revicon

Edit: This video was from a while back but I hadn't seen it before.

More info: https://www.eso.org/public/videos/eso1825e/

This show's use of music really made it magical

One of my favorite scenes: https://vimeo.com/151017533

Bookmark this page...

https://www.facebook.com/?filter=friends&sk=h_chr

its a filter to just show you posts from your friends, no groups, nothing else.

This is why I put the raw url to the script first in my comment. Downloading the script file, doing a chmod +x and then a ./script.sh to execute it is daunting for some.

But I'll add a caveat to my original comment as well.

edit: Looks like I can't edit my original comment anymore.

Here's a one-shot script that does the compare for you, in case it's helpful...

  https://www.sfbaylabs.org/files2/2026-02-11/bad_browser_extension_check_osx.sh

You can run it directly if you cut/paste this in your mac terminal...

  curl -fsSL https://www.sfbaylabs.org/files2/2026-02-11/bad_browser_extension_check_osx.sh | bash

If you're on a mac, you can list all the IDs of your installed browser extensions across all your profiles like this...

  find "$HOME/Library/Application Support/Google/Chrome" \
    -type d -path "*/Extensions/*" -not -path "*/Extensions/*/*" \
    -print 2>/dev/null | sed 's#.*/Extensions/##' | sort -u

Compare to the list of bad extensions. I stuck a stripped down list here...

  https://www.sfbaylabs.org/files2/2026-02-11/chrome_extensions_exfiltrating_history.txt

It seems there are a bots on here commenting and upvoting each other. If HN is susceptible to this, it feels like there's no chance for the rest of the web. Damn.

It is amazing how much they’re gaming the twitter algorithm, everything in my feed is claw/molt/whatever for the last week.

It’s a masterclass in spammy marketing, I wonder if it’s actually converting into actual users.

“the proposed law will also make sharing and possession of 3D files containing guns or gun components illegal”

The statistics on consumers evaluating the purchase of something that is $9.99 vs $10 is well proven.

Switching to round number prices would cost retailers a whole lot more.

https://www.sciencedirect.com/science/article/pii/S002243599...

https://www.researchgate.net/publication/23547242_Penny_Wise...

https://www.sciencedirect.com/science/article/pii/S002243590...

There are several US states where, by law, retailers are not allowed to give preferential treatment to credit card paying customers over cash paying ones. Which means, in those states, retailers will be required to always round transactions to the cash paying customer's benefit, where in other states the retailer is allowed to round to the nearest 5 cents. This is going to cost large retailers millions.

Interestingly many of them had already put the work into updating the cash register software to allow for this due to the penny shortages during covid.

Right up until your user opens your site in more than 6 tabs.

I have a bunch of little scripts and aliases I've written over the years, but none are used more than these...

alias ..='cd ..'

alias ...='cd ../..'

alias ....='cd ../../..'

alias .....='cd ../../../..'

alias ......='cd ../../../../..'

alias .......='cd ../../../../../..'

Is this different from using a remote docker context?

My workflow in my homelab is to create a remote docker context like this...

(from my local development machine)

> docker context create mylinuxserver --docker "host=ssh://[email protected]"

Then I can do...

> docker context use mylinuxserver

> docker compose build

> docker compose up -d

And all the images contained in my docker-compose.yml file are built, deployed and running in my remote linux server.

No fuss, registry, no extra applications needed.

Way simpler than using docker swarm, Kubernetes or whatever. Maybe I'm missing something that @psviderski is doing that I don't get with my method.

This is exactly what I do, make a context pointing to the remote host, use docker compose build / up to launch it on the remote system.

Taking a photo of my handwritten notes and passing it to ChatGPT works 95% of the time. Once in a while it gets a character or two wrong, but for the most part its magic for me.

The prefil method works pretty well...

https://docs.anthropic.com/en/docs/test-and-evaluate/strengt...

They've been constantly bugging me to upload my government ID, never did. Now I'm really glad I didn't.

I usually tell it to go read the relevant doc when I do the initial prompt to it when I start working on something. And sometimes I'll remind it during a conversation if I want to make sure it isn't re-inventing the wheel instead of using a feature that is already there in the lib. I run into that with TanStack Router from time to time as an example.

We've done some experimentation when using Claude Code and taken to just creating a "vendor" folder under our "docs" section of each of our repos and just pull down the readme file for every library we use. Then when I'm prompting Claude to figure something out, I'll remind it to go check "docs/vendor/awesomelib" or whatever and it does a fine job of checking the docs out before it starts formulating an answer.

This has done wonders for improving our results when working with TanStack Start or shadcn/ui or whatever.

I guess there's pieces of this that would be helpful to us, but there's too much setup work for me to mess with it right now, I don't feel like generating a Gemini api key, installing puppeteer, etc.

I already have all the docs pulled down, but reducing the number of tokens used for my LLM to pull up the doc files I'm referencing is interesting.

Is there a command line tool anyone has had luck with that just trims down a .md file but still leaves it in a state that the LLM can understand it?

Same! I prefer the CLI, way easier when I’m connected via ssh from another network somewhere.

A collection agency has no problem finding your SSN if they have your name.