JimmyChezPants 🇨🇦

@aksdb 🤣 @gravitas_deficiency

@quick_snail @ampersandrew yt-dlp works a charm for such occasions. You want the “nightly” build cause Google is actively resisting them, always, but they always win after a number of hours or days.

https://github.com/yt-dlp/yt-dlp

@curbstickle @early_riser I drive a school bus. I have to maintain a seating chart for my kids, in case of accident and whatnot.

Let me tell you about the locked-in software that the Division is probably paying thousands per year to use, which I have to use in order to update that chart.

It’s… from the 90s. No javascript whatsoever. To update a seat, I click one of those HTML1.0 popopen lists. After I choose the new student, I wait about five seconds for the page to reload. There is no option to do a bunch of updates then hit save. Only this Promethean pecking at my geekish liver with each passing second.

I could code something better in a weekend and I never touch LLM slop of any kind, and I offered to, for free, because I will give it to all other drivers, but so far, they are not taking me up on it.

Proprietary = you* are meat for shareholder leeches

*edit: The “you” here being my local community, who have to pay for this garbage, but don’t get to see how fucking appallingly bad it is.

@freeman @mittyta I have not figured out the arrs entirely yet, I believe they make em weird and ugly on purpose, but I have noticed that when I add a series on seerr, it will try to download a full season in a single torrent, but does not seem to search for individual eps if it doesn’t find it.

I then go into the arr and click search on the individual episodes, and it usually finds them then. It’s brilliant at grabbing new stuff, but I’m finding it’s just easier for me to manually download and move older stuff, more and more.

@Bakkoda What software did you use? Or did you code your own? I’m using Navidrome at the moment, it’s okay and there’s lots of apps that extend features using its api, but I’m open to trying out a better project, if you’re onto something you’re happuy with.

@los0220 I might have a go at that since I’ve got a public-facing cloud server already. I’ve opened ports on my home firewall once or twice, but decided I’d rather accept the devil’s bargain of cloudflare than do that again.

Tailscale will probably stay in my mix, because if I ever try to loop a non-techie friend or relative into this, it will be a lot easier to explain “free service that make you connect” than “I’m gonna plug your PC into my janky hacksternet” :>

@Cort @TechSquidTV (looks over at his FreshRSS reader) whatchoo talkinbout Willis

@los0220 That’s the bit that I don’t understand yet about wireguard - you would need something with a static address or dyndns before you could implement it yourself, yes?

As it happens I just rented a linux server from ultra.cc and they have Wireguard as an installable thingy so I might have a go at that myself.

@los0220 Just setup Navidrome on my home Debian server not too long ago, am quite happy with it so far.

I am not familiar with Subsonic, but they use implement its API much like all the other Fedi servers use the Mastodon API, and thusly there are tons of client apps available. I installed Castafiore on my fairly old android, and have not seen any reason to try anything else, so far.

Add in a free Tailscale account and you’ve got your tunes everywhere.

@MIXEDUNIVERS

edit: didn’t close my /del tag correctly. It will be interesting to see if this correction edit gets back to Lemmy.

@JensSpahnpasta Sorry meant to reply the other day - also, I’m kinda fascinated how well my Mastodon instance is communicating with this Lemmy thread.

I’m not the dev of AudiomuseAI, just an interested tire kicker.

But like I said, based on what I know about Machine Learning, this could potentially be better at this specific set of human problems (creating playlists, finding similar new artists to maintain the presence of novelty, etc) than trying to draw upon the crowdsourced metadata (meatdata?) of a species of incredibly opinionated apes.

Like for wo examples, we can only assume that the weights will not care about considerations like authenticity as defined by your typical music fan, nor how “commercial” the track is, which has more than once bit me in the ass.

Like for instance, a friend of mine a few years ago played me this record that was perfect mid-70s Heavy Rock of exactly the sort that I like. And another time I heard this other tune on the radio, similarly full of that 70s Overdrive that I love, and was all “holy shit who’s that?”

The bands in question were The Osmonds (yes, those Osmonds) and Abba. I would never have looked in their back catalogues for good music that I missed. This thing will do that, in theory, as long as the weights get trained on a given track.

Still to be solved, seems to me, is a repository where people can share their trained weights for a given song; we could save a lot of time that way.

Currently 40% finished initial scanning after 48 hours, by the way. My collection is slightly bigger than 1tb of mostly Flac files.

@aquantumofdonuts I used Plex until Jellyfin reached Minimum Viable Product stage and never went back, and when I used it I did not pay, so I will never have any idea about that 😜

@JensSpahnpasta @Routhinator I’ve been playing with just such a thing this morning - I just started the ~80hr initial (CPU) analysis of my flac collection. Once complete, it purports to be able to do quite a lot stuff that sounds both useful and credible, based on what I know of this technology.

Entirely local processing, no planetkiller concerns. It does have the option to use the nasties, and yes, tempting to pay a few bucks to some bondvillain and have it done in a few minutes, but I’m sticking with good ol’ patience to get the job done on the same CPU I leave running 24/7 anyways.

(I might do some experimenting with Kubernetes later, cause my understanding is that I could make all our household machines into workers, and we have another couple in the house…)

@irmadlad I don’t run this instance, but I do generate so much of the traffic, I’m told, that I might as well 🤣

I already registered chezpants.ca nearly a year ago and I run my own servers, I’ve been stuck in decision paralysis about regular Mastodon vs GoToSocial. I think I’ve decided, just need a day to get it up and relearn CloudFlare…

@irmadlad I have a couple of sites going through Cloudflare Zero Trust, strictly done to get my head around Cloudflare, but at this point I’m kinda settling into just using Wireguard for everything personal, and I’m gonna push my Fedi servers through Cloudflare, because they are all public-facing information; if someone has a secret lost on the fedi, it’s not because of Cloudflare, it’s because they posted a secret on a 100% public network.

@enchantedgoldapple

@Foofighter ZFS is not as difficult as you might think. In fact, I found learning how to do ZFS easier than learning to deal with a hardware raid at my first sysadmin job.

It is also so damn good that you can do it on regular commodity hardware, like any mobo with six sata ports, and have no concerns.

Feel free to @ me if you decide to have a go at it on a Linux system. I guarantee you, we could get on a chat session, I’d guide you through the commands with links to the docs, and all of a sudden you’ll be done and be like “that was it? It’s running?”

Not that you won’t have to get a few concepts packed in like datasets and snapshots, but those concepts enable you to have superpowers.

Superpowers.

@lka1988

@bernhoftbret Others have mentioned already, the main issue being if the network comes up before the Pi-Hole does, some devices will be confused.

One possibility to mitigate this, you can get a UPS for your Pi that uses one or more 18650 batteries, fairly cheaply.

I also thinks to myself, some routers can surely be made to act as a DNS cache for the Pi-Hole, and thusly, maintain a working table through disconnections/powerlosses. Whether that applies depends on each case, the hardware and the wetware.

@NightOwl

He has no compelling reason to do so. The railroad lockout put our entire economy and future health in jeopardy - it was a deliberate choice by the two corporations who coordinated it to threaten absolute scorched earth in Canada.

Airlines are not the vector for anything like that much of our economy. Lots of pain but not a deathblow.

@Fudoshin @weilawei

…and THAT being said, I’m well aware of what the CCP has been doing on college campuses over the last ten years, and they are not the only Asian nation that is operating unofficial police forces and hit squads right here in Canada (lookin’ at you India), so I’m also not necessarily *denying* these people were acting as CCP agents. Claimed to be media I think?

It’s also possible they’re just rich, with the usual rich people’s sense of entitlement to dictate to others.

@Fudoshin @weilawei

That being said, there is a strange presumption on the part of the guy walking up and trying to tell the guy not to film in a public space in the UK; I would likely have reacted exactly the same way, though I don’t remember him pointing out to the guy that he was livestreaming, and therefore the *only* move left is to get out of range of the already-live camera. Pointing out that it’s too late for them not to be seen might have been a quick end to it. ->

@Fudoshin @weilawei

I watched this yesterday, I’m well along the path of learning piano and was already gigging, so anyone doing these sorts of public performances is interesting to me right now.

I don’t see anything in the video that says to me that these are CCP members specifically; we Canadians tend to put a flag on our backpacks when abroad, and folks traveling with a language barrier are well-advised to do the same. It is a “communist flag” but it’s also their national flag. ->