Hacker Newsnew | past | comments | ask | show | jobs | submit | jml78's commentslogin

Just a high level question. My youngest son was adopted. He isn't white. He is naturalized. Should we go ahead and renew his passport so he can carry around a passport card at this point(we never got one in the past because they seemed like a waste). It seems crazy but with people being grabbed off the street for being brown, it seems prudent have a quick way to prove citizenship.

It's awful to have to consider these things. But since you ask, he also could carry his naturalization certificate. The other option is to get a Certificate of Naturalization but that's akin to getting a passport.

I’m not white (and never have been). Haven’t yet been “grabbed off the street for being brown.”

I've never been murdered either. Good to know that doesn't happen at all.

You forgot to mention it is also tied to provable namespaces. People keep saying that NPM is just the biggest target...

Hate to break it to you but from targeting enterprises, java maven artifacts would be a MASSIVE target. It is just harder to compromise because NPM is such shit.


Maven Central verifies the domain used for the package namespace, too. You need to create a DNS TXT entry with a key.

This adds a bit more overhead to typo squatting, and a paper trail, since a domain registrar can have identity/billing information subpoenaed. Versus changing a config file and running a publish command...


Maven central also requires package signing. You're not stealing my signing key. It's on a yubikey. Game over, you can't publish malware in my name using my key.

Based on sales of the air, it won’t be around in future years.


Yes. The US government requires offsite backups .

They also require routine testing distaster recovery plans.

I participated in so many different programs over the years with those tests.

Tests that would roll over to facilities across the country


Along the same lines, stop paying property taxes on it. See how much you really own it


Yep, I am 46 and never knew the name of it until AI . Never used them


Anyone working on HTML 20 years ago was very familiar with — as a layout/typography tool. I think I used them before that but learned the name from typing it so much.


Maybe the only cool thing to come out of this kooky obsession then: people are learning more about the wider world of punctuation.

I just thought it was cool when I learned that there were glyphs with names that indicated how wide they were.

And I believe the letter "x" is the standard for determining font height? Someone can correct me.


I'm 49 and they're my bread and butter. Em-dashes Army unite!!


surrounded by spaces... right?


Yeah, I do that. Maybe it's a reaction to badly kerned fonts I've encountered or maybe I just didn't notice the words were more or less joined by the em-dash. I guess I've been treating it as a long hyphen all this time.


Only after, not before.


Correct, I am 46, grew up with BBS. Early internet. I will be honest, never knew the name of em dash until it became a GPT thing.


    ... meaning that you have read some posts on this page a certain way.  (-:
    --- IM2000
     * Origin: Some WWW site named Hacker News (2:257/609.3)


# Dash Usage Guide

*Hyphen (-)* = word-joiner

*En dash (–)* = “to/between”

*Em dash (—)* = pause, punch, drama


To a point. If gpt5 takes 3 minutes to output and qwen3 does it in 10 seconds and the agent can iterate 5 times to finish before gpt5, why do I care if gpt5 one shot it and qwen took 5 iterations


It doesn’t though. Fast but dumb models don’t progressively get better with more iterations.


There are many ways to skin a cat.

Often all it takes is to reset to a checkpoint or undo and adjust the prompt a bit with additional context and even dumber models can get things right.

I've used grok code fast plenty this week alongside gpt 5 when I need to pull out the big guns and it's refreshing using a fast model for smaller changes or for tasks that are tedious but repetitive during things like refactoring.


Yes fast/dumb models are useful! But that's not what OP said - they said they can be as useful as the large models by iterating them.

Do you use them successfully in cases where you just had to re-run them 5 times to get a good answer, and was that a better experience than going straight to GPT 5?


That very much depends on the usecase

Different models for different things.

Not everyone is solving complicated things every time they hit cmd-k in Cursor or use autocomplete, and they can easily switch to a different model when working harder stuff out via longer form chat.


ChaptGPT5 takes 5 times the time to finish, and still produces garbage.


I currently use Cerebras for qwen3. One of the things I like is its speed(the TPM limit is rough). I am curious, how fast is qwen3 on your platform and what quantization are you running for your models?


I'm on plane wifi right now but I'll benchmark later today — when I benchmarked GLM-4.5, I could get 150-200tps in the Bay Area, California. Qwen3 is probably somewhat lower TBH. We have an open-source coding agent that includes a TPS benchmarker that works with any OpenAI compatible API, including ours: https://github.com/synthetic-lab/octofriend

To run the TPS benchmark, just run:

    octo bench tps
All it does is ask the model to write a long story without making tool calls (although we do send the tool definitions over, to accurately benchmark differences in tool call serialization/parsing). It usually consumes a little over 1k tokens so it's fairly cheap to run against different usage-based APIs (and only consumes a single request for subscription APIs that rate limit by request).

Edit: forgot to add — for Qwen3 everything should be running in FP8.


Just tried benchmarking from Mexico City, where I'm at for a wedding — looks like 130tps for Qwen3 Coder 480B here.


Which is crazy to me. I had purchased international airline tickets 9 months prior to COVID.

Covid happened and everything was cancelled. The airline refused to refund, only give credit. The issue is that it was on an airline that was useless to me because this trip was cancelled and we were going to be rescheduling.

Did a chargeback with Apple even though I was past the date, they still gave me my money back. I was shocked


That's just credit (and for that matter most debit) cards working as designed. A card payment is only considered final once goods or services have been delivered on the agreed-upon date. For travel, this can well be months after the actual transaction date, but doesn't change anything about your dispute rights.

Unfortunately some European banks aren't too familiar with these rules, especially when bankruptcy law is involved.


Ive had good experiences with Apple doing chargebacks although my cases were pretty open and shut. Can’t say that about other issuers though including amex surprisingly


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

Search: