399
points
thomascountz
Joined 1,117 karma
thomascountz.com
My career is guided by three core principles: to create, to learn, and to share.
[ my public key: https://keybase.io/thomascountz; my proof: https://keybase.io/thomascountz/sigs/M83KX58eHE-LOM7b3BsNT98J4Fx_TAlqbdMG7AvHqdg ]
For the record, planning to do something later than originally planned is the definition of "postpone." Nevertheless, coupling to any vendor is a form of technical debt, and it's always a good idea to take stock and evaluate if it's time to start repaying it."Postponing" means they will just do it later.- AFAICT the tool routed the PCB from an existing schematic. It did not "design" the computer.
Source: https://www.quilter.ai/blog/preparing-an-ai-designed-compute...NXP publishes full schematics and CAD files for this platform, originally designed in Cadence Allegro. Our goal was to keep the schematic identical and prove out only the layout portion with Quilter. That gave us a clear baseline: if the board didn't work, it would be due to our layout.
Source: https://www.quilter.ai/project-speedrunWe chose to base our System-on-Module (SOM) + baseboard designs on the NXP i.MX 8M Mini evaluation platform Staff Electrical Engineer Ben Jordan prepared the design and constraints for the boards and submitted the jobs. Quilter ran parallel seeded runs with varied constraints, completing the layout in 27 hours, returning multiple ranked candidates. Quilter took care of the repetitive design work while the engineer stayed in control. Automation handled placement, routing, and physics checks, freeing him to focus on firmware prep, documentation, and constraint refinement. Common supply-chain hiccups—a few connectors out of stock and a Wi-Fi module dropped—were resolved instantly, with no delay to iteration. Cleanup was minimal: PDN pours, via clusters, and minor footprint swaps—no rip-ups, no re-spins.- It's important to remember there's no "right" or "wrong," it's all about connection.
If a stranger says, "my bike tire is flat," in most western cultures, they might very well be asking for your help to reinflate their tire.
If your loved one says the same, well you have a lot more context to fill in their subtext with. If they're displeased with your reasonable attempts to help them—like you'd help a stranger—it might mean that they were asking for something else. Finding out what that "something else" is, and adapting to each other's differences in "what was said" vs "what was heard," is part of what it means to build a connection with someone.
- This is really inspiring. Doing whatever you gotta do to be a better support for your loved ones is commendable.
Can you give an example of what you record in your SR system? Is it the anecdote itself? Do you generalize the pattern? Is there a "front" and "back?" A cloze?
- I think this might be what many people think. Which is what brings upon problems of self-worth. However, "best" and "high skill" aren't always the reason why companies value work and workers, i.e. the economy is not a meritocracy.
- NTSB Statement: https://www.ntsb.gov/news/Documents/National%20Defense%20Aut...
Chairman and Ranking Member of the U.S. Senate Committee on Commerce, Science and Transportation proposed amendment: https://www.congress.gov/amendment/119th-congress/senate-ame...
SA 3969. Ms. CANTWELL (for herself and Mr. Cruz) submitted an amendment intended to be proposed by her to the bill S. 1071, to require the Secretary of Veterans Affairs to disinter the remains of Fernando V. Cota from Fort Sam Houston National Cemetery, Texas, and for other purposes; which was ordered to lie on the table; as follows: Strike section 373. - The choice release a non-rechargeable/non-serviceable product feels like something that shouldn't be dismissed with "...lasts for years..." and "...you''d probably lose the charger..." This language feels patronizing to me. Even the "...[it] asks if you’d like to order another ring," begs the question: at what cost?? 99$, I presume.
The target market might not be exclusively other engineers and tinkerers, but as an engineer and tinkerer, I'm eager for more details about the testing, verification, construction, etc., of such a solution.
On the other hand, cool!
- ORGAN2/ASLSP (As Slow as Possible)[1] will turn 25 next year. What a time!
- I really really enjoyed Learn Go with Tests, and I'd say Ziglings (though it uses patched compiler errors and not test failures) is functionally similar. Albeit with the former, you also learn how to write tests and end up writing a lot more code by hand.
- I just nearly finish Ziglings and it's great! Though the skipped exercises around async is confusing as a newcomer. From the Codberg issues and PRs, it seems like perhaps an update is in the works. Otherwise, I learned a lot and would recommend it!
Great article and great ethos. Thanks for sharing! I had no idea how LLM worked before and now I know a bit more....I heard many good and bad things about [using RL for training] and I must give it a try.- I do rather think it takes good engineers to successfully build around the real constraints imposed by large companies—which optimize their products around a set of business metrics. At the end of the day, engineering is all about building around constraints; bridges aren't built in thin air with idealized traffic patterns, atomically perfect metallurgy, and with unlimited budgets. Yes, bridges aren't software services, but nevertheless the things we humans build are rarely pure or serve a singular purpose.
- Lots of debating about the color to paint the fence in the design meetings for the nuclear reactor...
I don't know how this philosophy is applied at TigerBeetle. When I establish engineering guidelines I try to frame them as exactly that: guidelines. The purpose is to spawn defensible reasoning, and to trigger reflection.
For example, I might say this:
"Style," "philosophy," "guides," they're all well-meaning and often well-informed, but you should be in command of the decision as the developer and not forget your own expertise around cohesion, cognitive load, or any functional necessities.We use a heuristic of 70 lines not as a hard limit, but as a "tripwire." If you cross it, you are not 'wrong,' but you are asked to pause and consider if you're introducing unintentional complexity. If you can justify it, keep it—there's no need to code golf.There are staunch believers in gating deploys based solely on LOCs, I'm sure... I like the idea of finding ways to transparently trigger cognitive provocations in order for everyone to steer towards better code without absolutes.
- Ruby::Box (formally: Namespace)
https://docs.ruby-lang.org/en/master/box_md.html
Ruby Box is designed to provide separated spaces in a Ruby process, to isolate applications and libraries. - Here's a fantastic writeup about frozen strings in Ruby and the upcoming changes: https://byroot.github.io/ruby/performance/2025/10/28/string-...
There is more specificity around AI use in the project README. There may have been LLMs used during drafting, which has led to the "hallmarks" sticking around that some commenters are pointing out.The book content itself is deliberately free of AI-generated prose. Drafts may start anywhere, but final text should be reviewed, edited, and owned by a human contributor.- Yes! That's the one!
I have to dig up my old code. I remember It was difficult to observe and identify the replicators. I don't remember following their "tracker" idea.
As you've mentioned, it can be quite self destructive, so I've been experimenting with the instruction set itself.
Each cell is one of 256 values, where only 10 of those values are instructions. In addition, the original instruction set is not uniformly distributed. This means that the likelihood of mutating destructively is highly affected by ratio of valid instructions and their distribution. For example, a cell with value 0xD0 is much less likely to mutate to a valid instruction than a cell of value 0x0D (assuming UTF-8). By playing with these parameters to make the state space smaller, I've seen significantly different levels of stability.
I'd love to follow your work if you share it anywhere!
[1]: https://sonic-pi.net/