@p @amerika @jae @phnt @hakui this nigga got the "sex is icky" autism.

fucking grow up, Jimmy.

@amerika @hakui @jae @phnt

> it's gross and messy.

As a man, "gross and messy" are not deterrents.

@jae @phnt @p @hakui

You guys seriously overrate sex with living humans.

@lanodan@phnt@hakui@p@gray@coldacid suse’s thing is actually a daemon, but it’s there to keep things neat and orderly and exactly as configured.

ifup/down scripts and hooks have a bunch of problems. they’re not declarative, nothing is well-defined, they tend to have side effects and catch fire if something isn’t exactly as expected… and like, if you’re using libvirt, openvpn, wg-quick, lxc and so on, they all can leave your network configuration in an unpredictable state that those scripts will not be able to recover from. on debian attempting to restart the network will usually just take your system offline and then abort due to some error.

like, this network shit is messy as fuck and every single layer of it is a problem. some kind of daemon does make sense there. just… not networkmanager. unless it’s a laptop or something where you want to switch wifi networks and have your VPNs autoconnect on some of them.

@mia @phnt @hakui @p @gray Well on servers it often can just be configuring the interfaces at boot time and you're done, no events to listen to except maybe interface creation but that can be a hook rather than a daemon.

There's few times where it can be more complex than that but for those I'm pretty sure you'd be explicitly configuring special software (like with CARP kind of stuff).

Meanwhile I wouldn't be surprised if stuff like NetworkManager still flushes interface configuration the moment they go down, so a switch/router rebooting or changing a cable goes from few lost packets to TCP connection cuts.

@phnt@lanodan@hakui@p@gray@coldacid networkmanager on anything that isn’t a roaming device is pretty dumb yeah. it’s practically unusable without mainstream DE frontends to begin with, and also too limited for servers.

suse did its own thing there because the alternatives at the time (essentially shell scripts or systemd-networkd) were too fragile and unpredictable.

@mia @lanodan @hakui @p @gray
>what pisses me off about arch in particular is the attitude of “fuckups and breakage are a user skill issue”
99% of the time they are.

>there are no safety nets for anything.
Good, I know what I'm a doing and anything that tries to be smart about safety usually gets in my way (apt and dpkg being prime examples of that). Seriously, this is one of the reasons why I still use it after 7 years.

And yes, while rebooting into a live iso after an update broke something like once in a year can be annoying, there are packages that do the whole OpenSUSE™ thing with btrfs snapshots and bootable snapshots with GRUB. It's not magic and it's optional (like it should be).

@lanodan@phnt@hakui@p@gray@coldacid what pisses me off about arch in particular is the attitude of “fuckups and breakage are a user skill issue”

there are no safety nets for anything. if anything breaks, get the rescue system and painstakingly fix it manually. why yes, why don’t you just switch out the running kernel with something you’ve never successfully booted!

@p @gray @phnt @hakui It's not even unique to distros, bugs even when critical get forgotten, even when they are regular, they just need to be in a crowd of other bugs. Which isn't surprising, there's more bugs per day than there is meals, and most of us forget what we've eaten.

And arch users seem to be specially good at that, ended up calling it "works for me".

Also I don't think a statically linked version is really needed, like Alpine has one (official build too) but so far after years of using it it's only been useful to bootstrap a system as you can use apk.static as a seed.
And apk does update itself, but of course with resolving it's dependencies for it (with apk using sonames for libraries) so it just doesn't breaks like pacman.

@p @gray @phnt @hakui > why_arch_is_not_my_favorite_distro.png

Only distro where this type of shit regularly happens.

@lanodan @gray @hakui @phnt Yeah, that stuff's super annoying. The stupid "man page" triggers in Debian, I forget what you have to do to stop them but it is super annoying to do. "Install these 80 packages. After each of them, make sure you reindex the man pages instead of doing that once at the end. In fact, don't do anything in parallel, either: instead of downloading all of them and installing them as they come in, download, then install, then run a bunch of triggers, then download the next one."

CRUX just runs `ldconfig`. It doesn't even rebuild the font cache unless you go out of your way to run the post-install script.

@lanodan @p @hakui I do sort of miss old CRTs mostly because they had a character. They made nice noises, you could see them warm up, etc. Computers nowadays have no character, they are silent with zero feedback of what they do and just sit on your desk or lap.

I don't miss old color LCDs since those were universally garbage with bad contrast , ghosting and they now all have burned out backlights anyway.

>OSDs
I'm thankful to have button only monitors and not those nipple navigation ones that are impossible to use.

@p @gray @hakui @phnt And I feel like FOSS is too much virus-like and widespread to be made illegal, it's like if you'd try to make slang illegal.

Or like making encryption illegal/restricted, when virtually all computers today have and need to have encryption algorithms stronger than can be cracked today (hence why it's reality of hardware, bad implementations, and bad usage that allow cracks).

@lanodan @gray @hakui @phnt For dumb shit I have to use for work, I am able to accept it. For the government, though, I didn't pick the government.

There are occasional rumblings of requiring open-source software in governments in Europe but based on how often they actually do it, those seem to mostly be just negotiating tactics to get a better deal from Microsoft.

@p @gray @phnt @hakui True, and GDPR sucks for this.

But so far only places where you're directly forced to use software are public services, and sometimes jobs but for those there's contracts and so possibility of negociation (specially when done grouped, like if a team just refuses/doesn't-uses a bullshit software, say MS Teams).

Meaning a large chunk can be fixed by "public services may not force usage of a software on people" but I don't see that happening except maybe in some weird country like Japan where they love their faxes.

@lanodan @gray @hakui @phnt

> More like EU privacy laws kind of idea

It was already annoying when you had to do a EULA every time you installed some software; it is now required every time you visit a website. (fundingchoicesmessages.google.com apparently no longer blocks Tor so I had to kill it from this end.)

I don't know when Python started doing the "just shove binaries onto people's computers, it's fine" but I'm pretty goddamn close to going full-.

> It's something you get by having the right to refuse a software, possibly even refuse to have your data processed by one, hence why I'm thinking more of GDPR kind of stuff.

Well, I don't think there is a right to ownership over data. I think (but have no hope because of picrel) the solution is to stop letting these companies assume that they can collect it by actively preventing this.
everyonewillnotjust.jpg

@p @gray @phnt @hakui More like EU privacy laws kind of idea, I want it to be based on something like consent rather than some kind of pre-baked set of refuse/accept-lists like organised religions strongly tend to have.

And well even if it qualifies as Free Software, I'd still want to be able to refuse it.

In fact interestingly, the Free Software definition has:
> “As you wish” includes, optionally, “not at all” if that is what you wish. So there is no need for a separate “freedom not to run a program.”

But it isn't something you get via licences. It's something you get by having the right to refuse a software, possibly even refuse to have your data processed by one, hence why I'm thinking more of GDPR kind of stuff.

That said first step would at least be full rights to refuse EULAs, after all those require you to accept them to waive some rights away.

@lanodan @gray @hakui @phnt

> Kind of thing that makes me wish we could have the freedom to *not* use some software as it's a social problem.

:ignucius: ...Claim a religious exemption?

@p @gray @phnt @hakui Lack of shame of this feels like guys that would say sorry but then do the exact same thing afterwards sometimes not even 5 minutes later.

Kind of thing that makes me wish we could have the freedom to *not* use some software as it's a social problem.

Naw, I was actually testing a tripod solution that I created for the antenna (the JPC-12) and simply didn't add the radial harness that I normally use with the antenna.

In other words, I got in a hurry and did something I know better than to do.

@mcdanlj Couple of different answers...I make a habit of winding a choke on all my feedlines, at the radio end. Does not make a huge difference, but it's easy and cheap and may lower noise levels.

Chokes at the feed point need to be used on antennas that require them, like G5RVs. They're nice but not vital for plain dipoles.

Not having a choke on hand shouldn't keep someone from getting on the air.

@W6KME Ooops, forgot to trim the at-mentions. I meant that question for Wes, after being bit by RF! ☺ I was just curious whether RF had made it back through a choke, and if so, what kind of choke.

My main POTA coaxes have a maxwell choke of 5 mix 31 beads at the radio end, and are covered with PET cable sheath to help protect them from abrasion. (So far I haven't been bit.)

My QRP linked dipole instead has a 7-turn choke on the winder, so that the shield doesn't become an untuned extra radiating element. I just finished building that and have done only one activation with it...