cross-posted from: https://sh.itjust.works/post/29792601

Hi,

I want to take a little slice of the time I spend watching videos on Youtube and give that slice over to Peertube. I don’t mean all of it, given I watch far too much Youtube, but a slither which I hope to grow as Peertube hopefully matures.

As such I don’t want to have to make a Peertube account, I want to get Peertube videos straight into my Lemmy feed. To an extent I’ve already done this, but the videos are just links to the Peertube instance and aren’t embedded.

And the other issue is that the comments section is, well, a mess. Lemmy hardly syncs the comments and only does so for comments from other Lemmy instances and the videos Peertube instance.

The first issue feels relatively solvable, Peertube embeds. The second feels like something to do with how activity pub works and as such I have no idea.

I do feel that Peertube is a platform that needs the most support from other platforms in the form of integration, as it’s got a very uphill battle ahead of it, and it’s the first one I’ve noticed. but stronger integration between fediverse platforms on Lemmy would doubtless benefit not just those platform but Lemmy as well.

I thought I would post this and get peoples thoughts

  • Admiral Patrick@dubvee.org
    link
    fedilink
    English
    arrow-up
    2
    ·
    edit-2
    9 days ago

    I’m 100% planning on integrating loops, but there’s a problem that needs solved. The embed data returned when querying a loops link doesn’t include an embed video URL (which is how most but not all of the embed media is rendered). There’s also nothing in the metadata for a loops link that can derive the video URL (which is a trick I use for some other embed media or where I don’t want the user to have to use the special “share” links).

    I’ve emailed the Loops team asking if they would add the oc:video meta tag in the header info, but I have not heard back.

    If that doesn’t pan out, I have tested hotlinking the video URL, and I can get that to render.

    Ex: https://dubvee.org/post/2293134

    Since that works, and if all else fails, I can add a server-side API endpoint to Tesseract that will fetch the loops page, parse it for the video tag, and essentially scrape the page to get the video link. That’s messy, but I can make it work (though I’d prefer not to).

    • m_f@midwest.social
      link
      fedilink
      English
      arrow-up
      3
      ·
      9 days ago

      Cool, that’s good to hear. The good news is that if it comes down to parsing the HTML, that’s pretty straightforward. I’ve been poking at it, and it’s nice and semantic (gets turned into a regular <video> element by JS:

      <video-player
          poster-src="https://loopsusercontent.com/videos/85319640150794240/96649242004583579/8yjZniSLk64lxIPomBfUzr5SwMXRKGWXyHpTLruN.jpg"
          video-src="https://loopsusercontent.com/videos/85319640150794240/96649242004583579/8yjZniSLk64lxIPomBfUzr5SwMXRKGWXyHpTLruN.720p.mp4"
          username="thomasdorr"
          profile-picture="https://loopsusercontent.com/avatars/85319640150794240/v0.jpg"
          caption="Snowing in #DuluthMN"
          tags=""
          likes="8"
          comments="1"
          />
      

      And hopefully it’ll get open source soon or have some way of filing bugs/feature requests. I saw that the dev created https://joinloops.org/ and is planning on putting a roadmap there soon.

      • Admiral Patrick@dubvee.org
        link
        fedilink
        English
        arrow-up
        3
        ·
        edit-2
        9 days ago

        I just put together a basic API endpoint and am successfully scraping the video URLs. Have a very rough integration working in dev, so I think this is going to be supported soon :)

        Edit: Hell yeah!