HereIAm

joined 2 years ago
[–] HereIAm@lemmy.world 5 points 1 day ago (1 children)

The standard specify a ton features and formats. Thing like day if week so 2015-W4-1 would be the first day of the fourth week of 2015.

But the you have can have periods like "P1Y2M10DT2H30M", and you can specify start and end dates. So if you want to start an event that runs for 3 months, 20 days, and some time you could write it as "20220212T1133/P3M20DT7H15M".

And then there's more like giving the year as an exponent, so 2015 can be written as Y-2.015E3S4.

[–] HereIAm@lemmy.world 0 points 1 day ago (1 children)

Ooh, didn't know there was an open source engine for dungeon keeper. Anyone have experience if it runs on wine/proton? I'll give it a shot later either way :D

[–] HereIAm@lemmy.world 3 points 2 days ago

I assume it is GNU Hurd, their own kernel.

[–] HereIAm@lemmy.world 2 points 3 days ago (2 children)

The original idea was the machines using humans as a connected neural network. I don't think it would change much about the plot of the movies if they're used for energy or brain power, so it's easy to change it for your own head canon at least 🙂

[–] HereIAm@lemmy.world 1 points 5 days ago

Believe it or not, but if you wrote down the melody for Bohemia rhapsody (from memory or not) and then sold it, you could be fined for copyright infringement. You can memorise it, you can even cover it, but you can't just sell it. That part still applies to humans. It's the redistribution of that information that's important.

[–] HereIAm@lemmy.world 2 points 5 days ago (1 children)

I was wondering if it was the very same Chipzel.

[–] HereIAm@lemmy.world -1 points 1 week ago

You're completely wrong! This tweet by this very specific person is just a front to distract us from the Epstein files! /s

It is possible two hold two thoughts in ones head and even think both are bad, at the same time!

[–] HereIAm@lemmy.world -2 points 1 week ago (1 children)

Jpeg XL is such an unsexy name though. (In my opinion) JPEG is already seen as an ancient format, there are plenty of even main stream memes about it's artifacting quirks. And now you just want to slap more stuff on it? Make it extra large. Sounds like bloat to me!

To be real for a second; I'm sure it's a fine format and suited for the future, but the name really does put me off it a bit 😄

[–] HereIAm@lemmy.world 6 points 1 week ago

Now we pray Gaben's and Torvalds' names aren't mentioned in the Epstein files for one more day.

[–] HereIAm@lemmy.world 47 points 1 week ago (14 children)

I hope you mean 60hz is enough for TVs. Because I certainly don't want that regression on my monitor 😄

[–] HereIAm@lemmy.world 6 points 1 week ago

In the case of BF 6 and fornite it's "could run, but we actively don't want you to".

 

I finally got some online storage to backup my home server to, and I can finally sleep easy when my house inevitably goes up in flames. My initial push has been with rclone using the sync command. I plan on running a daily or weekly push.

I've gone with Hetzner and they have a variety of options on how to connect to their storage box as they call it. They have an option for automatic snapshotting of, as I understand it, the entire storage box that you can then recreate.

So my thoughts are around rclones sync and copy commands, and the storage snapshots.
Copy) If I use copy, it won't clear up deleted files automatically that I actually want gone.
Sync) With sync I'm worried that say a drive failed (my set up is a bit wonky currently until I can build a proper RAID), it will do a sync, see an entire directory is gone, and go ahead and delete it.
Snapshots) Here I'm worried that any new data that's been added in-between the snap shot and a server failure would be lost. And I don't you can retrieve partial data from a snapshot.

What I have in mind is to use rclone's copy command, but then somehow have a retention period on files that no longer exist on my server. So a deleted file might exist in my storage bucket for a month before being removed from there as well.

How do you guys manage this? Are there any ready made solutions, or do I need to whip up a script that goes through and looks at the last modified dates and removes the old ones?

 

Hey all. I'm starting to plan out how to build a home camera system. For now I just want to use it to keep an eye on the dogs while I'm out of the house, so all of it indoors and with audio, but with plans to expand in the future. My one hard requirement is that the camera themselves are only communicating locally and the streams are accessible outside my network in a secure manner.

I already have a server running some docker containers, including a reverse proxy*, with a GPU (Arc B580) installed for other video streaming. I also got a Google Coral on its way for future camera detection funs. Would the B580 be able to cope with say 2-4 camera streams (of say 1080p quality) and streaming a 4k HDR movie? This support page says it might be possible, but could stretch the limits a bit.

My imagined setup is PoE IP cameras with RTSP streaming to my home server running Frigate (I'm open to suggestions) with some Home Assistant on the side.

For cameras I've seen Dahua and Hikvision recommended. Do they all have/is RTSP a common feature on IP cameras? As none of the cameras I've looked at on Dahua's website has explicitly said they support it.

I've been thinking about installing a separate network card on the server as well just for the cameras. But this might be a bit over-kill, and might be enough to block them on the router? But I image I will need a special switch for PoE either way.

Outside of buying cameras, switch, and cables and then configuring it all, are there any big ticket items I've missed? Or is my set up kinda meek and a separate server for the video streams is recommended?

  • I know a reverse proxy isn't typically as safe as a VPN tunnel, but it's a balance with easy of use.
 

So I want to swap off of Spotify. Most of the time it works great, but the annoyances with their UX are starting to build up. From not ordering albums in release order on certain screens, to having to wait a good few seconds before turning off their shuffle+, and their shuffle not being very shuffle-y to begin with.

I have a couple of requirements:

  • A decent Linux client.
  • Be able to easily select playback device from other devices (for example start playback on my PC from my phone).
  • Preferably pretty straightforward UX philosophy, i.e. haven't started going down any enshitification with AI, "we know best" kind of elements.

I don't particularly care for the highest of lossless quality audio. I don't posses any audio equipment where I would have any shot of telling the difference. As long as its not the experience I had with YouTube music where some random persons heavily compressed upload of a song would start playing.

My main contenders are Tidal, Qobuz, and deezer. The latter two I have very little experience with.

I've tried Tidal before, but my main gripe with it was scrolling through large playlists (about 2000 songs) was very slow, as it loaded in songs as you scrolled through (think endless scrolling on ddg or Lemmy) making it tedious to go to artists starting with a later character in the alphabet. Maybe it was just the Linux client, an issue on my machine, or if they've fixed it since, would be great to hear if any of you have had the same issue.

Qobuz and deezer I haven't really tried or heard much about from a users perspective.

I know some people swear by buying (or ship in under the jolly roger) all their music and use jellyfin or just local files for playback. I'm not very keen on that idea, the convince and discoverability of music on a streaming platform is what made me go to Spotify and away from winamp in the first place.

 

In a recent update to the HSBC app they've added a screen to prevent you from using the app unless you use the default (google) keyboard.

They do a similar thing if you have an accessibility service running that can access the screens content. A fair enough security warning if you've happened to install a dodgy keyboard app, but highly frustrating when using an open source alternative that enhances the security and privacy over the default option (HeliBoard in my case).

I haven't found a way to circumvent the page yet. It would be useful if Android allowed you to block the permission to query all packages, but alas.

 

But it seems to only do this in the home tab. Search and subscription tabs still show the view count.

Now I don't think view count is much of an indication of quality for a video, but the number of likes even less so. It varies quite a bit even on video to video from the same creator depending on if a like is called out for, or audience type.

Certainly not the most egregious change they've made, but a bit of an odd one I can't quite figure out why.

view more: next ›