• 1 Post
  • 367 Comments
Joined 1 year ago
cake
Cake day: June 6th, 2023

help-circle
  • I see several Amcrest options that look like they have integrated AI object detection. Frigate on the other hand says you should get a “Google Coral Accelerator”. Do you know if Frigate (or RTSP, I guess) has a way to leverage the built in detection capabilities of a camera (assuming they are built in, and not being offloaded to the cloud)? Or am I better of looking at the “dumb” Amcrest cameras, and just assuming all processing for all cameras will happen on my Frigate hardware?












  • Ahh, yeah proton overhead would add up. I wonder if there’s a way to run a single proton instance that launches all of them. When I was running the wow launcher via bottles, I believe it was running both the launcher and the client it spawned using the same resources.

    And yeah, I’ve almost never had the minimize on lost focus issue, but I’ve mostly used tiling WMs. The cursor getting locked to the window bounds is way more common.



  • Sounds like NakeyJakey’s take from 5y ago. Hoping they take some lessons for GTA6.

    To be fair, it’s not easy to make a big open world that feels immersive and competes with linear games in terms of fidelity (art, rendering, sound, music, etc), even if you know exactly where the player will go and what they’ll do. Trying to then account for every possible permutation of game state and player action is an exponential explosion of work. Without some kind of AI figuring out a believable way for the game to respond in any given situation, your only practical option is to make some assumptions, pick a small set of “golden paths” and polish those.

    R* devs work their asses off to an ethically questionable degree as it is, I don’t think it’s fair to imply they’re not making the best possible experience at that scale with the technology available.





  • if they do, I imagine they have MORE than enough money to be able to afford a CDN or S3

    As long as they’re continuing to run ads or getting enough “subscriptions” to maintain it. I don’t think any twitch streamer, no matter how big an audience they have or how much money they have, would go live just to burn through their cash.

    sadly it’s the point that everyone instantly comes up with WHY folks shouldn’t use Owncast.

    Yeah, that’s not the argument I’m making. Again, I love the idea of owncast, for all the reasons you gave in your last paragraph, but mostly just to give people the option to not be dependent on a for-profit corporation. But like with youtube, tiktok, and other video-based social platforms, they’re costly to run and moderate, and thus difficult to federate. I’m just trying to understand where its practical limits are right now.

    streams are just bunches of files

    Are they? Very short lived files I guess? Because the delay on a twitch stream can be as low as a couple of seconds. Not sure about owncast.