Hacker Newsnew | past | comments | ask | show | jobs | submit | userbinator's commentslogin

Yes, as well as DSL lines, 56k modems, and whatever else you can use to connect to the Internet.

The latency is what is getting me though. 0.4 round trip every time. Tested from multiple machines including a phone on LTE to get the same response time.

https://en.wikipedia.org/wiki/Anycast


It's 3 cycles for float multiplication (and 1 for shift right):

3x faster

In throughput it's even less of a difference: 2 per cycle vs 3 per cycle.

50% faster


Many people genuinely want to find a solution that is better for the children, and telling them "if you are open to age verification you are either a fascist or a moron" is not constructive.

We know they'll take a mile if you give them an inch. Ditto with "trusted" computing and the rest of that wormcan. That's why the opposition has to be absolute.


Who is this "they" you speak of?

We have age verification for all kinds of things that can harm minors. Most of them have adequate penalties for breach such that operators of said harms ensure they comply (checks for ID when selling alcohol, entry to over-18s pubs/clubs, etc.)

There's nothing sinister going on here, just attempts to prevent social/mental harm to minors.


> There's nothing sinister going on here

There absolutely is you're just not aware of it.

This whole thing is meta financially backing right wing conservative groups that want age verification because meta wants to avoid liability for the harms their platforms cause.

In addition, this is the beginning of the end of any sort of anonymity on the internet, which has disastrous consequences for politically minded individuals, minority populations, or targets of stalking. This is a privacy nightmare bring pushed through in the guise of "muh children".


Sure but that absolute opposition hasn't, as far I can tell at least, achieved an iota of success. So it's largely a self indulgent merit badge than an actual strategy.

Many of the cheap unbranded/OEM'd ones seem to just use their MAC address, which is (supposed to be) unique.

I see someone has read K&R.

Congratulations, you've made the meme recursive.

I actually find the "AI idioms" rather less grating than emoji-vomit. That said, I don't know why certain LLM output seems to be full of the latter; certainly no real human writing I've seen has that style, but perhaps it's a result of training on data that probably should've been done without.

Patrick Wardle, the guy behind Objective-See, had that style in the 2010s when I first started following his work. I actually liked it at the time.

Some keys won’t work right away – you are responsible for making this keyboard work

Do not use any beam spring keyboard in a mission-critical environment.

I can say this is the first time I've seen this in the sales description for a keyboard. Are these assembled from NOS parts?


Having dealt with this vendor in the past he uses verbiage like this to weasel out of INAD claims (eg. if something is defective he’ll just point to this and deny your return) - I wouldn’t trust him with a $400 keyboard

I’m not going to call it dumb, but I will will say… reading that bit, I don’t understand this product at all and could not be further from their target market.

I don’t understand it either and I know people who are in it. Super expensive keyboards are one of the strangest hobbies I’ve heard of!

I'd understand it more if super expensive keyboards actually improved typing speed and accuracy, but I think they just like the sound and feeling of the keys.

I have a rare board with Blue Alps switches, from when most keyboards were mechanical; it's definitely very clicky and tactile, but I rarely use it because I can type much faster on a generic low-profile one with soft and cushy rubber domes.


These are keyboards for users of LLMs. Every time their keyboard falls apart they are reminded that LLMs can make mistakes.

Woah, how is this legal?! Actually, is it?

All it means is you have to follow the instructions to set it up; it doesn't come working out of the box. Yes, it's legal to sell products that require setup and are only partially assembled out of the box. Think of a more extreme example like a backyard swing set for kids that comes all packed up in a box. The keyboard is guaranteed to work after you follow the setup, and you can send an email for help if you are stuck.

This isn't saying "you have to do some assembly" it's saying "random parts may be broken and that's your problem to solve however you can". If the "g" key is broken it might be as simple as needing to pop the key in/out or it might be as complex as a defect in the board/just plain broken key you need to go source a replacement for yourself. It's very vague because it's saying any fixes are on the buyer, not just some assembly.

If you mean to say you have to assemble but you get support/replacements if that doesn't work out then that would be a lot better to put than the current text.


Yes, that's actually step 2 in the manual - check everything and email me if something broke in shipping. I get emails when things break so people do follow the manual! The manual was significantly updated recently.

If it's in the manual that's great, the problem is with what's on the purchase page though.

Where does it say you have to live with broken keycaps? Please send me a quote and link so I can check.

That’s the crux of it, he will use what’s on the purchase page to deny your return and any credit card chargebacks.

Probably, if "for parts or not working" is a valid description for items on eBay.

The second one is absolutely trivial if you've ever read K&R (even if you're not allowed to just call strcpy()), while the fourth one is also very straightforward if you know about https://news.ycombinator.com/item?id=15266331 ; but 32 years ago, knowledge definitely did not propagate as quickly as it does today.

Additionally, I was allowed to store Color however I wanted — so if I needed some precomputation, I was allowed to bake it in there.

I believe it can be done in three operations, not including the precomputation.


> The second one is absolutely trivial if you've ever read K&R (even if you're not allowed to just call strcpy())

The naive approach’s assumes you can iterate over the first string until it terminates.

It’s a bit trickier if you do not assume the memory regions cannot overlap.

See memcpy vs memmove: https://man7.org/linux/man-pages/man3/memmove.3.html


Just tried the:

y -= x >> 4;

x += y >> 4;

Certainly works, and seems to require 100 iterations to get a full circle.

Are there other approximations, taking smaller angular steps, to get a better circle?


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: