I'm using this on my HTPC. It's currently anemic but functional. I've got high hopes

This is the one that broke me

It wasn't even blue on Windows 10, it was the accent color.

That product description sounded to me like a mechanical (not chemical) sunscreen. Unlinke chemical sunscreens those tend to have a visible whitening effect when applied properly. Given that the Choice tests were blind and on human skin, I can imagine a scenario where it was "rubbed in" like chemical sunscreen until invisible, and gave the absurdly low score as a genuine result of misapplication

On the other hand, two independent labs getting similar awful results is damning.

It's unfortunate the responses from these companies are mostly along the lines of "nuh-uh". It's good that there have been some emergency retests, but I would have hoped that someone would have worked with Choice to figure out what was up rather than just telling them "you did it wrong".

A switch is an electromechanical component that can open or close one or more circuits. There are many types of switch, with whole sets of nomenclature depending on how specific you want to get. A keyboard switch is an example of a momentary normally-open single pole - single throw (mom NO SPST) switch [^1]

A button is a user interface / mechanical design component which protudes from a surface and can be manually actuated.[^2]

You can have each without the other.

While a keyboard switch can be used as a button, it's not designed for the purpose, it's designed to have a keycap installed, at which point you do indeed have a button. When we are talking about keyswitches though, we're specifically interested in the electromechanical component, not the portion that a user pushes.

[^1]: Source: domain specific education

[^2]: Source: https://en.m.wikipedia.org/wiki/Push-button

This nonsense is why I've never set my XDG_HOME_DIRS to their actual values. A convenient button in file browsers isn't worth the intrusion.

Microsoft Frontpage?

Unironically, what is that kind of strap called? I have a (much smaller) PC begging for one.

I've migrated from Lychee to Prusaslicer+UVTools. It's less ergonomic, and the auto supports seem a little light-on, but it fulfills my needs and being open source means a lot to me.

That's highly cultural.

Oh god you reminded me of this gem

https://serverfault.com/questions/780150/how-to-cache-contents-in-haproxy#780155

Someone asks how to do http caching in HAproxy.

The one answer:

don't use the wrong tool

haproxy is a wonderful tool. It does not provide caching. A quick scan of the fine docs can verify this. Unless you want to patch haproxy you need to use a tool that does what you're looking to do.

don't create impossible problems

By asking for haproxy to do something that it doesn't and excluding the tool that seems to do what you want to do you've create an impossible situation. There is no technical solution for this. Don't make choices that box you into a corner.

try varnish or anything that actually caches

If you get over that you might find this tutorial on using varnish with haproxy useful or try varnish by itself. Maybe squid or memcached would be more your speed.

In the comments to this ludicrous tirade we get this simple comment:

This was true and valid back then. Nowadays HAProxy does this.

And just in case someone found this looking for an answer, here's the example from that link

backend bck1
  mode http

  http-request cache-use foobar
  http-response cache-store foobar
  server srv1 127.0.0.1:80

cache foobar
  total-max-size 4
  max-age 240

This meme does meat pies dirty (unless "mincemeat pie" is one of those horribly deceptive terms like "mince pie")