timr

13 ms·

271.
by timr 5mo ago
Yeah, I have those, but it's still pretty hit and miss, and obviously, it ends up being a game of whack-a-mole for everything I find. I don't mean to over-state the importance of these little errors, just to say that agents do pl
272.
by timr 5mo ago
In this case, the visual display was fine -- I was instructing it to fix bad code from a previous round that happened to deliver the right results. Like I said, this is just an example that happens to be CSS. I see this stuff daily, if no
273.
by timr 5mo ago
While reading this thread, I literally just caught an agent putting in the following CSS selector in a rule: > .row > div > div, .alert This is fairly simple CSS, not multi-threaded systems development. A bar low enough that you co
274.
by timr 5mo ago
> Its not so much about same file, as reducing distance to understanding, whether visually or by some sort of easily traceable path. The metaphor remains valid. You can do this all the way down the abstraction stack, back to functions.
275.
by timr 5mo ago
There’s absolutely no tension between locality of behavior and separation of concerns in CSS: you’re putting styles on the elements in the document . The styles are defined elsewhere. It’s like arguing that all of your source code should g
276.
by timr 5mo ago
In any real application you will have far fewer semantic class names than combinations of style properties. Working with concepts is vastly easier than trying to remember the specific property differences between concept A and concept B. Ob
277.
by timr 5mo ago
> That brings with it the problem of naming a thousand things in a consistent way that everyone on your team needs to understand and remember, otherwise you end up with tons of duplicated classes, parallel systems, and bike shedding. Hav
278.
by timr 5mo ago
> Well, instead of repeating inline class names everywhere, you end up with CSS properties repeated everywhere. Not really seeing the difference. Erm... what now? That's so off-the-wall that I can't even wrap my head around yo
279.
by timr 5mo ago
> A Figma component has a certain set of styles, you apply those same styles to the corresponding React component. This is what CSS classes were made for . Of all of the arguments in favor of Tailwind, this is the one that drives me bat
280.
by timr 5mo ago
Yeah, that's a straw-man. I've worked on large-scale projects with many team members, and it's perfectly possible to use CSS as it was designed. But sure, like most tools, it starts with understanding how it works.
281.
by timr 5mo ago
It's worse than that; the common arguments for Tailwind literally derive from total ignorance of how CSS is made to work, and a disposal of guidelines that developers would worship in any other context (i.e. Don't Repeat Yourself)
282.
by timr 5mo ago
So? You’re trying to engage in tu quoque without saying it explicitly. If you think the argument is wrong, make a counter-argument. Don’t just say that the arguer hangs out with people you don’t like. Cliff in an expert, he worked in the
283.
by timr 5mo ago
...and AWS does it too. I can go right into an account and see an estimated cost per hour, and even pre-pay at a fixed discount for longer terms if I want to. They tell me right there what it will cost. They do this for everything that is r
284.
by timr 5mo ago
> OK, now do that for 369,293 stars, 76,193 forks, 138 releases and 2,133 contributors. You're counting forks and stars as code metrics now? Oy. Look, those aren't nothing -- they're a decent enough proxy for populari
285.
by timr 5mo ago
Also, you step in poop on the way to school.
286.
by timr 5mo ago
No it isn't. There's basically no upper bound on the number of commits an LLM can generate. If the LLM takes 10,000 commits to do what a human would do in 10, then the comparison is meaningless. I don't know anything about th
287.
by timr 5mo ago
Seriously? Commit count is right up there with lines of code as a classically dumb measurement of productivity.
288.
by timr 6mo ago
> The main controversy was indirect, e. g. several actors - including Craig - trying to patent ESTs. You're under-selling it. Celera filed thousands of patents on expressed sequence tags, long before anyone knew anything about them.
289.
by timr 6mo ago
He wasn't from the human genome project. He (in)famously led a competing company (Celera Genomics) that was trying to use shotgun sequencing to do the same thing as the official project, but "faster". It was a fairly big co
290.
by timr 6mo ago
Sure, but it doesn’t matter. Confounding means that there’s an uncontrollable factor in the analysis. You can’t just assume that the uncontrollable factor is the same over all measurements in the dataset. Just to illustrate the point, th
291.
by timr 6mo ago
> A 1,200 year time series.. that's definitely in the climate area. Superficially, sure, but this is not a good dataset for any climate related argument. Cherry trees live about 100 years under optimal conditions, so you’re talkin
292.
by timr 6mo ago
You really think the owner of the marketplace doesn’t have an incentive to convince you to make a sizable transaction?
293.
by timr 6mo ago
Yeah, but you have to scale the projections for uncertainty about the future, and exaggeration by the seller. In particular, if someone on the internet tells me they’re making $x a month from spammy ads on a squatted domain, I immediately d
294.
by timr 6mo ago
> Claude has no taste and its opinions are mostly those of the most prolific bloggers. I often think that LLMs are like a reddit that can talk. The more I use them, the more I find this impression to be true - they have encyclopedic know
295.
by timr 6mo ago
Agreed on both points. We’re dealing with a cost/benefit analysis, and to this point, coders have been subsidized, coerced…maybe even mandated into using the most expensive option as if it was a limitless resource. Clearly not true,
296.
by timr 6mo ago
I don’t use Haiku for planning of big tasks, so we basically agree on that. But even just Sonnet 4.6, on a fairly large codebase, only truly goes into the weeds maybe 10% of the time for me. I also write pretty specific initial prompts, and
297.
by timr 6mo ago
More information required. What are you working on? What languages? How do you define “small tasks”? What are “well-defined boundaries”? What is your workflow? Most importantly, define your acceptance criteria. What do you mean by “disappoi
298.
by timr 6mo ago
It’s not snarky. It’s literally the argument people are making: I am special, my use case is exceptional, therefore I need to use the special tool, even if you don’t need to.
299.
by timr 6mo ago
It’s deeply ironic that the folks who want to outsource as much thought to the model as possible are saying that my stance - use your brain to decide the right tool for the job - is tantamount to “vibes”.
300.
by timr 6mo ago
That’s certainly an opinion. Not one I agree with, but sure, if you entirely outsource all of your thinking to the magic box, then you probably want the box to have the strongest possible magic.

More ›