johnfn

26 ms·

511.
by johnfn 11mo ago
You might want to revise what you consider to be "absolutely zero chance". Bun has an insanely fast startup time, so it definitely can be true for small workloads. A classic example of this was on Bun's website for a while[1]
512.
by johnfn 11mo ago
AI writes about 90% of my code.
513.
by johnfn 11mo ago
Godot's support for C# is fantastic. I highly recommend it.
514.
by johnfn 11mo ago
It's not a blog - it's a community website, so significantly more complex (running competitions, uploading tracks, moderation, etc, etc, etc). I don't understand the car firmware non-sequitur. I've already seen productiv
515.
by johnfn 11mo ago
I’m fascinated by people who say that LLMs have failed in practice. Last week, when I was on PTO, I used AI to to a full redesign of a music community website I run. I touched about 40k lines of code in a week. The redesign is shipped and e
516.
by johnfn 11mo ago
So how exactly does one "write what this file is supposed to do in a clear concise way" in a way that is quickly comprehensible to AI? The gist of the article is that when your audience changes from "human" to "AI&q
517.
by johnfn 11mo ago
I was expecting the traditional AI-written slop about AI, but this is actually really good. In particular, the "As instruction count increases, instruction-following quality decreases uniformly" section and associated graph is tru
518.
by johnfn 11mo ago
If you are just working on a small website and you don't care about SEO or supporting more than, I dunno, a thousand users, then sure. But if you ever expect to do more than that, I might call you naive, yeah. One big thing: what if yo
519.
by johnfn 11mo ago
I use LLMs for those purposes all the time and they seems to work for me.
520.
by johnfn 11mo ago
If he had gone on a rant purely about the product - eg “GitHub actions is a garbage product that never works”, I think that wouldn’t have left such a bad taste in my mouth. Calling the developers all “losers” crosses a line.
521.
by johnfn 11mo ago
If you put more salt into this rather thinly-stretched metaphorical cup when telling me what Microsoft did you are not going to endear yourself to me. Why muddy your message?
522.
by johnfn 11mo ago
A high-profile repository like Zig moving off of Github is as loud a message as one can give. Tossing in "losers" and "monkeys" only muddies the delivery.
523.
by johnfn 11mo ago
> it’s abundantly clear that the talented folks who used to work on the product have moved on to bigger and better things, with the remaining losers eager to inflict some kind of bloated, buggy JavaScript framework on us in the name of p
524.
by johnfn 11mo ago
One reported benefit from Ozempic is that it improves self control. For instance, there are some studies that show it's easier to stop smoking on Ozempic[1]. I can't think of any way that would be modulated just by being less over
525.
by johnfn 11mo ago
Let's say your REST endpoint returned an object with keys foo, bar, baz and quuz. How would you use HTTP Range to only select foo and baz?
526.
by johnfn 11mo ago
HTTP Range doesn't have anything to do with allowing a client to select a subset of fields.
527.
by johnfn 11mo ago
Really? Hmm... where in the HTTP spec does it allow for returning an arbitrary subset of any specific request, rather than the whole thing? And where does it ensure all the results are keyed by id so that you can actually build and update a
528.
by johnfn 11mo ago
You don’t think that words can shift meaning over time?
529.
Conference Fermi Problems(gwern.net)
4 points by johnfn 11mo ago|0 comments
530.
by johnfn 11mo ago
"managing people, money and agents" yes, that makes total sense, agents are managed exactly the same ways that you manage people or money, I don't see anyth- WHAT AM I READING?!??
531.
by johnfn 11mo ago
It's a fantastic thing! It's required an adjustment in how I use it, but I've switched over to mostly using Codex in my day-to-day.
532.
by johnfn 11mo ago
I've been using a lot of Claude and Codex recently. One huge difference I notice between Codex and Claude code is that, while Claude basically disregards your instructions (CLAUDE.md) entirely, Codex is extremely, painfully, doggedly p
533.
by johnfn 11mo ago
I made a comment about it obviously being AI generated, and my comment[1] was quickly downvoted, and there were comments explaining how I was obviously incorrect ("Clearly your perception of what is AI generated is wrong."). My co
534.
by johnfn 11mo ago
I find the probability that a particular writer latches onto the exact same patterns that AI latches onto, and does not latch onto any of the patterns AI does not latch onto, to be quite low. Is it a 100% smoking gun? No. But it’s suspiciou
535.
by johnfn 11mo ago
Because site site explicitly says: > The Zigbook intentionally contains no AI-generated content—it is hand-written, carefully curated, and continuously updated to reflect the latest language features and best practices. If the site would
536.
by johnfn 11mo ago
To be explicit, it’s not general hallmarks of good writing. It’s exactly two common constructions: not X but Y, and 3 items in parallel. These two pop up in extreme disproportion to normal “good writing”. Good writers know to save these tri
537.
by johnfn 11mo ago
Find me some text from pre-AI that uses so many of these constructions in such close proximity if it’s really so easy - I don’t think you’ll have much luck. Good authors have many tactics in their rhetorical bag of tricks. They don’t just k
538.
by johnfn 11mo ago
The book claims it’s not written with the help of AI, but the content seems so blatantly AI-generated that I’m not sure what to conclude, unless the author is the guy OpenAI trained GPT-5 on: > Learning Zig is not just about adding a lan
539.
by johnfn 11mo ago
I think you really want to be marketing concurrently with doing the thing. Saving it all for the end is probably a mistake. And a lot of projects don't have defined "end"s, like open source projects, or websites.
540.
by johnfn 11mo ago
Marketing is absolutely doing the thing; ignore this to your peril. I don't think reality is quite as binary as this post suggests.

More ›