My "production stuff":
- https://github.com/festivus-es/festivus - public holidays calendars for Spanish cities
- https://github.com/remote-es/remotes - companies hiring in Spain for remote positions
Usable WIPs:
- https://alexpdp7.github.io/selfhostwatch/ - track self-hosting package updates (such as YunoHost)
- https://github.com/alexpdp7/ubpkg/ - package manager for "upstream binaries"
- https://github.com/alexpdp7/termflux - Miniflux terminal client
So I wanted to try Chimera Linux. But then I realized (again) that it does not have an installer. Well, I installed Gentoo in 2002, I should be able to do this?
But I have not produced a bootable system. Perhaps systemd-boot + ZFS was too daring. Or well, this laptop always gives me issues with booting.
Maybe I'll play with something easier :D
I'm not sure Google is indexing this, but just in case, a Kobo Libra Color does not have a working Internet connection if DHCP pushes multiple routes (as in a VPN setting).
@dabeaz I'm in a similar place. However, I focus more on the externalities. I have judged others as "luddites", so I must constantly question if I'm being a luddite myself.
IMHO, if authors who wish their works not to be used for training LLMs had a magic wand to enforce their will, the LLMs would crumble. To me, that's enough to prevent me from using them.
(I also hope that the same authors would make exceptions for accessibility and similar uses- but that's complex.)
I sometimes wonder if I'm being too stubborn on LLMs. I can acknowledge their utility for certain tasks, but I also look at their many negative externalities. I'm just not on-board with any of that. Also, I derive the most enjoyment from slowly figuring things out for myself and honing my own skills. Maybe an LLM would help me solve problems faster, but I'm much more interested in figuring out how to solve problems better. Of course, maybe it's just a pride thing at this point.
@coder oops, it's before taxes.
In any case, you can use their M4 offering for any non-consecutive 112 days before you pay more than buying the equivalent Mac Mini from Apple.
If you get the M1 offering, the number goes up to 225, although of course you are getting lesser hardware. But if you don't care about performance and you only need macOS...
Some time ago I found a company that rented Mac Minis affordably. However, I think the minimum 24h rental that Apple imposes had killed that.
Today I found that Scaleway has Mac Minis from 0.11€ to 0.24€. Of course, this is subject to the 24h Apple rule, *but* 2.64€ for a day is bordering on reasonable for testing. (It works out to 75€/month, which is a lot of money, though.)
@modulux @glyph what surprises me is the number of fast typers. Over half of the responses do 100 WPM or more. I just did a test and only did 87 WPM, and I don't know so many people who type noticeably faster than me.
(This keyboard is not great, but I think I've never tested over 100 WPM, 90+ is my max.)
I could quote this entire article by Dan North: https://dannorth.net/best-simple-system-for-now
undefined behavior is pretty well understood at this point, but a piece of the puzzle that has always been missing is "how well could a compiler like LLVM optimize, without leaning on UB"
here's a very cool new paper that takes a crack at answering this, for LLVM:
"Vendoring" by Carson Gross
https://htmx.org/essays/vendoring/
"You get more of what you make easy, and if you make dependencies easy, you get more of them."
"This demonstrates significant cultural problem with dependency managers:
They tend to foster a culture of, well, dependency."
Some little-used #Python regular expression features that *aren't* confusing punctuation! https://nedbatchelder.com/blog/202504/regex_affordances.html
@vaurora what I'm finding more frustrating is that I'm not trying to get others to "help me". I'm trying to "help others", and I'm not succeeding.
(Of course, I am likely doing it wrong.)
Tired: Teachers using tools to find students cheating with AI.
Wired: Teachers using tools to figure out how many of their new students are bots who are just there to submit enough AI-completed assignments that they can claim financial aid in someone else's name.
This story from the Voice of San Diego is worth a read:
"When the spring semester began, Southwestern College professor Elizabeth Smith felt good. Two of her online classes were completely full, boasting 32 students each. Even the classes’ waitlists, which fit 20 students, were maxed out. That had never happened before. "
"By the end of the first two weeks of the semester, Smith had whittled down the 104 students enrolled in her classes, including those on the waitlist, to just 15. The rest, she’d concluded, were fake students, often referred to as bots."
"The bots’ goal is to bilk state and federal financial aid money by enrolling in classes, and remaining enrolled in them, long enough for aid disbursements to go out. They often accomplish this by submitting AI-generated work. And because community colleges accept all applicants, they’ve been almost exclusively impacted by the fraud."
"That has put teachers on the front lines of an ever-evolving war on fraud, muddied the teaching experience and thrown up significant barriers to students’ ability to access courses. What has made the situation at Southwestern all the more difficult, some teachers say, is the feeling that administrators haven’t done enough to curb the crisis."
Someone accidentally ordered me to write about the problem with containers, and I did:
@vaurora even using them to bounce off and discuss ideas has always struck me as a sign of the loneliness of modern age.
I think we need to make ourselves more available to others to "help" with their projects, even in detriment of our own projects.
Today we've added a #Kubernetes Cluster API provider for #Incus to the ever growing list of Linux Containers projects! Thanks to Angelos Kolaitis for leading this great initiative!
https://github.com/lxc/cluster-api-provider-incus
Interesting, beyond sshuttle, apparently NetworkManager has an SSH VPN connection plugin, and it's at least in Debian:
One thing I will say about Rust, is that simply learning it has made me a better programmer. Similar to how a static-typed language teaches its users concepts about types, (which you may take for granted if you learned programming with Java and have never met a python developer,) Rust has taught me a lot of concepts related to memory safety and concurrency simply from using it to write software.
I just learned about https://privacypass.github.io/ ... like anything from Cloudflare, better be cautious, but this looks interesting. I wonder how well it works.
Some writing after reading an article about the coding profession some days ago:
https://github.com/alexpdp7/alexpdp7/blob/master/programming/the_tragedy_of_the_geeks.md
If you want to ensure your content does not get indexed by big tech LLMs, just keep it in your robots.txt file.
https://www.truenas.com/docs/scale/25.04/gettingstarted/scalereleasenotes/
TrueNAS Scale adding Incus to the upcoming 25.04 is huge news. LXC support in Proxmox was my biggest reason to stick to Proxmox. Proxmox is still fantastic, and I have everything set up in it, but having another ZFS-supporting appliance with LXC and VMs is fantastic news.
The upcoming #TrueNASScale release comes with #Incus built-in so I took it for a spin to see how the integration was done!
https://youtu.be/ENrQ1RWKZuI
So Lilygo just released this beauty! I’ve always thought e-paper will go excellent with LoRa. Now I want this, so I can put #meshtastic on it! https://lilygo.cc/products/t-deck-pro