alain94040

If the article title had been "why does the hash function never return -1", you would agree it's not a random behavior and serves a very clear purpose. It just so happens that the original author didn't know that until they did all the digging (nice job by the way).

Have you used the previous generation of MP3 players? I had one with a tiny LCD screen that would only fit half the song title (no space for the artist). To go to then next song, you had to press the "next" button (which makes sense). Except that action would take at least 0.5s. You press next, you wait, you see the display refresh with the next song's partial song name. Not the song I want, press next again. Very quickly, to skip 10 songs takes 10 seconds of effort. It was a painful device to use.

The iPod cam with a large screen and a click wheel. I could find songs on it. That was a revolution for me.

MP3 was the enabling technology (if you can't fit many songs on a small device, then this is moot).

Very true. To paraphrase a saying, CPU amateurs argue about micro-benchmarks on HN, the pros simulate real code.

That project seems pretty much dead, running on fumes. A message from mid-2024 after a year of silence:

>At this point all of the areas we are working on could progress much faster with money for hiring people. That was not always the case…

>Also, we could disclose a lot more if we could file patents for what we still have as internal ideas, but that requires money as well.

And in 2020:

>In some ways the plague [COVID] is much less a problem for the Mill project than for other businesses. We have always been a distributed virtual company, so we already had work-from-home worked out. And as a sweat-equity organization with a burn rate of zero we have an infinite runway, while so many others are shut down and going bust.

At some point (now), you have to give up on the hope for a white knight to show up and fund you, and just go the H&P way and publish what you are working on. No one will steal your inventions.

> if this report were to quit today, are you able to step in and complete their project?

I don't think that's the main goal for a manager (even technical). I'd say generally, the manager should understand why the team is building something, and the engineers should know how. The why includes who is going to use what was built, when do they need it, what trade-offs can be made, etc.

$20K/MO sounds exaggerated. A quick Google gives me

Palo Alto was deemed the seventh most expensive location for assisted living in California, according to a new study by Mirador, a platform that helps people research nationwide assisted living locations.

The average yearly cost of assisted living in Palo Alto is $91,177, compared to the state average of $63,927, according to the study.

Source: From https://www.paloaltoonline.com/seniors/2024/10/04/study-reve....

Sounds like the person who designed the board followed a very simple and wise rule: always connect the negative side to the ground. Can't go wrong with that...

until you have to deal with negative voltage (-5V). Another out of bounds bug.

> find me an "advanced amateur" (as the article says of the LLM's level) chess player who ever makes an illegal move

Without a board to look at, just with the same linear text input given in the prompt? I bet a lot of amateurs would not give you legal moves. No drawing or side piece of paper allowed.

You can do a search on GitHub and there is quite a bit of code that is using event.screenX > 0. Maybe someone should file some bug reports.

Also, stack overflow suggests that exact code to "differentiate between mouse and keyboard triggering onclick": https://stackoverflow.com/questions/7465006/differentiate-be...

I know several founders who went through YC in the chip design space, so even if the people running YC don't have a chip design background, just like VCs, they learn from hearing pitches of the founders who actually know the space.

I agree with most of the technical points of the article.

But there may still be value in YC calling for innovation in that space. The article is correctly showing that there is no easy win in applying LLMs to chip design. Either the market for a given application is too small, then LLMs can help but who cares, or the chip is too important, in which case you'd rather use the best engineers. Unlike software, we're not getting much of a long tail effect in chip design. Taping out a chip is just not something a hacker can do, and even playing with an FPGA has a high cost of entry compared to hacking on your PC.

But if there was an obvious path forward, YC wouldn't need to ask for an innovative approach.

Don't know what the latest is, but this from Teradyne was probably close: https://www.teradyne.com/products/ultraflexplus/

But if you assume that the opponent knows that they shouldn't pick between 37-64, doesn't that change your odds?

The game theory here is similar to another quiz "Guess 1/3 of the average".

You are really trying to guess how deep the other has thought about the problem, so you can tell which strategy they settled on, and then you adapt your strategy based on that. Of course, it's a loop.

There's a bug in your logic. Yes, if a new contract is proposed and a party declines the terms, there is no new contract. But that has no impact on the old contract.

So the old contract is still in effect, as it was not replaced by a new contract. So who terminated the old contract? That's the question here, and the conclusion seems to be that the employer is the one who then fired the employee, the employee did not quit by refusing the new contract.

Ken Kocienda is known for writing the iPhone software keyboard, an experience he wrote a book about that is well worth a read.

Accept a subset of html as a tool output? So if a CLI tool wants to output a list just do:

  printf(“<li>item 1”);

And legacy tools are obviously supported as raw text since their output doesn’t start with a magic html tag.

After 4 years of $0, I'm not sure it's a matter of finding the strength to keep going. Something in your target market doesn't seem to be working out.

If you can, find experienced friends that are willing to tell you the ugly truth about your business effort. Talk more to your potential customers. Did you just happen to pick the wrong market?

6174 is only remarkable if you count in base 10. This is HackerNews, so we all use hexadecimal. Sadly, according to https://kaprekar.sourceforge.net/output/sample_hex.php, there is not a simple Kaprekar Constant in base 16.

So you are basically at the mercy of the OS caching algorithm. That sounds like a bad plan for a benchmark. You are not measuring what you think you are (your code), you are measuring the OS caching policy.

Why isn't it linear?

Not 100% sure about cancel vs. stop.

Say I'm creating a new calendar entry. I type a lot of information, then I get to save or cancel. Technically, cancel is safe because I'm not messing up my existing calendar. But all the typing I did (especially on mobile) would be lost. So is cancel really safe? But Stop wouldn't the right choice either...

What do you suggest?

A physics textbook is a great start, but it has one major drawback: it's linear, and can't adapt to the reader. Maybe I already understand half of the textbook, but I can't figure out this one section. I'm out of luck, because the textbook won't expand on the part I struggle with.

One major value of the teacher in the classroom is to be able to sense when students are getting lost, and have ways to slow down, re-explain concepts that they missed.

A basic AI sounds like it could soon do a great job of providing the content of a physics textbook, with the adaptability of a one on one teacher.

Is your math wrong by order of magnitudes? I try to visualize the amount of waste I generate every year and try to imagine one cell phone next to it. Doesn’t feel very significant.

You are missing what happened next, which Cruise tried to cover up.

The key results:

> Cascade discovered 12 bugs that produce wrong computations under certain microarchitectural conditions (Uarchvals) in Kronos and VexRiscv

> Cascade discovered 3 bugs that cause hangs in Kro- nos, PicoRV32 and VexRiscv

These are the hard bugs to find in the implementation of the CPU

> Cascade discovered 4 bugs in BOOM and CVA6 that produce wrong output val- ues regardless of the microarchitectural state

These are unacceptable bugs, showing a lack of architectural tests. It means no one ever ran those instructions and checked the result. The community should be able to fix this.

And notice how his comment is constructive despite the lack of name calling or abusive language.

Agreed. I see the job of an experienced maintainer as a facilitator who should bend over backward to help other people's contributions land cleanly. The help often involves feedback on style and architecture consistency, but a sign of a great maintainer is someone who credits others, and mostly stays behind the curtain making sure everything goes smoothly.

Yes, not testing at all on a real device meshes nicely with their company's signature:

  Solutional is an *agile* software development company

Emphasis on agile.

There is also a dynamic aspect to that: the original team who started the startup has battle scars, they used to take risk, but now they have a natural incentive to be more conservative, preserve the way things were done.

Rands has a great article on that topic well worth reading: https://randsinrepose.com/archives/stables-and-volatiles/

It matches my experience with startups experiencing challenges as they grow.

Full array in L1 is not a typical scenario for binary search. Binary search is usually for large data sets, that are in DRAM.