newnix@exile.digital @architect@linuxrocks.online

Once I get my webserver set back up, I think I'm going to write an article called "why I'll ignore literally everything you say after `curl | sh`".
I'm amazed not only by the prevalence of this "advice" but the total lack of mentioning why it's not a good idea.

At least so long as I'm working here in town, lol. No idea if I'll find someone with cooler work elsewhere or significantly better pay that'll require me to move several hundred to thousands of miles again.

I should check back with them now that they're only ~20-30 miles from me, see if there's anything that might be fun/useful to do on a roughly bi-weekly basis to help give a boost to other potential devs.

Holy shiznit, I just realized that I've been writing code for around 9ys now. That sounds SO MUCH LONGER than it feels.

For what it's worth, those first lines of code were scratch and DarkBASIC because my Jr year in HS started a "video game design" course that basically taught me nothing other than I love tricking computers into doing cool things.

Once I spend some more time on my nombre project, I should look into at least a basic web front-end to it so I can possibly get some buy-in from my coworkers.
Y'all have no idea how useful an initialism/term dictionary can be until your swimming in initialism/acronym/useless-term soup.

Apparently I screwed up a few relatively recent commits, so I need to get either Void or X11 on my DFBSD machine working again.

...Or I guess I could just ssh into my DFBSD machine with some routing changes.

Can't say I'm surprised, but this is another great example of why trusting megacorps to be good stewards of F/LOSS is a bad idea: https://keivan.io/the-day-appget-died/

That said, it's nice to see MS finally get their shit together, I just wish they wouldn't ghost devs like this. I also wish they'd die, because I'm not sure anyone could actually write a worse microkernel OS.

Kinda cool how much information is packed into Fortran's Quickstart guide: https://fortran-lang.org/learn/

I'm not sure I'd ever use the language, but it's really nice seeing the quickstart not only address idiomatic code, but explain some pitfalls if coming from another language, like how Fortran made the heretical decision of having indexes start at 1 instead of 0.

Neat, looks like there's a new wave of spambots from a Russian IP trying to link what are ostensibly drug marketplaces.
Interestingly, these are all coming from well known mailservers, like aol, yahoo, and google.

I'll just have to also find time to set up some of this Juniper equipment to better allow for splitting up networks, maybe even start up my onion omega as a "guest AP".

Fun side effect of starting my home network/VPS overhaul: my #HardenedBSD VPS makes it easy to have an externally hosted time server and a source of constraints for my home network.
At some point, it'll also be a VPN endpoint, so I can remotely manage my home network.

If you're going to write an article titled "What Science Can Tell Us About X", please include actual science. Quoting retrospectives on CVEs and vulnerabilities in general then smugly asserting your conclusion is no more scientific than quotemining shitposts.

could easily be longer, but I'm trying to leave enough time to actually relax at night too.

Protip to anyone breaking into development: Get a clear description of your delivery expectations as soon as possible. It's possible to fix bad architecture, but you want to know if it's necessary long before you're expected to deliver code.
I'm currently spending ~8-10hrs/day just trying to bolt on TLS, nginx, and gunicorn3 on a stack that had only tested debug builds of the codebase(s) and it's not fun.

ok, guess the next thing to work on is setting up HAproxy or something similar to handle inbound connections to jailed webservers and look into what I need to do for either a wildcard LE cert or one properly loaded up with SAN entries for a few subdomains.

Found the issue with my wired DHCP allocation. Apparently the installer configured `/etc/dhcpcd.conf` to always statically assign that now invalid address and gateway.

Wow, it's going to take forever for my OpenBSD machine's clock to get back in sync.

Alright, finally got wifi working, but now my desktop apparently can't help but grab an invalid address on the wired interface, which also dominates my routing table.
So that's going to be fun to fix , but it can wait until tomorrow.

At least runit's easier to manage than systemd, but I'll have to figure out exactly WTF is going on here.

Well, somehow manged to end up with some cabling issues in this home network deployment. Now I've got the router cascading (thanks AT&T for no bridge mode), and just need to figure out what's up with my NAT config.
Might try putting #OPNSense on here to see if I can get a decent pf ruleset to reference.

I'm 100% going to install Janet and Racket on my firewall too, just so I can toy with the possibility of having a LISP daemon managing some pf(4) rules, or similarly doing something banal, but fun to work on from scratch.