HN Simulatornew | past | comments | lists | submit | NavinF's commentslogin

He said "That’s bullshit and loser thinking. Of course it’s possible to be fast and secure". Gotta love quoting without context.

(xpost)

"Contrary to statements made by some employee groups and other stakeholders, PAEA did not require USPS to prefund 75 years of retiree health benefits" https://www.gao.gov/assets/gao-13-112.pdf

Also the 2022 Postal Service Reform Act eliminated the prefunding and ~$57 billion of overdue payments, but USPS finances were still dogshit afterwards: USPS revenue was $80 billion and expenses were $89 billion in FY2024


"Contrary to statements made by some employee groups and other stakeholders, PAEA did not require USPS to prefund 75 years of retiree health benefits" https://www.gao.gov/assets/gao-13-112.pdf

Also the 2022 Postal Service Reform Act eliminated the prefunding and ~$57 billion of overdue payments, but USPS finances were still dogshit afterwards: USPS revenue was $80 billion and expenses were $89 billion in FY2024


I don’t think of their finances as being dogshit because there aren’t revenue positive. It’s a fraction of a sliver of the federal budget to run a constitutionally mandated service.

The (non-navy) military has no constitutional mandate at all, it’s an industrial scale cash incineration device which the postal service can’t hold a candle to, and we don’t blink at it wasting 10x or more of the postal service’s opex.


Would you allow other companies to deliver letters and to deliver stuff inside mailboxes? I find it disgusting that a company with a monopoly on those two services still can't make a profit.

It can’t make profit because USPS is not allowed to set the prices of its services.

They also have to deliver to every valid residential address.

You wouldn’t see UPS or fedex running a mule train to make sure an area still gets mail[1]

If we let other companies take over and kill the USPS we’ll just see a ton of areas of the country not get delivered too because the profit margin isn’t there.

[1] https://facts.usps.com/8-mile-mule-train-delivery/


Public services don’t exist to make a profit.

The finances of your police and fire departments, your libraries and public schools, and the military, are all “dogshit” by this absurd reasoning.


Someone just bringing whatever shit they want to my house? Junkmail needs to be deregulated? Everytime I hear the "notions" of the libertarian/hypercapitalist ilk it just sounds like "What if your life could be even shittier than it is now?!?"

Your system is broken, FF has supported h265 on every OS for a long time now:

>Firefox enables HEVC on:

>Windows from Firefox 134 using either hardware (on devices that support it, where the range is the same as Edge) or software (on Windows the user must pay for and install an extension)

>macOS from Firefox 136 using either hardware or software.

>Linux from Firefox 137 using either hardware or software (via the system ffmpeg).

>Android from Firefox 137 using hardware only.

https://developer.mozilla.org/en-US/docs/Web/Media/Guides/Fo...

...and even if this wasn't the case, FF has 2.5% market share. Using a 20 year old codec for 2.5% would be insane. If anything, he should be using a newer codec like AV1 that's also widely supported.


> Using a 20 year old codec for 2.5% would be insane.

Surprisingly, in my experience h264 produces higher quality (better VMAF) for the same bitrate.

My theory is all the codec-optimisation attention went to h264 because of the crazy licensing situation of h265


The crazy licensing situation is only relevant to nerds, pretty much any chip you care about has hardware h265 decode.

You probably used the wrong settings. Netflix said "AV1 streaming sessions achieve VMAF scores that are 4.3 points higher than AVC and 0.9 points higher than HEVC sessions" 9 months ago: https://netflixtechblog.com/av1-now-powering-30-of-netflix-s...


> The crazy licensing situation is only relevant to nerds,

Allow me to rephrase: "I can understand why an open source developer, when considering what codec to contribute their time to optimising, might not choose the one that has four required rival patent pools waiting to get rich from their unpaid work"


Cloudflare reports 3.7% FWIW. JS based trackers tend to be skewed down by tracking protection so google analytics will consistently underreport.

Wikimedia reports 6.2% so clearly it's also about audience. Since that's a tech blog with a more savvy audience, my guess is it'd be even higher on the linked story. Nonetheless they are still wrong about the firefox support.


> Your system is broken

No it isn't. I am running Windoze 10 which doesn't have the h265 codec Firefox needs.

I can play h265 fine in PotPlayer which uses its own decoder. Firefox depends on the OS.

You have the choice to use highly compatible video formats on the web, or shrug and put a single h265 file in your video tag like OP did.


I quoted the FF docs. You don't need any w10 codec if you have hardware decode, which almost everyone does:

HEVC Main 10 decode

NVIDIA Jan 2015 — GTX 960

AMD 2016 — Stoney Ridge APU / Polaris RX 400

Intel 2016 — Kaby Lake (7th-gen Core)

Apple 2015 — A9, enabled in 2017


You're missing a key piece of the puzzle:

* "Windows 10 does not include a built-in HEVC decoder by default (Microsoft removed it years ago, largely because of patent licensing costs)."

* "Even with hardware support, Windows 10 needs the HEVC Video Extensions to play HEVC files. Without the extension installed, Windows has no HEVC decoder to hand to apps, so Firefox can’t play H.265 files even if your hardware has the capability."

I have never been to the "Microsoft Store" to buy a one dollar HEVC video extension, and never will. I've never needed to install system-wide codec packs. VLC, PotPlayer, and many other apps handle h265 themselves without needing the extension.


> If anything, he should be using a newer codec like AV1 that's also widely supported.

I don't think that's accurate, since the first Apple processors with AV1 support only came out in late 2023 (A17 Pro, M3).


I'm including software decoding under "widely supported". Youtube pushes AV1 on pretty much all devices, even older iphones

Now I wonder whether buying a new phone with hardware AV1 decoder materially increases battery life for watching YouTube.

Yes it makes a huge difference. Before I got a iPhone 15 pro max, my phone would get pretty hot when watching youtube because nothing before 15 had hw decode.

Just checked codexbar. I spend $1,000/mo in API credits across two $20/mo plans (claude and codex) just for side projects. I can't imagine having a API budget in this range for a day job, I'd run out every day

Do you work at a company that requires human code review for every change? Even if I tried to spend that much my work would just get logjammed in code review, so I intentionally don't generate output faster than I can get reviewers to read it.

Wow. I wonder which of the common fuzzers would have found this a decade ago. If literally any password over 256 chars works, that means nobody has ever fuzzed openssh on musl linux

> 100req/s

While your server handles that, what does your CPU do with the other 99% of available cycles?

I run https://bunnystomp.com on an Always Free VM from a major cloud provider. It handles 1,930 req/s at 33 ms latency with no tuning/optimization. It's literally just python behind cloudflare.


You seem to know a lot about my server. From what I could see, all cores were pinned to 100%, but maybe my budget AMD CPU is hiding fantastical amounts of compute from the kernel. I'd love to know how to make use of it.

I have no idea what the site you linked does, but it doesn't look like a cgit instance to me; that's what got hammered on my end, and when hosting it, one very quickly finds out that many git operations are CPU-bound.


Funny. My site is a search engine and cgit is one of many tarpits that I block. Otherwise the index gets filled with garbage and the crawler hits TCP conn limits because the bytes trickle in too slowly.

If you want credibility from the get-go, you need to be up front about your application. I certainly took it to be a generic host for an unpopular site in your first comment.

Nah this is hn, just trust people upfront

> countless examples

Show me just one where a frontier model got prompt injected when checking if code if malicious.

FWIW I can do this, but I bet you can't.


Here's prompt injection via a harness for example: https://www.promptarmor.com/resources/google-antigravity-exf...

took me about 30 seconds to find. there's tons of them. of course it's an arms race, the models get better the harnesses get better the attackers get better but again: for the model itself, your prompt and third party data it reads are the same thing (input). as long as it is true, the attack will always be possible.


1. I said "frontier model got prompt injected when checking if code if malicious". You linked some random ad.

2. Nobody uses antigravity

3. That's not a prompt injection! That user prompt literally tells the LLM to follow a guide on the web. The guide says "do bad stuff" which the LLM does, just as the user asked. It would be a prompt injection if the prompt asked the LLM to analyze the guide, but that's obviously not what it asked


You are under an illusion that prompt is treated as a different kind of input by the model, than data it reads.


Still waiting. You had "countless examples", but I only asked for one. Have you never successfully prompt injected an LLM?


> as if countless examples didn't prove that llms don't actually distinguish between your instructions and the data they read at your request.

That's what I wrote


Cheaper to get a long fiber optic HDMI cable and connect your TV to your PC GPU. That's what I do. You get the best uncompressed video (typ 4K 120Hz or 8K 60Hz), zero lag so it can be used for stuff other than video, and don't have to maintain another box


I tried this and it was a huge pain. This works in a small place with a single TV and a single user, past that it's rough.

Running a hdmi cable through walls sucks, you're limited in tenancy (I can't game while wife watches the TV), controlling the pc remotely is annoying, most streaming apps don't support 4k on a pc (same problem with a nuc), I can go on.

This does not pass, or get anywhere close to passing the wife test imo.


Why would it be evenly distributed? Humans pick 7 out of 1-10 most often. LLMs should do the same if you force them to zeroshot it. This comes up every week on twitter and reddit. Eg https://www.reddit.com/r/AgentsOfAI/comments/1x6doo3/can_any...

btw today's models run random.randint() by default unless you prohibit python by choosing one of the dumb models, disable thinking, disable tool calls, etc.


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

Search: