If you could improve PeerTube, what would you improve?
If you could improve PeerTube, what would you improve?
What keeps you away from PeerTube? What features does PeerTube lack? If you were the developer of PeerTube, how would you improve it?
About a year ago I started trying to check out peertube to see if it was worthwhile for uploading my videos to. My first challenge was just finding instances to sign up on. Most of them didn't allow registration. Then for the ones I did find, streaming videos was very slow and laggy. In some cases, I couldn't even view videos. And then, it seemed that I could only search for videos that existed on that particular instance.
Like I said, this was a year ago so maybe it's improved. But in general, it seemed totally unusable for someone just looking for a way to share videos.
Most of them didn't allow registration. Then for the ones I did find, streaming videos was very slow and laggy. In some cases, I couldn't even view videos. And then, it seemed that I could only search for videos that existed on that particular instance.
As a user, I feel this pain. The defenders of how "easy" it is to join federated services, it really isn't. There's choice paralysis as well as waiting for registration and verification. It's not smooth at all. Take a stop watch and set up a YouTube account, upload a video of a cat's butt, and it's like 2 minutes.
As a person who rolled out my own servers, I understand the barriers. I don't have registration because my $5/month server can't handle it, performance-wise or spammers. Even uploading things on my server is slow.
Hard problem to solve tbh and I don't have a solution
yeah for some reason you have to type some sentence for most, better off suggesting something like kbin.earth where you only check a box. But I actually have no idea if you can login with that to lemmy apps or if it just works on their site?
I went looking the other day and had the exact same experience. They have an instance search on the main peertube website but even if you filter stuff out you still get results that don't match e.g. filter for English only instances that allow content creation and you still get french and German instances or instances that don't even let you register.
Mainstream websites require an email and a password. That's it. No thinking. It's done. A lot of the people on Lemmy are internet savvy or software engineers. Of course it's easy for them. Johnny Offstreet wants to open the app store and have it done for them. Which is why, for better or worse, decentralized social media will never reach the moon.
I'm pretty sure that part of the reason Reddit took off was that it didn't even require an email address at a time where most sites did require one to create an account.
I think the biggest thing for everyone is discovery.
I also think the worst parts of youtube are baked into the videos themselves now so re-uploads are just a worse experience even without the buffering.
I cannot agree more about discovery. Easily the biggest issue for me.
Lack of content and lack of recommended page to find content
Content, no clue how to fix it though.
I'm considering starting a channel on Peertube about woodworking. There's an instance that focuses on makers, arts & crafts, DIY and such, sounds like a damn good home for "Howdy folks, today we'll be building this cherry coffee table."
If I make these videos, will you watch them?
Right at this moment: content and performance.
If I start looking around at Peertube, will I find anything I'm interested in watching? There's a LOT on Youtube right now, what's on Peertube?
There's a tendency for alternative platforms to be wretched hives of scum and villainy. LBRY for example felt like the place racist shitheads were banished to when banned from Youtube. The difference between LBRY and Peertube seems to be "something something blockchain."
I noticed a video from a channel called The Giddy Stitcher titled "How to start a (good) Flosstube channel" which for a second I took to mean Free Libre Open Source Software tuber, as if she was going to give tips on how to run an channel on Peertube. No, apparently "floss" = textile arts/string/whatever and that term is similar to "woodtube" for Paul Sellers et al. She apparently uses Peertube to mirror her Youtube videos, and gets thousands of views per video on Youtube and maybe a dozen views on Peertube.
Anyway, watching this video, it buffered HARD. It got better after awhile but...you remember how, back in the day, you could just pause a video and it would buffer? And how it kinda doesn't anymore? It was also that. They don't offer lower qualities below 720p50 which probably doesn't help; I've seen Youtube jump all the way down to 144p to keep the video playing at all.
I know, Peertube is "some people" while Youtube is run by Alphabet. But, maybe we should standardize on lower resolutions and aspire to HD later on, huh?
- Hard to get into and discover channels
- Services like Grayjay, which can put together both recommendations from PeerTube and YouTube and other platforms for some reason only pull the oldest content from PeerTube
- A lack of community outside the biggest Linux YouTubers
A few solutions might be to..
- Somehow find a way to have the PeerTube app work with YouTube
- Include subscriptions for channels so YouTubers can use PeerTube as a better patreon or ko-fi for storing videos and sharing content
- Promote better videos which may entice others to stay longer, have users consistently promote the service personally
I definitely could see a lot more usage of PeerTube if Grayjay could work better together with it, maybe enabled by default
The inverse of this is just doing what Grayjay does but on the PeerTube app instead
Regarding the first question, for me PeerTube has a similar problem as with other fediverse stuff, which besides a lack of greater adoption is a scarcity of sites with a positive, distinct identity/community (last I checked at least) to encourage more people to use it.
Off the top of my head in terms of PeerTube, only TILvids.com comes to mind, which is cool, but remains primarily tech and specifically Linux-related educational videos. I don't mind that, but it'd be cool to see a broader range of educational content on there.
Proper grayjay integration. Though last time I tried peertube, there wasn't much to watch anyway.
The PeerTube Android App is a huge step forward, but it needs two important features to be a viable YouTube alternative:
- possibility to login with my account
- functionality to cast to Chromecast or similar devices
- more content (not a feature)
PeerTube overall lacks decent support for subtitles, which is my answer to all of the questions. Veronica Explains is (so far) the only channel I make an effort to watch on PeerTube because she provides subtitles.
I’d be more than happy with terrible machine-generated subtitles at the bare minimum since they are enabled by default on YouTube when uploading a video. Accessibility is quite important imo
What keeps you away from PeerTube?
I don't think that the economics work for a revenue-free video hosting service, not in 2025. Maybe not ever, depending upon what happens with bandwidth costs.
Text is pretty small. In 2025, it's reasonable for someone to serve text to many people at relatively little cost. That lets the Threadiverse work.
Streaming video to many people is a lot more bandwidth-intensive.
Can you do something in software about that? Ehhh...well, there are limited improvements you can make.
If you give up the on-demand aspect and manage to get people Internet-wide multicast access again a la the Mbone, maybe use forward error correction to provide redundancy to permit reconstruction of dropped packets, you can get somewhat-more-efficient network utilization, kinda like TV with recording. It's still soaking up network resources, just not as much at one point, and if it scales up, at some point network providers are going to want to be paid one way or another.
Maybe people have enough upstream bandwidth today that you can require that everyone run a servent, as some P2P systems do, and only provide downstream bandwidth if they provide upstream bandwidth. Think BitTorrent or Mojo Nation or similar. I don't know to what degree that's practical, and it's basically relying on ISPs not to crack down on some services that are a lot more bandwidth-heavy than others, having users offload costs to other network users; it's really more of gaming a pricing strategy, like FidoNet. Also, NAT and firewalling is going to be a pain in the butt on the network as things stand -- I suspect that the average user isn't gonna be able to punch a hole (if their network provider even supports it).
If we assume that bandwidth gets far cheaper in the near future, to the point where everyone is just running around with so much bandwidth that streaming massive amounts of video just doesn't matter, then, okay, maybe that would do it. But I doubt that that's actually going to happen, and there may be fundamental physical limitations that prevent it. Also, while one could maybe do a video streaming service akin to those today, my guess is that if the availability of bandwidth became that much more available, that people could figure out other desirable things to do with that bandwidth that is a substitute for video, and that might largely supplant traditional video. Like, maybe instead of 2D video, you send a 3D voxel field or something, give the viewer freedom of movement.
Honestly, I'm still not sure that the Threadiverse is going to be able to handle free hosting of high-resolution images in the long term in its present form if usage scales up much. A revenue-free video streaming service seems far harder.
Never heard of MBone, thanks for sharing. I think the fact that there is currently basically one successful video streaming service is indicative of how hard it is to run one. The bandwidth and performance it takes is huge, so unless a federated service is a paid-for service or receives huge donations there are never going to be enough resources to anywhere near match the performance of YouTube.
Ad economics: https://lemmy.ml/comment/16289543
Yeah, though if you go the ad route, then it's kinda doing what YouTube is. Google kinda killed traditional media advertising because they profiled users and could show highly-targeted ads. Maybe they could have some kind of lower level of profiling but still show ads and that'll pull in enough to make a more-privacy-friendly mix possible.
Google may have shown us a revenue model because of how mouthshit they are. "We've demonetized this video for no reason we're willing to articulate." Okay, I won't rely on AdSense then, I'll monetize my channel by:
-
Merchandising. Sell T-shirts with catch phrases on them or things related to my topic at hand. The New Yankee Workshop made its money selling project plans on broadcast TV decades before Youtube was even a thing.
-
Patreon or similar. Allow fans to subscribe to the creator out-of-band, and reward higher donations with anything from a name in the credits to early access to content or even patron-only content. Youtube accidentally allows for that kind of thing. Wouldn't it be funny if Peertube became a good place to put patreon-only videos instead of having them on Youtube at all?
-
Independently negotiated ad-reads. When Tom Scott looks to camera and says "This episode is sponsored by NordVPN" Youtube didn't get a cut of that.
Assuming an audience large enough, this could support the creator, and then perhaps the instance can then bill creators.
I don't know about technical help, like...Peertube has a P2P aspect to it where if you watch a video, you might help seed that video to other viewers, which helps shoulder the bandwidth load for the instance. If I'm a video creator, can I become a permanent peer for my instance and always seed my videos? I figure that would help more than...not doing that.
Youtube did what Internet Explorer did: It got the general public used to the idea that an entire category of something on the internet is free. IE set the maximum price for a web browser at $0. Youtube did the same for video hosting. Can that be reversed?
-
Controversial opinion: I think PeerTube should add support for paid-access content as you could then persuade the creators on Nebula to join. Since they are promoting their YT-alternative already, this would mean extra PeerTube publicity and users for free.
I don't pay for YouTube Premium because (a) I can already block ads and (b) YouTube doesn't offer what I would pay for, which is a no-log, no-profiling policy. As it is, paying just means that YouTube can link my financial details to my profile.
I do pay for Kagi.
I'd be willing to pay for something like YouTube Premium if someone could make something like YouTube Premium, adopted a no-log, no-profile policy, and charged $15/month or something like that. I'm not sure how many people would be willing to do that, though. And there are other challenges:
-
I don't know if a privacy-oriented commercial streaming service is what the people who are working on PeerTube actually want to make.
-
I think that it will be difficult, though not impossible, to build the kind of content and userbase that YouTube has -- it has network effect, which favors large incumbents.
-
What is peer tube? Honest questions
It's basically YouTube, but FOSS, federated and has P2P functionality to lessen the stress on the server that will be streaming the video to the user.
Semi-official search engine: https://sepiasearch.org/
It's another Fediverse service.
What the Threadiverse -- lemmy, mbin, piefed -- is to Reddit, or Mastodon is to Twitter, PeerTube aims to be for YouTube.
EDIT:
Peertube is to Youtube what Lemmy is to Reddit. Individual instances host the content uploaded by that instance's members, viewers can browse the entire federation. To help with server load, Peertube has a P2P function; if 100 people are watching the same video at the same time, some of them will upload parts of the video they've already downloaded to others.
I mostly watch youtube on my firetv stick and I don't bother with sideloading. ill check to see if an app has come up but their annoying store is well. annoying. its not really a problem with peertube. I am going to try using it whenever I would otherwise watch youtube on my laptop but its not very often.
ok I changed my mind. what would really be great for the federation in general is something like saml where once I sign up for one federation thing I could use the same login other places.
Sepia search is great but it could use some modernizing.
The set-up process
Currently, unless you are already familiar with Docker etc, you run headfirst into a brick wall when trying to set up a PeerTube instance.
Until that process is easier to follow and understand, there won't be widespread adoption