aard

aard , 17 hours ago

Everything is deployed via ansible - including nameservices. So I already have the description of my infra in ansible, and rest is just a matter of writing scripts to pull it in a more readable form, and maybe add a few comment labels that also get extracted for easily forgettable admin URLs.

aard , 15 hours ago

Unless you are gunning for a job in infrastructure you don’t need to go into kubernetes or terraform or anything like that,

Even then knowing when not to use k8s or similar things is often more valuable than having deep knowledge of those - a lot of stuff where I see k8s or similar stuff used doesn't have the uptime requirements to warrant the complexity. If I have something that just should be up during working hours, and have reliable monitoring plus the ability to re-deploy it via ansible within 10 minutes if it goes poof maybe putting a few additional layers that can blow up in between isn't the best idea.

aard , 8 hours ago

I was referring to work setups with the overengineering - if I had a cent for every time I had to argue with somebody at work to not make things more complex than we actually need I'd have retired a long time ago.

aard , 7 days ago

Meanwhile over in Europe - went to the doctor in spring as a cough didn't go away for ages. As suspected nothing he could do much - irritated throat, and just at the time when cold season was giving way for allergy season. So he prescribed some nose spray - and asked if he should also add some antihistamine to the prescription to save me a few eur (didn't check, but it probably is single digits. That stuff is cheap)

aard , 23 days ago

I nowadays manage my private stuff with the ansible scripts I develop for work - so mostly my own stuff is a development environment for work, and therefore doesn't need to be done on private time.

aard , 24 days ago

There is nothing like this availlable currently. Framework probably comes closest, but they only sell in a few countries, and there is lots of stuff to dislike about their solutions - but building your own around a framework board might be feasible.

I have two mnt reforms - as you said, slow and expensive. They have their use for work prototyping for me, but generally wouldn't recommend. They also have the worst keyboard I've encountered in a notebook in the last decade.

aard , 1 month ago

A lot of the Zen based APUs don't support ECC. The next thing is if it supports registered or unregistered modules - everything up to threadripper is unregistered (though I think some of the pro parts are registered), Epycs are registered.

That makes a huge difference in how much RAM you can add, and how much you pay for it.

aard , 1 month ago

Short version: A bunch of shitty companies have as business model to sell open databases to companies to track security vulnerabilities - at pretty much zero effort to themselves. So they've been bugging the kernel folks to start issuing CVEs and do impact analysis so they have more to sell - and the kernel folks just went "it is the kernel, everything is critical"

tl;dr: this is pretty much an elaborate "go fuck yourself" towards shady 'security' companies.

aard , 1 month ago

Making an exception for one organisation, pressured by politicians, would be harmful. BBC has the following policy about neutral reporting:

We don't use loaded words like "evil" or "cowardly". We don't talk about "terrorists". And we're not the only ones to follow this line. Some of the world's most respected news organisations have exactly the same policy

aard , 1 month ago

Ability for AM radios to interrupt other playback for announcements has been around at least since the 90s. Back then it was commonly used to pause cassette playback when traffic announcements were made.

This just requires for the device to monitor radio when on, and to be on - and with how integrated it is in modern days cars functionality I'd say the chance for them to be on is higher than it was in the 90s. So having that functionality is a pretty good way to reach a lot of car drivers.

aard , 1 month ago (edited 1 month ago)

RDS and related protocols like TMC have specifications for both FM and AM transmitters. Those are used to stop playback if an urgent message comes. I'm assuming you have AM stations with such signals in the US (I don't think we have in the EU) - otherwise the AM radio mandate would indeed be stupid.

edit: did some digging (it's been almost 30 years since I cared about that stuff) - seems the US was pretty late to the party for radio data channels, and side channels for AM (which wasn't of that much interest here due to the FM heavy radio landscape in Europe) only was discussed in the early 90s for the US specific variants. I couldn't find any details if that actually ever got implemented. Given that most documentation available on that topic is heavily focusing on EU I'd guess it never got that much use in the US.

aard , 1 month ago

I find this situation rather entertaining. It shows yet again how important it is to educate people on the basics of how LLM work, including how they are being executed - I'm guessing with just a tiny bit more knowledge it'd also have been obvious nonsense to you.

aard , 2 months ago

It has been a while since I touched ssmtp, so take what I'm saying with a grain of salt.

Problem with ssmtp and related when I was testing it was its behaviour in error conditions - due to a lack of any kind of spool it doesn't fail very gracefully, and if the sending software doesn't expect it and implement a spool itself (which it typically doesn't have a reason to, as pretty much the only situation where something like sendmail would fail is a situation where it also wouldn't be able to write a spool) this can very easily lead to loss of mails.

I already had a working SMTP client capable of fishing mails out of a Maildir at that point, so I ended up just doing a simple sendmail program throwing whatever it receives into a Maildir, and a cronjob to send this forward. This might be the most minimalistic setup for reliably sending out mail (and I'm using it an all my computers behind Emacs to do so) - but it is badly documented, so if you don't care about reliability postfix might be a better choice, or if you don't just go with ssmtp or similar. Or if you do want to dig into that message me, and I'll help making things more user friendly.

aard , 2 months ago

It surely is a bubble - so probably a bit different than many other bubbles.

I think OpenAI made the right call (for them) to commercialize when they did - as that pretty much was their only chance to do so. Things has moved fast over the last 1.5 years - and what used to take a decade in tech has happened within months: OpenAI is the dinosaur company grandfathered in, while for already about a year it's been more sensible for anybody wanting to do something with LLM to selfhost (or buy hosting capacity, but put up own data) one of the more open language models, and possibly adjust or re-train it.

As a company owner I get a ridiculous amount of spam for a year already from all kinds of companies building products on top of OpenAI stack, or are trying to sell training or conferences. All those companies will be left with nothing once all the slower users realize technology has moved on. It's like somebody trying to build all their product offerings based on VMWare stack nowadays.

If you as a company want to offer something around AI right now the safest option is probably offering hosting, or if you want to do more hands on, adjustment of open models. Both of those are very risky, and many will go bust in years to come - but not as suicidal as building on top of a closed dinosaur.

aard , 2 months ago

A problem of this bubble is that it is making AI synonymous with LLM - and when it goes down will burn other more sensibly forms of AI.

aard , 2 months ago

All my software can be configured using dedicated configuration files (.c)

aard , 2 months ago

JBOD relies on an optional SATA extension, which most of your controllers won't have.

That leaves you with RAID in the controller - which is a bad idea, as you don't have much control over what is going on, and recovery if it fails will possibly messy.

aard , 2 months ago

Because it does JBOD if the controller supports it. Pretty much none of the controllers you'll find in consumer hardware support that.

aard , 3 months ago

They used to link to my dig wrapper on my homepage for having their clients debug DNS problems for many years - even with translations of my UI in the various language help sites. I always found it amusing that a hoster of their size does that, instead of spending a lunchbreak to throw something together that integrates with their help page.

There also was a non significant number of users which didn't understand that my homepage had nothing to do with OVH, and ended up mailing me about their DNS problems.

aard , 3 months ago

His first colony will be a death trap.

That's a feature, not a bug. His family got rich with mines in south africa, exploiting the locals. For getting more rich by mining mars you'll have to bring your own locals to exploit, and there's no need to make it to comfortable for them.

aard , 3 months ago

https://kyu.de/pictrs/image/ebfe40ed-a98c-468b-89f0-57bed3803d99.png

aard , 4 months ago

Das dachte ich auch, bis ich das erste mal in nem Bus sass wo die Werbung auch auf die Fenster geklebt war. So mit kleinen Loechern damit die Werbung fuer Leute im Bus nicht stoert, was ueberhaupt nichts bringt. Also, wenn ihr was bespruehen wollt besprueht bitte alles wo Werbung ueber die Fenster geklebt wurde.

aard , 4 months ago

Und wieso traegt man es einfach gar nicht ein?

aard , 4 months ago

What kind of monster stores bananas in the fridge?

aard , 4 months ago

I've been trying that for a while until I ran out of searches, and was trying to pay - after getting unsolvable captchas thrown at me several times by their payment processor I eventually gave up. Having a captcha at that point also doesn't make any sense at all - as I'm in the EU my card will have to go through strong verification before adding it. For a US audience the experience might be different - I guess that's also their main initial target.

They also just did a bare minimum job of supporting non-javascript - while it nowadays is pretty much impossible to handle payment without allowing some javascript they also have their own account logic unusable without javascript, and they don't have a way to easily open that in a private session when you get stuck. That'd be trivially solvable by just giving you a URL with an account key attached you can paste into a private instance to do your payment.

metager does that way better - they're usable without javascript, and don't force you to create an account with them. You can create a key with tokens tied to it to unlock search features. You can just use that to enable it in other browsers - and you easily jump into a private instance from the key workflow to just add money to the key.

I might revisit kagi later to see if they fixed some of those problems - but for now metager seems to be the best option. I'm a bit amazed they still exist - it was my main search engine back in the 90s before google came around.

aard , 4 months ago

You nowadays even need to pay attention on Android as you might get an "software related to what you're currently trying to install" with an install button on top of the install button you want - in the location where the install button used to be in googles playstore. Whoever came up with that stupid idea needs their computer privileges revoked for the rest of their life.

aard , 4 months ago

https://kyu.de/pictrs/image/11cfb7f1-b495-41e0-a673-423547a9f95f.png

aard , 4 months ago

He got purged a few years after the war.

aard , 5 months ago

On a real UNIX (not only AiX) killall is part of the shutdown process - it gets called by init at that stage when you want to kill everything left before reboot/shutdown.

Linux is pretty unique in using that for something else.

aard , 5 months ago

killall typically sends SIGTERM by default. It accepts a single argument, the signal to send - so shutdown would call it once with SIGTERM, then with SIGKILL. killall is not meant to to be called interactively - which worked fine, until people who had their first contact with UNIX like systems on Linux started getting access to traditional UNIX systems.

It used to be common to discourage new Linux users from using killall interactively for exactly that reason. Just checked, there's even a warning about that in the killall manpage on Linux.

aard , 5 months ago

Back then I was testing modelines to see the maximum I could push to my 14" monitor. I then backed it with a 1200x1600 virtual screen.

My girlfriend got sick from watching me scrolling around and bought me a 19" display which could do that resolution - and ended up frustrated when I added a larger virtual screen.

aard , 4 months ago

A good starting point for a wikipedia rabbit hole covering the software aspects on how to drive a display: https://en.wikipedia.org/wiki/XFree86_Modeline

aard , 5 months ago

The same people who, in my field (software engineering), don’t know the difference between Java and JavaScript

The same people who don't understand that zip codes are not unique to a country, and do a zip code search on recruiting platforms without also setting a country. And then offer you to move to their country (at your expense) when you explain them the concept of zip codes and countries.

aard , 5 months ago

The real danger is when people start believing the artist more because of how much more aesthetically pleasing they can make their misunderstandings, and trust me it is a real danger.

We already have that a lot in our field - people just buy a shiny UI, and don't care about the rest.

One of the first times I've encountered that was when a customer bought a ridiculously overpriced firewall box because of the easy to use GUI, and asked me to implement a pretty complex rule set. The irony of having bought the fancy UI so the can do it themselves, and then hire an expert to do it instead was completely lost on them.

This thing had troubles doing what was needed there, and had pretty much zero debug functionality exposed - so eventually I suggested they give me one of their old desktops and half a day to see if I can get the ruleset done the old fashioned way with OpenBSDs pf (that was before Linux kernel 2.4 was released, so Linux firewalls couldn't do stateful filtering yet, which was required there) - got everything running in a morning, they decided to just stick with it, and the expensive fancy box was collecting dust.

aard , 5 months ago

RAM is cheap, and even if you're just doing absolute basic shit your current PC will work better with 16GB of RAM (also looking at you here, Apple). If it's not a phone you're buying don't get anything with less than 16GB.

aard , 5 months ago

On a phone the additional power draw of larger modules can be an issue - plus phones are designed to freeze background apps to conserve memory, so you can get away with less.

I currently have 6GB in my phone, which mostly is fine. In a few situations I'd have preferred having 8, though. 4 or less hasn't made sense for a few years now.

aard , 5 months ago

I guess it depends on how you are using your phone. If you're mostly using it between charges (possibly replacing other devices) it indeed doesn't matter. If you care about standby time, or use it as music player or similar tasks more than active use it does matter.

aard , 5 months ago

Wall is a linguist, which influenced several of his design choices. You have a wide variety of expressing what you want in perl, just as with natural languages - some ways are maybe a bit harder to read for newcomers, while others are not worse than something like python. Typically you'd have coding guides for projects.

I did a webchat in perl in the 90s, and eventually rewrote it in php3 - php was easier to manage properly isolated between users than perl via the CGI interface, so it became popular with hosters very quickly. I went back to doing all my web scripting in perl once I started hosting my own servers,though.

aard , 5 months ago

I'm in my 40s and therefore generally in the "get off my lawn, kids" age.

But I totally agree with that article. I've converted quite a few legacy devices with barrel jack to USB-C - and got rid of a huge box of junky old power bricks. Especially for devices I only use occasionally I don't want to search for the matching power bricks - I just want to plug it into one of the 4 USB-C PD sockets I have installed into my desk.

aard , 5 months ago

A major difference is how they interact with feedback - the main reason I never did my own mastodon instance is the developers attitude. "We're not interested in helping you because you didn't set it up exactly as in the guide" was (and maybe still is) all over the mastodon bug tracker.

That was the first thing I looked for when lemmy became popular - and found they were taking deployment issues to even the most absurd system seriously.

Additionally they treat suggestions seriously - even if they personally think it is stupid - and even implement some of that. Pretty much no chance of anything of that happening with mastodon.

aard , 5 months ago

I wasn't quite sure what to think about this, so I've asked my local LLM. Seems it is fine.

https://kyu.de/pictrs/image/75740d07-1543-4702-8f61-3a71212aa04b.png

aard , 5 months ago

Friendly reminder: just don't buy nvidia

aard , 6 months ago

The bit where you have a small view on a large virtual display exists in xorg (I assume it is still there - when I used that it was XFree86).

You'd configure a virtual screen with whatever resolution you want, and your physical resolution generates a view on that which is moving with the mouse focus. I used to run a 1200x1600 desktop on a 640x480 screen until my girlfriend said she got sick watching me and bought me a large screen.

Might be useful if you quickly want to prototype the general idea.

aard , 7 months ago

Make sure you're helping make Lemmy a welcoming place for non-males

I'd phrase that as "make sure you're helping make Lemmy a welcoming place for everybody"

Being active in a pretty friendly tech scene in the late 90s/early 00s I've seen things being ruined for quite a bunch of people who enjoyed being where nobody was judging them for who they were or wanted to be after a bunch of newly joined women decided to try force a bunch of "women only" policies.

Just don't be dicks to each other, no matter who's on the other end. And don't try to force talking about who you are in places where nobody cares - there are specific groups for that.

aard , 7 months ago

Even in this thread I'd rather phrase it explicitly to include everyone - I've seen statements like this interpreted by some individuals as "make it welcoming for women at all costs, which may include making it openly hostile for people not meeting my specific definition of woman", which didn't have a very pretty end result.

aard , 9 months ago

WoL works as Ethernet¹ broadcast, while Wireguard routes IP, one level above that. So for the purpose of WoL the two ends of the Wireguard tunnel are in two different, not connected networks. In theory you might be able to make it work using subnet directed broadcasts - though creating some means to trigger the WoL packet on where you're terminating your Wireguard might be easier to manage.

Simple option would be just logging in via SSH to trigger it (you could script that - define a host in your SSH client config that just executes a command on connection), or something like a simple web frontend which will then trigger the WoL event.

¹ it is probably fair to assume nowadays that you're using Ethernet, and not something like Token Ring. In case you do it still works the same, just the terminology is different.