• 1 Post
  • 48 Comments
Joined 3 months ago
cake
Cake day: January 16th, 2026

help-circle
  • Isn’t this what every major social media site does? It’s certainly what security and privacy experts have been warning us about for years.

    Once can hope LinkedIn pays a heavy price for this, but they’ve probably done it intentionally knowing the value 100x exceeds the likely penalty. This will probably end up with all of us being offered to join a class action where our settlement is a free month of LinkedIn premium.




  • Ah, I see the unclear part. I read this line…

    I imagine sitting on coach, searching for show. Then you want to watch some, and then you have to wait half an hour for full episode (or even season?) to download.

    As if OP already had a media library, and was outside of their home, sitting on a coach (bus?) and wanting to watch something from their existing library on their phone/laptop/tablet, thinking they’d have to wait for the entire thing to download. This would not be the case. If OP had no content library, and wanted to browse for something new, then yes, you’d need to download the entire thing and add it to your media library first.

    1. Getting stuff into your media library require downloading the thing.
    2. Watching stuff (even remotely) that already exists in your library does not require downloading the whole thing.


  • You can’t watch media before it’s completely downloaded.

    This is not true for just about any use case.

    If you use *arr, you’ll likely use Plex or Jellyfin for a media server. That server will do progressive streaming. Netflix by contrast does dynamic adaptive progressive streaming.

    Progressive streaming means that playback will start once your client has downloaded and buffered enough of the selected content from the server. The amount is typically a fairly small portion of the stream, like 10 seconds or so, though the specifics are left to the server and client configs.

    Dynamic adaptive progressive streaming has a multiplicty of streams optimized for different devices, formats, and quality levels. This might be a few hundred copies of the same video asset, but in a few different codecs, a few different color encodings (ie HDR, SDR), and a quality ladder of maybe 10 steps ranging from low quality SD to moderate quality UHD (like maybe 300kbps at the low end, and 40Mbps at the high end. And these will be cached around the world for delivery efficiency. On playback, the client (player) will constantly test your network throughput in the background, and “seamlessly” adjust stream quality during playback to give you the best stream your network and client can support without stopping to rebuffer.

    For example, if you’re on a 4K/HDR TV with Atmos sound, and great network throughput, you’ll get the highest quality HDR streams and Atmos audio. Conversely, if you’re on mobile that doesn’t support HDR and only stereo audio, you’ll get much more efficiently coded HD video (or maybe SD) and stereo audio streams that are more suited to playback on that device. It would be impractical (huge cost and minor benefit) to try to replicate dynamic adaptive streaming just for yourself.

    In any case, even if you’re just pulling off a NAS, you shouldn’t need to wait for the entire file to download before you can start playback. If your files are properly coded, you should be able to do progressive streaming in just about any use case.







  • I have opnsense, and it was pretty easy. I use DNS overrides and a local reverse proxy. When I’m on the home network, the local dns overrides point to the local reverse proxy. When I’m outside the home, public DNS records point to my VPS, which reverse proxies the traffic to my home machine. This way I’m only hitting the VPS when I’m outside the home. Much more efficient.

    I think Side of Burritos’ youtube channel has a guide on how to set this up, but it’s fairly straightforward.






  • Could it be that he observed that the so called “agentic” operating systems (current versions of Windows, MacOS, iOS, Android) are essentially screen-scraping everything people do, and funneling it to the intelligence apparatus? Security researchers have been squawking about this for a while, and even recently the Signal Foundation CEO pointed it out. Or is that too mundane? Is it much worse? Intelligence gathering tools like Microsoft Recall are an intelligence agency’s wet dream.

    The election interference thing certainly doesn’t strain credulity, but wouldn’t he be able to disclose something so wildly illegal? That is the whole point of congressional oversight.


  • OK, so after a bit of poking at it:

    1. I agree. The OnlyOffice mobile Android app (called Documents) is a much better mobile spreadsheet viewer/editor than Collabora.
    2. What’s even cooler is that the app works with Nextcloud as a cloud backend. So I can log into my existing Nextcloud instance and get the benefit of the better sheets editor on my existing files with no extra work at all!
    3. They say that OnlyOffice supports markdown as of version 9, but I think they mean the broader platform itself, not the Android app. For example, you cannot create a new .md file from the mobile app, and if you try to open an existing .md file, it displays a “wrong file type” error, but it does successfully open it as a .docx.

    In any case, since it works with Nextcloud, the app, out of the box, is already a more functional mobile spreadsheet editor. That’s a big win in my book. Thanks!