Hacker News new | past | comments | ask | show | jobs | submit | jamesnorden's comments login

It's CPU intensive JS code that must run to get an output that must match something server-side, the idea is that it makes attacks/spam not economically viable to run.


The problem is that it doesn’t do anything. Maybe you slightly slow down a volumetric spam attack, but you’re just putting a sleep() before letting spam through which might be the worst solution. As for economic viability, it’s still just a sleep(). Even if it somehow did cost extra money to use more of the CPU, botnets don’t even use their own hardware.

And if you make the PoW so hard that it takes very very long to solve then you basically made a captcha that bots have no problem doing (it’s just time) and humans don’t want to do at all especially on their phone.



BuyVM (Frantech) is a good one, you can get a lot of cheap storage "blocks" even on the lowest plan, if you wish.


Think he recently sold the company


Delete this. Tell him if they marry they can just have a shared bank account, if he insists, get a lawyer involved.


Ah yes, the classic "hacker" tip, just pay the corporation.


It's not wrong, the cap is 8, if it was max() it would use 8 on a single core system.


>In its current state it is not possible to complete the game using OpenRW, or make any significant progress through the game.

Plus last commit was 2 years ago.

why even post this?


it gratifies one's intellectual curiosity.


Is the OpenAI key paid-only?


Similarly, is there any half decent general LLM that can run on consumer hardware?


Thinking to add LLaMA or other OSS LLM models soon


Please consider LLaMA.cpp (https://github.com/ggerganov/llama.cpp), which supports a lot of models and doesn't need an expensive GPU.


Why do you believe Google "pushes" Russian software? That seems odd.


I think your definition of "not complicated at all" differs from most people.


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

Search: