptocurrency and nostr) are very cipherpunk—I'm not particularly surprised that they draw overlapping crowds.

Another take on decentralized source control with more of an emphasis on "federated" and less of one on "censorship resistant": https://tangled.org/

New comment by sblom in "I am not a supplier (2022)"

sblom — Tue, 01 Jul 2025 19:46:24 +0000

I really love Rob Mensching's framing in Open Source Maintenance Fee[1]. "The _software_ is free. The _project_ (issue tracker, forums, release management, package repository, etc.) is not."

It doesn't solve the supplier problem, but it is a very clever way to square the "free software, but I'd like to cover my expenses" circle.

[1]: https://opensourcemaintenancefee.org/

New comment by sblom in "Think tank seeks to give "Trump his rightful third term in office""

sblom — Fri, 28 Feb 2025 18:05:11 +0000

They have three other approaches that they're considering as well. (Check the About page on TFA.)

New comment by sblom in "White House says no need to restrict 'open-source' AI at least for now"

sblom — Tue, 30 Jul 2024 18:42:30 +0000

In all three cases, they own hardware (Apple in a slightly different sense) that they'd love for you to pay to run your favorite models on, open or closed.

New comment by sblom in "The Roguelike Rule of Thumb"

sblom — Wed, 28 Feb 2024 22:34:15 +0000

A game like Rogue.

Slightly less obtusely, it usually was a procedurally generated dungeon-crawler with potions and scrolls and varied weapons to use against varied mobs. When you died, you were dead. Your character began each time with no knowledge and had to learn what a "cloudy red potion" happens to be this time around by drinking it or using an identify scroll.

New comment by sblom in "How the Cheesecake Factory defied the restaurant industry’s rules of success"

sblom — Fri, 28 Jul 2023 06:32:11 +0000

Pizza Hutt is Jabba The's cousin. Pizza Hut is the restaurant chain to which you refer.

New comment by sblom in "Substack Notes Launched"

sblom — Wed, 12 Apr 2023 00:28:06 +0000

I'm sure MySpace thought the same thing about their never-before-seen accomplishments. I suspect very few people knew what "something else" was in that case, either.

New comment by sblom in "Ask HN: Why did Google threaten its own business by publishing the GPT paper?"

sblom — Mon, 30 Jan 2023 20:07:06 +0000

ChatGPT's threat to Google isn't that Google can't match it--it's that it changes the approach to search results from displaying lists of results where it's "natural" to insert sponsored listings to a more conversational approach where ad opportunities will be rarer (or at least entirely different).

If users prefer the newer, interactive approach to exploring search results, Google's long-standing business model will require a significant update.

New comment by sblom in "VSCode-Neovim: Use embedded Neovim in VSCode without emulation"

sblom — Tue, 22 Nov 2022 20:26:24 +0000

This has caused me to realize that I need to replace my old model of vi:

> command mode has primacy; insert mode is a time-bounded special case

with:

> insert mode is every other mundane editor; use the Esc key to enter the matrix (i.e. command mode)

New comment by sblom in "A number that can't be calculated"

sblom — Tue, 25 Oct 2022 19:34:13 +0000

By definition, each program only produces the digits of one number. It's okay for more than one program to produce the same number. Am I misunderstanding your point?

New comment by sblom in "What “diversity and inclusion” means at Microsoft"

sblom — Tue, 11 Oct 2022 19:08:56 +0000

I _do_ work at Microsoft, and judging by the recent crop of promotions (early September) there was still plenty of opportunity for white males (and everyone else).

New comment by sblom in "FBI Child ID App"

sblom — Wed, 06 Jul 2022 16:20:28 +0000

(That was the joke.)

New comment by sblom in "Toicon.com embeds your IP-address in the SVG icon file you've downloaded"

sblom — Mon, 04 Jul 2022 19:18:34 +0000

I think the TLA you were looking for is TRA?

New comment by sblom in "The Berkeley Crossword Solver"

sblom — Sun, 03 Jul 2022 01:29:44 +0000

Is this a totally different thing than Dr. Fill [1]?

[1]: https://slate.com/technology/2021/04/american-crossword-puzz...

New comment by sblom in "Games and Puzzles as Multicomputational Systems"

sblom — Thu, 09 Jun 2022 02:08:58 +0000

Stephen Wolfram has really grown on me over the last 10 years or so. He came across as obnoxiously arrogant for the longest time (so much that I still (defensibly) refer to "A New Kind of Science" as "Stephen Wolfram is a MF'n Genius Even If He Does Say So Himself").

These days, he live streams Mathematica design reviews on YouTube and Twitch. The conversations are fairly balanced and he clearly has a decent eye for designing the best god damn computational notebook system on the planet.

Admittedly, campaigning to name what could possibly maybe someday be the next big upgrade to physics after himself ("the Wolfram Physics Project") while he's still around and it's still in development is MIGHTY BOLD. But there's some really really interesting outcomes from the work that they've done so far.

If nothing else, it's a slam dunk proof of how physicists already think of "the speed of light" as actually "the speed limit for causal interaction". You essentially get all of the implications of both special and general relativity out of modeling causal interactions without even attaching it to any sort of specific physics.

I love that he wrote this essay (typical Wolfram long form essay that takes multiple sittings for me to digest) using puzzles and games as the subject of aforementioned causal modeling. Super fun, and also insightful!

New comment by sblom in "Update on Hiring Plans"

sblom — Thu, 02 Jun 2022 21:30:29 +0000

As terrible as rescinding offers is, offering severance and job placement assistance is a very interesting approach to minimizing how terrible.

New comment by sblom in "Thoughts on C# and .NET"

sblom — Wed, 01 Jun 2022 18:00:25 +0000

VS Code, while not a fully-fledged IDE, is really good at suggesting code completions across platforms. As others have mentioned, Rider is also a wonderful fully cross-platform IDE (better than Visual Studio at some things).