• KillingTimeItself@lemmy.dbzer0.com
    link
    fedilink
    English
    arrow-up
    7
    ·
    2 months ago

    But people are so used to seamless voice and video chat nowadays - and that’s a technical hurdle that AFAIK, no open-source self-hostable projects have come close to solving.

    this is unironically such a big problem, there are great voice chatting solutions, mumble, and the handful of other ones that exist out there.

    There are basically 0 good usable video conferencing/sharing softwares out there. The same goes for desktop streaming. If we just focused like, a little bit more of our energy on these two things, i think the world would unironically get better. It’s 2024, h264 runs on a CPU like nothing, why haven’t we figured out how to do these things yet?

    The ones that do exist are likely to be web based, and thus, webRTC, the dreaded behemoth of both web support and also, generally poor implementation. I just want mumble but with support for video streaming, how hard is it >:(

    • lambalicious@lemmy.sdf.org
      link
      fedilink
      English
      arrow-up
      7
      ·
      2 months ago

      It’s 2024, h264 runs on a CPU like nothing, why haven’t we figured out how to do these things yet?

      It’s not about the hardware. (Not like it’s that ubiquitous anyway; I’m daily driving a machine from 2017)

      I’m going to guess part of it is because for the things that matter to the people who do end up having to code, test and distribute stuff, something like “seamless screen sharing” or “video conference” doesnt really matter.

      And IMO, that’s good if we want to Recover the Web.

      The idea behind being in something like a jabber chatroom, or a web forum, is that I can pay attention to 12 channels (or whatever) at a time, read one or two, reply in three others, etc. Text is so un-invasive that I can just explore without bothering myself or anyone else.

      In comparison, something like audio chat or video chat is more presence-encompassing. You can’t really “push to talk” three different things to three chatrooms at about once, and you likely can but won’t want to listen to three chatrooms full of people at the same time. For something like a videoconference you not only need a camera, but a good behind-you because not only who knows who or what will be showing back there.

      In the end, something like a simple jabber-like chatroom is far easier and more productive to work on, even before we get to the coding part.

      Not to mention: this is computer stuff. No one really likes to work on “debt”, which is what “Foo has to have ‘screen sharing’ because Discord has it” ultimately boils down to.

      • KillingTimeItself@lemmy.dbzer0.com
        link
        fedilink
        English
        arrow-up
        2
        ·
        2 months ago

        I’m going to guess part of it is because for the things that matter to the people who do end up having to code, test and distribute stuff, something like “seamless screen sharing” or “video conference” doesnt really matter.

        this definitely makes sense in the OSS community, but i feel like someone should’ve already done it as a semi pet project already. I know i would’ve done it.

        And IMO, that’s good if we want to Recover the Web.

        that’s an interesting take, but personally i think the web should stick to pretty much static web pages, the browser is turning into a secondary operating system, which is being run on an operating system, which is just, stupid.

        Personally i don’t think any of this stuff should be done over the web, period.

        The idea behind being in something like a jabber chatroom, or a web forum, is that I can pay attention to 12 channels (or whatever) at a time, read one or two, reply in three others, etc. Text is so un-invasive that I can just explore without bothering myself or anyone else.

        yeah, my main complaint though is that we do have things like jabber, this is already incredibly accessible, there is almost no need for expanding the current landscape because it’s been around for like 30 years now.

        In comparison, something like audio chat or video chat is more presence-encompassing. You can’t really “push to talk” three different things to three chatrooms at about once, and you likely can but won’t want to listen to three chatrooms full of people at the same time.

        no but that’s not the immediate use case either, something like mumble is really nice if you’re playing games with other people and just want to VOIP so you don’t have to use a text chat, you can talk and play video games at the same time pretty easily. It’s also nice if you just want to casually hang around other people without having to be physically near them, or at a keyboard typing on it constantly.

        For something like a videoconference you not only need a camera, but a good behind-you because not only who knows who or what will be showing back there.

        i mean, you don’t need a camera, maybe in a professional setting, but in a casual setting, screensharing something to show someone else for example, you don’t even need a camera.

        Not to mention: this is computer stuff. No one really likes to work on “debt”, which is what “Foo has to have ‘screen sharing’ because Discord has it” ultimately boils down to.

        this is fair, and tbh i don’t even really want a discord clone, you could very easily just adapt one of the many existing text chat protocols IRC being the most obvious, and VOIP is basically a solved problem, that’s not hard either. Mumble has a pretty good low latency implementation of it, but you don’t always need low latency. Video sharing/video conferencing is harder, but we have things like youtube and netflix, so the actual video streaming part isn’t the hard thing. We have entire video manipulation libraries like FFMPEG as well, which will do everything you need it to do.

        Mumble i think is the perfect example of a “minimalist” application, it does VOIP and it does it really well. I pretty much just want mumble but for video sharing and i’d be happy.

        • lambalicious@lemmy.sdf.org
          link
          fedilink
          English
          arrow-up
          2
          ·
          2 months ago

          this definitely makes sense in the OSS community, but i feel like someone should’ve already done it as a semi pet project already. I know i would’ve done it.

          Pet project, yes; production-ready, that’s a whole 'nother story.

          Ultimately some things are too complex to deliver out on tem “just because”. Such as web browsers, hence ATM there only exist about 2.

          • KillingTimeItself@lemmy.dbzer0.com
            link
            fedilink
            English
            arrow-up
            2
            ·
            2 months ago

            Pet project, yes; production-ready, that’s a whole 'nother story.

            to be fair, linux was also a pet project, until it wasn’t. I’m not expecting people to drop zoom2 electric boogaloo over this or anything.

            Ultimately some things are too complex to deliver out on tem “just because”. Such as web browsers, hence ATM there only exist about 2.

            web browsers i could see, because they fucking suck, though there are a few alt browser projects currently going on, so there is that.

            but something like VOIP and video sharing i would imagine is probably going to be magnitudes easier than something like a web browser.