alyandon

KDE doesn't need 4 GB of ram - that's just what my laptop happens to have which is more than ample to run the OS + KDE + native applications.

I'm on an older LTS kernel so no support for lru_gen but I will definitely check out zramswap - thanks.

It's the web browser and electron based apps that are the primary consumers of ram on my desktops with the DE and OS ram usage being minimal by comparison.

I have an ancient laptop from 2008 with 4GB of ram that runs a modern KDE desktop and related applications just fine that I use for troubleshooting stuff. However, the moment I open a web browser it basically falls to pieces.

I hate everything about this. :-/

  https://antonz.org/go-1-26/#httptest-example

> To fix this issue, the HTTP client returned by httptest.Server.Client now redirects requests for example.com and its subdomains to the test server

I'm really conflicted about the idea that the https client should silently hijack requests to example.com in order to make a dummy certificate work... or am I just really misunderstanding?

Yeah, it's one of the reasons I use a Microsoft account to collect the PC entitlements and then create a local user account that has a sane profile name and never touch the online account again.

  (1) is already true today. There is no way for services to enforce whether a passkey is stored in software or hardware.

Challenge: Go and try to register a non-blessed passkey type with PayPal and come back and share your experience.

  (2) I understand you don't like the user experience

Pretty much my complaint. Passkeys allow for service providers to do dumb things that result in terrible UX. With Password + TOTP, I don't get asked to touch a sensor, enter a PIN, enter an unlock password, etc.

I'm a bit of a curmudgeon about this.

Until service providers are no longer allowed to:

  1) force the type of passkey stores used (e.g. hardware vs software) when I am providing the passkey store
  2) force me to MFA (e.g. forcing touch ID, entering pin or unlock password, etc) when attempting to use a passkey

I'll continue to stick to plain old boring password + TOTP. I fully understand the security trade-offs like phishing resistance but password + TOTP is secure enough for me.

I can't take anyone that views regulation as universally evil seriously when history is full of stuff like https://en.wikipedia.org/wiki/Radithor

I had put it off for several years because everyone kept telling me how awful the prep was. My doctor kept pestering me to agree to do it so I eventually relented.

Turns out, I needed to go to the bathroom frequently during the day which was an annoyance but I never had anything close to an "accident" nor did I feel any strong urge to evacuate at night. So the whole experience turned out to be a huge nothing-burger and I had a few polyps that got sniped that weren't cancerous - so now I have peace of mind that I didn't have before.

  They say there’s no Emacs — only your Emacs.

This hit home for me. I spent about 6 months working exclusively with emacs to get past the "this is weird/hard because it is unfamiliar to me" stage. At the end of the experiment, I went back to using vim and IDEs.

My take personal takeaways from the experience:

1) capslock/ctrl switching is helpful in so many other areas - so I kept that

2) emacs is something you want to "live in" (e.g. learning to rely on eshell) if you want to really become proficient with it

3) emacs is something you have to be willing to tweak/adjust via elisp to suite your personal preferences if you want to really really really be proficient with it

I didn't hate emacs but it also wasn't for me.

Same for my parents - one of their desktops that they were going to leave on Windows 10 needed to have tpm and secure boot turned on in bios and suddenly it was Windows 11 compliant.

I use powerline ethernet adapters to hook up the media center in the living room. They aren't super fast (~100 mbps) but they are so much more consistent than wifi.

  What purpose outside abetting in avoiding a DUI is there for publishing a live map of DUI checkpoints?

That is easy to answer - letting law abiding citizens going about their personal business know that if they go through an area they are likely to be stopped and subjected to being searched by police without cause.

For me personally it's just old configs I have that redirect from port 80 to 443. You're right though - probably unnecessary in this day and age.

For me, I'll be replacing Windows with Linux on any machine that can't run Windows 11 in a supported fashion.

I honestly thought Microsoft would blink on this but I guess their leadership believes everyone is going to run out and buy an entirely new computer just to run Windows 11. My older parents (for example) aren't going to throw away their perfectly functional desktops/laptops and will instead continue to use Windows 10 with the attendant security risks (they won't pay $30 for security updates) unless I convince them to switch to Linux.

Perhaps instead we can just rely on the plain meaning of common language around "lifetime" income streams and sweepstakes companies that offer those kinds of prizes should be required to purchase an annuity to guarantee that payment.

100% spot on. I do this with subcontractors quite frequently for house related stuff and most of them are quite happy to work with me directly.

LTS in general being slow to uptake new versions of software is a feature not a bug. It gives predictability at the cost of having to deal with older versions of software.

With Ubuntu at least, some upstreams publish official PPAs so that you aren't stuck on the rapidly aging versions that Canonical picks when they cut an LTS release.

Debian I found out recently has something similar now via "extrepo".

Yeah, the lack of an official upstream .deb that is kept up to date (like the official Docker .deb repos) for Ubuntu really kills using podman for most of my internal use cases.

I'd really like to see some combination of Tinc that manages the layer 2 mesh routing with wireguard underneath for the point to point tunnels.

My relatively technically literate older parents fell for that crap as well. Their tablets were virtually unusable because of full screen scam ads being shown as notifications.

Also,

  * No Autoplay for Video or Music

I'm having a hard time getting myself worked up over this. There should be no universe where any application (web or native) is allowed to perform autoplay without explicit user consent.

When I bought my home in 2009 (suburbia in major metro area) - AT&T was offering a whopping 6mbps DSL connection for wired internet. At the end of 2024, they were still only offering 6 mbps DSL.

This year - they rolled out fiber. So, it only took 16 years - maybe you'll get lucky.

I had a similar experience with NixOS and eventually gave up. I liked the idea of NixOS but I do lots of weird things off the beaten path and was constantly troubleshooting why things wouldn't work.

For me, building/running environments in containers is the least amount of friction.

I took the plunge and don't regret it. Despite the condition of the web site the extension is very useful and relatively free of any annoying bugs.

At some point in time, I'd like to take the time to invest building a custom version of the extension to bump dependencies to access more modern support for plantuml/mermaid diagrams.

Nope - fresh install.

I just tried to install the community edition and it fails claiming the license key generated by the site is invalid. Not a great onboarding experience.

  Cloudflare's 1.1.1.1 Resolver service became unavailable to the Internet starting at 21:52 UTC and ending at 22:54 UTC

Weird. According to my own telemetry from multiple networks they were unavailable for a lot longer than that.

  The KDE dev disagrees with this premise, thinks that additional clicks don't matter much in the grand scheme of things, 

I wonder how much that dev would care if someone changed the workflow of their favorite C++ IDE setup to constantly require multiple extra clicks/steps to perform common activities.

I know I would.

Spot on - that is exactly my experience when working with LLMs.

I lean a bit on LLMs now for initial research/prototype work and it is quite a productivity boost vs random searches on the web. I generally do not commit the code they generate because they tend to miss subtle corner cases unless the prompts I give them are extremely detailed which is not super useful to me. If an LLM does produce something of sufficient quality to get committed I clearly mark it as (at least partially) LLM generated and fully reviewed by myself before I mash the commit button and put my name on it.

Basically, I treat LLMs like a fairly competent unpaid intern and extend about the same level of trust to the output they produce.