Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Fast

catherinejue.com

Post reply on HN
Print view View original

111–120 of 450 posts

Re: Fast

#111

Post by Karrot_Kream »

skydhash wrote: post #94
citizenpaul wrote: post #62

Earlier quoted context omitted.

The only thing I've found that LLM speeds up my work is a sort of advanced find replace. A prompt like " I want to make this change in the code where any logic deals with XXX. To be/do XXX instead/additionally/somelogicchange/whatever" It has been pretty decent at these types of changes and saves time of poking though and finding all the places I would have updated manually in a way that find/replace never could. Tho…

I supposed you haven’t tried emacs grep mode or vim quickfix? If the change is mechanical, you create a macro and be done in seconds. If it’s not, you still got the high level overview and quick navigation.

emacs macros aren't the same. You need to look at the file, observe a pattern, then start recording the macro and hope the pattern holds. An LLM can just do this.

Re: Fast

#113

Post by monkeydust »

Trading software by it's nature has to be fast, fast to display new information and fast to act on it per users intent.

Re: Fast

#114

Post by Eji1700 »

hn_throwaway_99 wrote: post #104

Just want to say how much I thank YCom for not f'ing up the HN interface, and keeping it fast. I distinctly remember when Slashdot committed suicide. They had an interface that was very easy for me to scan and find high value comments, and in the name of "modern UI" or some other nonsense needed to keep a few designers employed, completely revamped it so that it had a ton of whitespace and made it basically impossibl…

Information density and ease of identification is the antithesis of "engagement" which often has some time on site metric they're hunting.

If you can find what you want and read it you might not spend 5 extra seconds lost on their page and thus they can pad their stats for advertisers. Bonus points if the stupid page loads in such a way you accidentally click on something and give them a "conversion".

Sadly financial incentive is almost always towards tricking people into doing something they don't want to do instead of just actually giving them what they fucking want.

Re: Fast

#115

Post by aidenn0 »

Aurornis wrote: post #96
Liftyee wrote: post #54

I always have to remind myself of the bank transfer situation in the US whenever I read an article complaining about it. Here in the UK, bank transfers are quick and simple (the money appears to move virtually instantly). Feel free to enlighten me to why they're so slow in the US.

The US actually has two real-time payment systems: RTP and FedNow. The number of participating banks is growing rapidly. https://real-timepayments.com/Banks-Real-Time-Payments.html Prior to that, you could get instant transfers but it came with a small fee because they were routed through credit card networks. The credit card networks took a fee but credit card transactions also have different guarantees and reversib…

From the linked RTP site: "Because of development and operational costs most banks and credit unions will offer "send only" capabilities"

Which means nobody can send me money.

FedNow on the backend is supported by fewer banks than Zelle is, which is probably why hardly any banks expose a front-end for it.

Re: Fast

#116

Post by dist-epoch »

calibas wrote: post #49

Efficient code is also environmentally friendly. First, efficient code is going to use less electricity, and thus, fewer resources will need to be consumed. Second, efficient code means you don't need to be constantly upgrading your hardware.

Energy used for lighting didn't decrease when the world moved to LED lights which use much less energy - instead we just used more lighting everywhere, and now cities are white instead of yellow.

Re: Fast

#117

Post by Karrot_Kream »

PaulHoule wrote: post #2

Kinda funny but I think LLM-assisted workflows are frequently slow -- that is, if I use the "refactor" features in my IDE it is done in a second, if I ask the faster kind of assistant it comes back in 30 seconds, if I ask the "agentic" kind of assistant it comes back in 15 minutes. I asked an agent to write an http endpoint at the end of the work day when I had just 30 min left -- my first thought was "it took 10 min…

I guess it depends? The "refactor" stuff, if your IDE or language server can handle it, then yeah I find the LLM slower for sure. But there are other cases than an LLM helps a lot.

I was writing some URL canonicalization logic yesterday. Because we rolled this out as an MVP, customers put URLs in all sorts of ways and we stored it into the DB. My initial pass at the logic failed on some cases. Luckily URL canonicalization is pretty trivially testable. So I took the most used customers from our DB, send them to Claude and told Claude to come up with the "minimum spanning test cases" that cover this behavior. This took maybe 5-10 sec. I then told Zed's agent mode using Opus to make me a test file and use these test cases to call my function. I audited the test cases and ended up removing some silly ones. I iterated on my logic and that was that. Definitely faster than having to do this myself.

Re: Fast

#118

Post by nashashmi »

hn_throwaway_99 wrote: post #104

Just want to say how much I thank YCom for not f'ing up the HN interface, and keeping it fast. I distinctly remember when Slashdot committed suicide. They had an interface that was very easy for me to scan and find high value comments, and in the name of "modern UI" or some other nonsense needed to keep a few designers employed, completely revamped it so that it had a ton of whitespace and made it basically impossibl…

Is that when they went fully xhtml?

Re: Fast

#119

Post by breuleux »

calibas wrote: post #49

Efficient code is also environmentally friendly. First, efficient code is going to use less electricity, and thus, fewer resources will need to be consumed. Second, efficient code means you don't need to be constantly upgrading your hardware.

Well, that depends. Very inefficient code tends to only be used when absolutely needed. If an LLM becomes ten times faster at answering simple prompts, it may very well be used a hundred times more as a result, in which case electricity use will go up, not down. Efficiency gains commonly result in doing way more with more, not more with less.

Re: Fast

#120

Post by raldi »

chamomeal wrote: post #28

Only sorta related, but it’s crazy that to me how much our standards have dropped for speed/responsiveness in some areas. I used to play games on N64 with three friends. I didn’t even have a concept of input lag back then. Control inputs were just instantly respected by the game. Meanwhile today, if I want to play rocket league with three friends on my Xbox series S (the latest gen, but the less powerful version), I…

Or how channel surfing now requires a 1-2 second latency per channel, versus the way it was seemingly instant from the invention of television through the early 1990s.
Post reply on HN