1vuio0pswjnm7

"your ssh key is your identity."

"don't trust us? use a throwaway key"

Text-only, no Javascript annoyances

   tnftp -4o"|grep -o '<p>.*</p>'|tr -d '\134'|sed '1s/^/<meta charset=utf-8>/'" \
   https://assets.msn.com/content/view/v2/Detail/en-in/AA25oHBP > 1.htm

   firefox ./1.htm

The submission begins with these examples:

    for f in *.txt; do
      wc -l "$f"
    done
   
      Or this?
   
    find . -name '*.sh' -exec wc -l {} +
   
      Or maybe you pipe into xargs and pray your filenames don't have spaces:
   
    find . -name '*.log' | xargs rm

Looks more like "system administration"

Bash is not the default scripting shell on Debian

Given that its manpage states that Bash is slow, perhaps it is not well-suited to be a scripting shell (cf. an interactive shell)

I always avoid Bash

I prefer Almquist (ash), eg., NetBSD sh, Debian sh with tab completion, busybox ash, etc.

Powershell must be slow and/or bloated like Bash; someone wrote xargs in C++ for Windows:

https://github.com/idigdoug/TextTools/tree/main/wargs

I have read that some HN users do not like "AI summaries"

Using a computer without a GUI, i.e., nographics, is standard practice when using networked computers. A majority of computers today are connected to networks

This nographics computer is called a "router"

This computer may only be used for limited purposes, hence the names. But its use is standard

A "router" can be used as a nographics "general purpose computer", i.e., it can be used for more than just routing packets

A "general purpose computer", graphics or nographics, can also be used as a "router", e.g., by enabling packet forwarding in the kernel

"Mobile apps are easy to make with artificial intelligence. That doesn't mean people are excited to use them."

Make apps for self not others

Actual title, currently: How Google's A.I. Search Is Imperiling the Open Web

Let's tell analysts we're developing a more efficient chip

NB. This litigation is pending, there has been no settlement

Why use the term "you"

Does it mean the HN reader

What if that reader is "non-typical"

HN comments have argued for many years that HN users are not representative of the majority of www users

Thank you to the author for questioning this ridiculous terminology

"Content" implies there is something else, besides content, that is found on the www

It's garbage, not The New Yorker, for example

When data collection, surveilllance and advertising is the sole "business model" then traffic is the business goal, not "content" (The New Yorker does not rely 100% on advertising, it also has subscription revenue)

"Content is king" but it is not an absolute requirement for traffic

That's why "Big Tech" companies are intermediaries (middlemen), they do not need to create "content"

Website operators without content have been buying traffic since the early years of the www

The www is infested with middlemen with no content

They profit from traffic

Content makes good bait for traffic but Big Tech cannot charge subscriptions for it because it isn't theirs; they have to pay for a license

May I ask what makes these experiences fascinating

Genuinely curious

"You get a Gemini summation including several links."

Who does "you" refer to

Me, I don't get a Gemini summation (Tested with old version of Chrome)

As such I do not believe that "Google search is basically Gemini now"

I believe Google search is still scanning through a doclist to find which documents, if any, contain words parsed from a query. These documents are pointed to by the URLs I get in the SERPs

I do not get any Gemini summation

The reason for this comment is that the parent cites "Raspberry Pis" instead of "the software and configuration recommended by the Raspberry Pi vendor"

The recommended software, derived from Debian,^1 and the default configuration of that software is what causes the SD card corruption, not the RPi development board

1. If I am not mistaken the OS recommendations have changed since the RPi was released, e.g., the company was originally recommending RISC OS. Challenge me on this and I'll find the reference

"Install an OS onto boot media

To use your Raspberry Pi, you must first install (flash) an operating system (OS) onto your storage device, making it a boot device. By default, Raspberry Pi devices check for an OS on any SD card inserted into the SD card slot. Some Raspberry Pi models allow you to boot an OS from other storage devices, such as USB drives, storage connected through a HAT, and network storage. The following instructions assume that you're using a microSD card.

We recommend installing an OS to your storage device using Raspberry Pi Imager on another computer. For instructions, see ~Install using Imager. If you don't have access to another computer for writing an OS image to the storage device, you might be able to install an operating system directly on your Raspberry Pi from the internet. For instructions, see ~Install over the network."

https://web.archive.org/web/20260715154847/https://www.raspb...

What is the difference between a "recommendation" and a "default"

Here, we can see that this is a recommendation not a default

The owner must make a choice

The company has not pre-selected a default choice

It has only provided recommendations

Contrast "defaults" with "recommendations"

The former is opt-out; a choice has already been pre-selected by the company

The later is opt-in; no choice has been pre-selected, a choice must be made by the owner

A few alternative sources

https://fortune.com/2026/07/19/data-center-eminent-domain-pu...

https://www.pbs.org/newshour/science/when-can-a-power-compan...

https://theaicronicle.com/en/news/policy/data-center-eminent...

Last one appears to be generated using AI, with no attribution to the original author

By "RPI owner behaviour" I mean the decision made by the RPi owner in choosing what operating system software to run and how to configure it

When I bought RPi Model B the hardware did not come with an OS pre-installed

It booted only from SD card

It was up to me to choose which OS to run. I did not choose any of the "recommended" operating systems. I was already booting other "diskless" computers using USB sticks with root mounted on tmpfs so I wanted to use that same OS for the RPi

I never experienced SD card corruption with RPi hardware

However I have experienced SD card corruption with other SBCs that had an OS pre-installed, an OS not chosen by me

By choosing to run a "recommended" OS on the RPi perhaps I would experience SD card corruption

But if that happened I would not attriubute it to the RPi. I would blame myself. And I would make different choices going forward

"Raspberry Pis are notorious for SD card corruption with unknown causes."

This is not specific to RPis. When an SD card is subjected to "unpredictable" writes it creates potential risk of corruption. The writes are triggered by software, not the RPi hardware

There is no rule that says the RPi user must mount the root filesystem r/w on the card. It can be mounted r/w on mfs or tmpfs, for example, and the card can be removed after boot. Been doing this since 2012

Zero writes to the card, no corruption

It's true these corruption issues are "notorious" but that's due to RPi owner behaviour, not the RPI hardware

CPU Intel Atom N270 1.6GHz

Memory 512MB DDR2 667MHz

Storage 160GB SATA

10.1" screen. For me, this was great in VGA textmode

I prefer to keep OS and personal data separate (This is one important difference from ChromeOS that appeared a few years later)

The USB stick contained OS only, HDD was used for data storage only, if at all. For example, I might not use the HDD at all, leaving it as is when I bought the computer. or I might use the HDD to store public data like public zone files

I primarily used external drives for personal data storage (This is another important difference from ChromeOS, no "cloud" storage)

As such, when the computer eventually dies, I have the OS on USB stick and data on external drives, easily portable to new computers

The "AI" strategy does not seem to be helping the Mets

Compare

https://github.com/ggml-org/whisper.cpp/

(Use quotes around the URLs or escape the ampersands)

LG Electronics display monitor as Trojan Horse

Except this Trojan Horse isn't offered for free

*adtech

"These 2008-2010 era netbooks are impossible to use as a desktop."

I used one as a replacement for a "desktop computer" for 7+ years

Here, "desktop" means the form factor not the interface

I used NetBSD as the OS. I never tried to use Windows. It was during this time that I stopped using X11 entirely, i.e., no "desktop" metaphor, no terminal emulator programs, and began staying in VGA textmode 100% of the time

If I needed to view graphics I sent the files to another computer running graphical OS on the LAN but not connected to the internet. At the time, this was mainly an iPad

As a matter of practice I never connected Apple computers to the internet

These ASUS netbooks indeed had a slow processor but the amount I accomplished with this computer was substantial. I created bootable USB sticks that booted to rootfs in RAM and never touched disk, an immutable, custom OS that resembles ChromeOS today, but better (no Chrome or other software written and controlled by an adtrech corporation). I could pull out the stick after boot and use the USB port for something else.^1 NetBSD kernels compiled quickly enough and I did not use QEMU for testing

1. It was perplexing to me to read about the problems people had with SD cards when the RaspberryPi appeared. I pull the SD card out after boot, the OS runs entirely from RAM

I still have this netboook. There is some issue with the power. I have thought about trying to fix it

"Blanket claims like "Everybody" are trivially falsifiable."

As are claims like "No one" or "99.9%"

How many times per week are such claims made on HN and in similar forums with respect to software and hardware

   curl https://hn.algolia.com/api/v1/search?query=no+one&tags=comment |grep -Eo '\"nbHits\":[0-9]+'

How often do HN commenters use the trivially falsifiable term "everybody"

    curl https://hn.algolia.com/api/v1/search?query=everybody&tags=comment|egrep -o '\"nbHits\":[0-9]+'

I read the article but found no hint of "aggressive-denialism that anyone could find AI genuinely useful (bordering on miraculous)"

Instead I found a statement that the author finds AI "incredibly" useful:

"An AI summary would have been incredibly useful"

Where, specifically, is the "aggressive-denialism that anyone could find AI genuinely useful (bordering on miraculous)" in this article

The author's point as I understood it is about "AI" _hype_ not "AI" usefulness. The issue she has is with the _people_ who are hyping "AI" and the actual or potential effects of that hype

Corrected URL

https://dn721508.ca.archive.org/0/items/the-myth-of-artifici...

Review from 2024

https://eprints.whiterose.ac.uk/id/eprint/216381/1/Myth%20of...