I think the broader point is that both genders have seen a decline in mental health over the same time period. Scoping by gender is a choice. One I am highly suspicious of when people are happy to chime in with their change suggestions.
Seems like a good time to ask: what do people use, if not ubuntu, on a web server? Like for example for a rails app, on EC2... asking for a friend (who is me).
In the age of AI, there are only more and more security vulnerabilities to be discovered, and backporting all of them to outdated kernels will be nigh impossible. I’d not rather make such security patches dependent on the timeliness of backports especially if it is done by a distro rather than upstream. I suggest you use a rolling distro rather than anything with LTS kernels. The CVE in the original post CVE-2026-98432 is fixed since 7.3. If a distro doesn’t offer kernel 7.3.1 today, rule it out. (7.3.3 is also out today.)
In general I find distros using old kernels annoying because Linus treats compatibility seriously and “don’t break use space” has been a mantra for many years. On desktops in particular, GPU driver improvements like https://www.phoronix.com/news/Linux-7.3-DRM make it super worthwhile.
As for exactly which distro, it matters less than you think. I use openSUSE Tumbleweed but plenty of my friends prefer Arch.
That is a wild take, especially consider the GP asked for a server use-case. Do you just assume that all new versions should be taken as soon as possible for all packages?
There is a balance: too outdated is bad, too eager for new versions is bad. The tradeoff that LTS tries to take is older, proven versions but with relevant security fixes quickly.
Not a wild take at all. Most server use cases already use containers to precisely specify what the server workload depends on, and can only access what's in the container. I see no reason for the base system to move slowly with proven versions. Even if you don't use containers, you most likely already use language-specific tools to lock in your dependencies. Leave these to the app developers and the sysadmin can move at any speed.
I use Rocky Linux on my web servers and hypervisors.
However, since I run a VPS host most of my customers run Debian or Ubuntu, due to both being more approachable than RPM-based distros.
I'm not a huge Ubuntu fan due to it being a very heavy downstream of Debian, versus Rocky which is 99.9% RHEL. I hold no grudge against Ubuntu, they made desktop Linux very approachable when Mandriva, Knoppix, et al. weren't.
But Canonical is too much of a downstream versus Red Hat, constantly losing standards wars. And unlike BSD (where I can understand) Canonical is a downstream by choice. It's like Sony who keeps losing standards wars to rivals.
for personal use I'd go with NixOS. NixOS "Unstable" is both pretty reliable and receives latest feature updates a lot faster than stable branches. Nix is simply the fastest to iterate with, for me personally, and having a VM or a Docker image declared right in my flake is something I love.
for something more serious than a personal project, I'd go with Debian because of their vast free support windows and some incredible companies providing extended support
Me three. I was surprised how same it all was. Except for the snaps! Debian doesn't force you to use snaps. This was perhaps the top reason I switched.
Ubuntu has been the easy default for things like WSL (Windows Subsystem for Linux) for my coding on Windows, and a linux VM on my home server. But I don't really like the idea of having a distro that doesn't treat security patching with urgency in the age of AI-fueled exploits. Need a standard distro with decent hardware support, low churn, and fast patching.
People are saying Debian. What are the tradeoffs? (other than faster patching on Debian of course)
I am someone who believes ALL Linux distros are basically the same (with outliers like Nix) but on the other hand Ubuntu does use some things like it's own network manager that is not the same as Debian as far as I know. And it has a working remote upgrade functionality which I used just yesterday. On Debian the process is manual and is not recommender to actually do.
> What are the tradeoffs? (other than faster patching on Debian of course)
Older packages, but that's one of the reasons I use it. On the stable branch you will only get security fixes and very rarely updates for critical bugs.
If you need to use newer packages the testing branch is surprisingly stable (it is the base for Ubuntu) but there are a lot of updates.
Honestly Ubuntu doesn't even come to mind as a candidate... Alpine, debian, nix, distroless, scratch, Amazon Linux, coreos are what come immediately to mind that I'd expect to see.
My customers (non technical people) don’t know how to install a CLI. They’ve never opened a terminal.
But they can use an MCP. That’s the main difference.
I've been pleasantly surprised how well it has worked and how robust it is. It's installed at my parent's house and I am many states away so I was paranoid about making it resilient since i'm usually not there and have to fly to get there. So, i really tried to think of everything I could. I thought it might overheat the first afternoon or have other problems, but it's been running for months now uninterupted. Sometimes the OpenRouter credits run out and it stops catching birds , but it still runs normally and after I re-up the credits, it continues working normally. No need to restart even.
I had been using gemma-4-31B-it llm for cost/accuracy and it was super cheap, about $0.65/month. I had also downscaled the size and back tested on bird/no-bird images from my camera.
But, then I noticed that it was missing birds that were at the back of the pool and I needed to use higher resolution and better models and cost went up 10x. Now, it's to high for my liking and I need to do more model testing to get it down to a few dollars a month that seems reasonable. I can also reduce frequency, I'm doing every 2 mins during daylight hours.
My sense has been that accuracy that I need is only possible with the pretty advanced AI models and that OpenCV type approach wouldn't be good enough. Birds are pretty small in the image of the overall pool that I am working with and discerning them accurately is hard. I think it would be possible for other use cases and basically depends on the image source you are working with and the difficulty of detection within that image.
If you had, say, a camera on the inside of a birdhouse and the whole scene was either empty of taken up by a bird, then that would be easy for an on-device model.
I'm seeing some pretty cool projects for bird detection with YOLO online. They may do more than you need; you probably don't care about the kind of bird, right?
Maybe it would be worth it to spend a few tokens on an implementation, just to try!
Pretty sure he was not when I posted my comment. It currently says "more on him at the end" and that bit seems not to have been written yet. Also there are two mathematicians called "John Conway" so it's good form to say "John H Conway".
Author here, good call! I credit Conway in the center of the article, the content for "more on him at the end" was hidden after reader finish the challenge in part 6. I guess it worth to show it by default, updated already.
I'd also need to thanks for Dr. Arthur Benjamin. I saw his "That math show" with my 9 years old where he introduced this Doomsday rule, I thought it was really cool and try to create some interactive tutorial for it. Would love to see if this can help other kids or adults https://www.thatmathshow.com/
reply