On that front I was about to make a post asking for people to link some music. Basically I also want to make a baseline playlist. I was going to merge people's suggestions with what is in /s/Music and my own Spotify like list.
But maybe it is just better to make the base playlist equal to the front page of /s/Music. The one reason why that doesn't work currently is it needs actual media urls or urls that yt-dlp can help me convert to a direct media url. So youtube doesn't work in the current setup because yt-dlp does something different for actual youtube.
Another task I'm working on that I'm hoping I can actually get to is just using the pug (html template) from cytube and working back from there. Essentially getting the same front end of cytube at /player instead of /chat, as well as anywhere for 1/4th to 7/8th of the backend with it depending how much come with that speghetti pull. But sadly I have other things I'm working on that keep dragging me away from that goal. It's in the cards of something I hope happens soon.
But if you want me to do your suggestion faster if you guys could post a lot of music linking either to gvid, catbox, my.mail.ru.video. I'd suggest avoiding pomf because the pomf2 has SSL issues with some peoples browsers and the pomf1 doesn't seek with http-206 which we need for syncing. But if it's gvid it can't be a youtube embed from there. Top way someone could do it if they want to try hard is downloading music, preferably video, uploading to catbox, doing a catbox hotlink on gvid (reduces the media I have to store and deal with copyright for, just a bonus) but also makes it browsable like youtube unlike catbox, and then drop it in goatmatrix to share. With that I could make playlists that work in the current player.
I have code that can bulk do that to a lot of youtube videos. But IDK if anyone would want to do it at that level.
(reduces the media I have to store and deal with copyright for, just a bonus)
What are the copyright issues and how do you deal with them?
How much media do you have, and would it help if we decentralized it?
I always thought it would be good to have a nice storefront-site that could call to the warehouse-archive to then serve it to the user. Ideally that warehouse would be a server at my home or work, under my direct ability to back up and add storage. The much smaller site could be a VPS anywhere in the world.
If you are offering space I can do that. And I have done that before. Gvid is distributed across a few vps's and has had a home server at one point. Home storage really is the cheap storage.
The way I've built it I can make use of both nodes that can serve out to public and ones that can't.
On copyright I have never gotten a single takedown request of any video. Only thumbnails. But I don't want to invite it. So if I want to get an absolute shit ton of music on there I'm pretty ok with it being on someone else's server if I can. I used to be sheepish about over using catbox. But he's actually cool with it, so there is no reason not to. If catbox wants to be a file host that's great. And when it's media that exists elsewhere on the internet the demand for archival stewardship and trusting a 3rd party is at least a little relaxed.
The problem is the url they give (if done manually), does not play globally. From what I can tell it's just for that one browser. I did make some code updates that may expand the domains it will work with, but it didn't quite get there with invidious. If you are talking about reading the code invidious uses and copying it, the problem is invidious has to change how they do it all the time and it plays cat and mouse with youtube. I don't want to play the same cat and mouse game for a player we use once a week.
What I want to do instead, two paths, will be entirely better and less work. One is automating what I suggested there so doing that is as easy as dropping a url into gvid. The other even better one is just using a fragment of cytube's code and having literally the same abilities. I'm working on option 1 ATM, but I'm kind of over doing, it making it as an excuse to get into using Erlang nodes. That will give me better IP rotation in case someone uses it a lot.
Maybe we'll have an actual cytube interface next next time.
Log in
On that front I was about to make a post asking for people to link some music. Basically I also want to make a baseline playlist. I was going to merge people's suggestions with what is in /s/Music and my own Spotify like list.
But maybe it is just better to make the base playlist equal to the front page of /s/Music. The one reason why that doesn't work currently is it needs actual media urls or urls that yt-dlp can help me convert to a direct media url. So youtube doesn't work in the current setup because yt-dlp does something different for actual youtube.
Another task I'm working on that I'm hoping I can actually get to is just using the pug (html template) from cytube and working back from there. Essentially getting the same front end of cytube at /player instead of /chat, as well as anywhere for 1/4th to 7/8th of the backend with it depending how much come with that speghetti pull. But sadly I have other things I'm working on that keep dragging me away from that goal. It's in the cards of something I hope happens soon.
But if you want me to do your suggestion faster if you guys could post a lot of music linking either to gvid, catbox, my.mail.ru.video. I'd suggest avoiding pomf because the pomf2 has SSL issues with some peoples browsers and the pomf1 doesn't seek with http-206 which we need for syncing. But if it's gvid it can't be a youtube embed from there. Top way someone could do it if they want to try hard is downloading music, preferably video, uploading to catbox, doing a catbox hotlink on gvid (reduces the media I have to store and deal with copyright for, just a bonus) but also makes it browsable like youtube unlike catbox, and then drop it in goatmatrix to share. With that I could make playlists that work in the current player.
I have code that can bulk do that to a lot of youtube videos. But IDK if anyone would want to do it at that level.
What are the copyright issues and how do you deal with them?
How much media do you have, and would it help if we decentralized it?
I always thought it would be good to have a nice storefront-site that could call to the warehouse-archive to then serve it to the user. Ideally that warehouse would be a server at my home or work, under my direct ability to back up and add storage. The much smaller site could be a VPS anywhere in the world.
If you are offering space I can do that. And I have done that before. Gvid is distributed across a few vps's and has had a home server at one point. Home storage really is the cheap storage.
The way I've built it I can make use of both nodes that can serve out to public and ones that can't.
On copyright I have never gotten a single takedown request of any video. Only thumbnails. But I don't want to invite it. So if I want to get an absolute shit ton of music on there I'm pretty ok with it being on someone else's server if I can. I used to be sheepish about over using catbox. But he's actually cool with it, so there is no reason not to. If catbox wants to be a file host that's great. And when it's media that exists elsewhere on the internet the demand for archival stewardship and trusting a 3rd party is at least a little relaxed.
Both Inshittious and Freetube can help you get the media URLs although those aren't cli tools but perhaps they can simply help you figure it out.
The problem is the url they give (if done manually), does not play globally. From what I can tell it's just for that one browser. I did make some code updates that may expand the domains it will work with, but it didn't quite get there with invidious. If you are talking about reading the code invidious uses and copying it, the problem is invidious has to change how they do it all the time and it plays cat and mouse with youtube. I don't want to play the same cat and mouse game for a player we use once a week.
What I want to do instead, two paths, will be entirely better and less work. One is automating what I suggested there so doing that is as easy as dropping a url into gvid. The other even better one is just using a fragment of cytube's code and having literally the same abilities. I'm working on option 1 ATM, but I'm kind of over doing, it making it as an excuse to get into using Erlang nodes. That will give me better IP rotation in case someone uses it a lot.
Maybe we'll have an actual cytube interface next next time.
I have a idea I will mail jewtube asking them to please make they are website more friendly to scrapers and provide direct access to the media URL
^