> it seems that Funkwhale would be the easiest one to get people out of Spotify/Apple Music/Youtube Music/etc, and I am trying to figure out what is missing to get more people interested and using it
As someone who's subscribed to Apple Music despite thinking it's overpriced and would love a good alternative, I guess I can be a case study here.
I use Apple Music for roughly three situations:
- I'm driving and want to listen to music. Usually a mixture of music I hear all the time and like, and new music that is similar to the music I already know I like
- I want a random assortment of chill classical music to lull me to sleep or get into a relaxed mood
- A song comes up in conversation and I want to quickly pull it up ("oh, you think that noise we just heard sounds like that riff from song X? Let's check. Hey siri, play song X")
I'm not very well informed about Funkwhale's feature-set, but I'm under the impression that it wouldn't do any of that well, much less all of it. Am I wrong?
The main difference from Funkwhale to the "streaming" services is that you need to "build" your music catalog, either by uploading your own music collection and/or following other people's libraries. So, "out of the box" you are not going to get any music available unless you join some instance which has an extensive catalog of copyright-free music or gives public access to songs they own copyrights.
For your use cases:
> I'm driving and want to listen to (...) mixture of music
Yes, Funkwhale can create "radios" which generates playlists like what you describe. You can choose to generate the playlist according to how often you listen to it or whether it is your "favorites" list or not.
> I want a random assortment of chill classical music to lull me to sleep or get into a relaxed mood.
If you have already access to this music, the radios can be set according to a specific genre.
> A song comes up in conversation and I want to quickly pull it up
That will depend on whether this song is available to you on the instance.
Well that makes the problem pretty clear. People are sticking with Spotify/Apple Music/Youtube Music because those platforms enable algorithmic discovery of new music that they want to listen to.
Somewhere on my hard drives, I have hundreds of gigs of flacs comprising my "old" music library that I used to maintain, because that's how I used to interact with music. I haven't accessed or updated that library in 5+ years now because manually maintaining a personal music library completely fucking sucks for the casual music listener, even with perfect interfaces. It's a lot of effort, and the end result is that you get to listen to the same songs over and over again and are completely on your own for finding any new music that you want. Compared to algorithmic discovery services that can cut through the monotony of repeat plays with fresh songs that I've never heard or barely remember but am still very likely to enjoy, it's no surprise that a lot of people are moving away from manually curated libraries. I don't want my music listening to be limited only to songs that I've previously discovered myself - I want something that has awareness of more or less "all" music to pick good songs for me without having to lift a finger, and existing services have gotten pretty good at that.
I recognize that for copyright reasons at the very least (not to even get into technical issues), projects like funkwhale can't possibly provide this functionality. That's okay and understandable, but it's also the reason why people aren't switching to it.
Funny I lean on YouTube for music discovery but mostly listen out of my own library. The worst problem I have managing it technically is that M.F. Doom hasn’t recorded two albums under the same name.
What does your process look like for discovering music on youtube and then getting that music into your personal library? If you're not discovering while you're typically listening, then you must be setting aside some separate time for discovery, and then curating those discovered songs, acquiring their files somehow, and integrating them into your library, right?
If I was smart about it I'd get one of those YouTube downloader scripts and attach it to something that puts on metadata and do it all in one step.
I do it in bursts and frequently it is motivated by reading something about music, then looking up the music on YouTube and listening to it there for a while before I decide something is worth putting in the library.
For what it's worth I am not particularly interested in music that came out in the last year or so, not like I am fundamentally opposed to it, but given that the event horizon for decent music recordings is roughly 1965, I think there is more older music than I haven't heard yet than there is new music.
Also my approach is album and discography oriented partially because I really like listening to albums and partially because it's often easy to get a whole discography. So it is not like I am keeping track of tracks I like as much as it is albums and artists that I like.
I do the same thing with Youtube & Spotify. Use the algorithmic feeds when I feel like listening to new music (or being notified of releases), buy downloadable versions or CDs & rip them if I like something new that showed up in the algorithm. Most of the time I listen to music from my library, but sometimes I feel like a change.
I don't think I was clear. If you already have the music collection, your "job" will be mostly about uploading it, but after you do that you will be able to enjoy similar features as the ones you are describing. The "radios" are meant exactly to give some "algorithmic generated playlist" that requires very little input from the user.
I think you're being perfectly clear, but I also think you're missing my point. "If you already have the music collection" is the deal breaker. If I want to listen to my own music library, there has always been plenty of solutions for that. Long before any modern music streaming service existed, I was making streamable radio stations of my personal music library available with software like IceCast. But these days I don't want to listen just to music I've already discovered myself, or even just to music that me and a collection of friends have discovered ourselves. I want exposure to essentially all music, whether I or anyone I know has ever heard of it before. I don't seem to be able to get that with services like FunkWhale, and I'm not willing to go back to limiting my music exposure to only songs that me or a select few others have personally discovered and added to our libraries on our own. Mass discovery is the killer feature, while being able to listen to music I own (in whatever presentation) is unimpressive table stakes. And I want that discovery to be happening while I'm listening to music normally, not have to set aside time to discover new music before adding it to my library so my normal "music listening" sessions can pull from it.
> I don't seem to be able to get that with services like FunkWhale
The problem is not with Funkwhale. The problem is with archaic IP laws and our collective unwillingness to fight for copyright reform.
If you don't mind straight out piracy, there is a ton of solutions that could give you a virtually infinite music collection and just get the algorithms to select the songs for you.
> The problem is not with Funkwhale. The problem is with archaic IP laws
Yeah, as I mentioned a couple posts up: "I recognize that for copyright reasons at the very least, projects like funkwhale can't possibly provide this functionality." But at the end of the day it doesn't matter what the source of the problem is or whether or not it can be reasonably fixed. The problem exists, and it makes FunkWhale a non-starter for people like me, as unfair as that is.
> If you don't mind straight out piracy
I don't at all. Pretty much the entirety of those hundreds of gigs of flacs I mentioned earlier are pirated.
> there is a ton of solutions that could give you a virtually infinite music collection and just get the algorithms to select the songs for you
The thing about getting the algorithms to select the songs for me is that if I'm doing that, the service has already bundled a very convenient way to actually play those songs as well, so I don't have any reason to maintain my own library. I'm paying for the service primarily for its algorithm, and the service just happens to also cover the trivial case of actually listening to the music it recommends, so I don't really need another service to handle that part. Even if I did have a reason for wanting to separate out the discovery from the playback, I don't think it would be possible to do without significant friction, and I've found that any suggestion to the contrary usually just ignores or minimizes any use cases that the suggester doesn't personally value.
If you don't mind the piracy and you are willing to pay for a service that can make the music selection for you, one could just host a Funkwhale instance in a jurisdiction that is not aligned with the MPAA and feed it directly with a seedbox. I'd also love to see something like that, but I honestly think we are in a very small group that of people that would support this.
That sounds fine by me and is essentially what I used to do, but I don't see how it's any less effort or any better of an experience than just continuing to subscribe to the well established and polished services that are already widely used. Like, why would I go through the effort of setting up and paying for an illegal service that provides me most of the functionality I already have, but needs more maintenance and is almost certainly rougher around the edges and less well integrated than the corporate services I already use?
Because by "paying to the existing corporate services" you are effectively contributing to their continued dominance and control over the business model?
If ending my Apple Music subscription was going to shift the future to be one where I can conveniently self-host or access a similar but open non-corporate implementation of a global music discovery service with a good an unencumbered recommendation algorithm and player, I would happily do so. Unfortunately I'm forced to live in reality, so I know those things aren't connected at all and I'll just be degrading my own experience for no reason.
Of course they are connected. You may be right that Apple would be largely unaffected by a small group of people dropping them, but you also need to look at the other side of the market. If a small-but-meaningful minority puts a strong signal that they refuse to pay for corporate-controlled services but do want to pay for "independent" alternatives, then you have the right incentive for some player to show up and satisfy this demand.
You'll be "degrading your experience" only in the beginning, but with continued support the alternatives would be surpass the incumbent. It could be seen as an investment.
It's not that different from the "Linux on the desktop" case. Yes, using Linux over Windows was "difficult" in 2008-2012, but if it were not for the people willing to drop Windows then we would never had companies like Canonical and Valve willing to make a bet on it. And Linux might have ~4% of the desktop share, but you'd have a hard time to argue that Windows is objectively "better" and today it's more likely that people will migrating from Windows to Linux than the opposite direction.
>subscribed to Apple Music despite thinking it's overpriced and would love a good alternative
It's not perfect, but a YouTube Music subscription also gives you ad-free regular YouTube. That's a pretty nice bonus if you're going to be paying for music streaming anyway.
> I think that its main problem is that no instance provides a “public” catalog of songs that can be streamed/downloaded by its users, but its “social aspect” is really cool. You can, e.g, connect with your friends and choose to share your private libraries with each other. So if you have a good mix of friends who kept/ripped their CD/Vynil collection or bought songs from their favorite indie musicians, you can end up with a pretty extensive library. This makes it a decent (and legal) alternative to sneaker-net piracy.
This sounds like it shares a lot of the UX of piracy, so I'd be interested to know the bounds of it's legality. What happens when I join the largest instance and upload Universal Media Group's catalog?
There is no magic loophole, but I don't see how this would get "sued into oblivion". There is no "central instance", so even if someone sets up a server which starts to make copyrighted material available publicly, only that server would receive nastygrams and/or be ejected by the hosting provider.
And those that really want to circumvent the law could just set this up in a place which offers more protections. If someone sets up an instance on 1984.hosting and uses a domain from njal.la... how would anyone be able to take it down?
As soon as the industry gets wind about this practice people or servers or companies will get sued (there is after all nothing else these companies can do) and as soon as this happens some people in the line will try to protect them.
Could be server hosting company, the server hosters, source code Providers etc etc.
Btw that doesn't change the law: dependimg on your country there are already clear legal rules regarding lending digital copies to your friends.
They might fake users and start collecting data through those etc.
But only if it becomes a thing.
I think the main reason I'm not seeing anythinguch about lawsuits today is due to streaming services and the amount of legal media traffic compared to things like torrent.
If you upload it and make it public, you will very quickly get kicked out of the instance as soon as someone reports your library as containing copyrighted material.
If may try to be coy about it and abuse the "invite-only" mechanism to share with lots of people, but you are still risking getting reported.
There are also moderation tools in place that can let the instance admin see beyond the "privacy sharing" settings, so a more zealous admin would be the first to want abusers out of their servers.
One of the problems with Funkwhale is that it only works well when all files fed to it are correctly tagged with artist, album and track data which is not the case with many 'discovered' tracks. Untagged (or badly tagged) tracks are hard to impossible to find in Funkwhale while e.g. Airsonic has no such problems since they have the option to navigate the collection through the file system. Add to that the fact that 'sharing' music - which is part of the concept of Funkwhale - has been demonised and sometimes criminalised and suddenly there seem to be few upsides to running a 'Pod' (which is how a Funkwhale collection is supposed to be called). I've been following the project for years and have been running an instance on and off but always went back to one of the Subsonic forks.
33 comments
[ 4.3 ms ] story [ 80.9 ms ] threadAs someone who's subscribed to Apple Music despite thinking it's overpriced and would love a good alternative, I guess I can be a case study here.
I use Apple Music for roughly three situations:
- I'm driving and want to listen to music. Usually a mixture of music I hear all the time and like, and new music that is similar to the music I already know I like
- I want a random assortment of chill classical music to lull me to sleep or get into a relaxed mood
- A song comes up in conversation and I want to quickly pull it up ("oh, you think that noise we just heard sounds like that riff from song X? Let's check. Hey siri, play song X")
I'm not very well informed about Funkwhale's feature-set, but I'm under the impression that it wouldn't do any of that well, much less all of it. Am I wrong?
For your use cases:
> I'm driving and want to listen to (...) mixture of music
Yes, Funkwhale can create "radios" which generates playlists like what you describe. You can choose to generate the playlist according to how often you listen to it or whether it is your "favorites" list or not.
> I want a random assortment of chill classical music to lull me to sleep or get into a relaxed mood.
If you have already access to this music, the radios can be set according to a specific genre.
> A song comes up in conversation and I want to quickly pull it up
That will depend on whether this song is available to you on the instance.
Somewhere on my hard drives, I have hundreds of gigs of flacs comprising my "old" music library that I used to maintain, because that's how I used to interact with music. I haven't accessed or updated that library in 5+ years now because manually maintaining a personal music library completely fucking sucks for the casual music listener, even with perfect interfaces. It's a lot of effort, and the end result is that you get to listen to the same songs over and over again and are completely on your own for finding any new music that you want. Compared to algorithmic discovery services that can cut through the monotony of repeat plays with fresh songs that I've never heard or barely remember but am still very likely to enjoy, it's no surprise that a lot of people are moving away from manually curated libraries. I don't want my music listening to be limited only to songs that I've previously discovered myself - I want something that has awareness of more or less "all" music to pick good songs for me without having to lift a finger, and existing services have gotten pretty good at that.
I recognize that for copyright reasons at the very least (not to even get into technical issues), projects like funkwhale can't possibly provide this functionality. That's okay and understandable, but it's also the reason why people aren't switching to it.
If I was smart about it I'd get one of those YouTube downloader scripts and attach it to something that puts on metadata and do it all in one step.
I do it in bursts and frequently it is motivated by reading something about music, then looking up the music on YouTube and listening to it there for a while before I decide something is worth putting in the library.
For what it's worth I am not particularly interested in music that came out in the last year or so, not like I am fundamentally opposed to it, but given that the event horizon for decent music recordings is roughly 1965, I think there is more older music than I haven't heard yet than there is new music.
Also my approach is album and discography oriented partially because I really like listening to albums and partially because it's often easy to get a whole discography. So it is not like I am keeping track of tracks I like as much as it is albums and artists that I like.
The problem is not with Funkwhale. The problem is with archaic IP laws and our collective unwillingness to fight for copyright reform.
If you don't mind straight out piracy, there is a ton of solutions that could give you a virtually infinite music collection and just get the algorithms to select the songs for you.
Yeah, as I mentioned a couple posts up: "I recognize that for copyright reasons at the very least, projects like funkwhale can't possibly provide this functionality." But at the end of the day it doesn't matter what the source of the problem is or whether or not it can be reasonably fixed. The problem exists, and it makes FunkWhale a non-starter for people like me, as unfair as that is.
> If you don't mind straight out piracy
I don't at all. Pretty much the entirety of those hundreds of gigs of flacs I mentioned earlier are pirated.
> there is a ton of solutions that could give you a virtually infinite music collection and just get the algorithms to select the songs for you
The thing about getting the algorithms to select the songs for me is that if I'm doing that, the service has already bundled a very convenient way to actually play those songs as well, so I don't have any reason to maintain my own library. I'm paying for the service primarily for its algorithm, and the service just happens to also cover the trivial case of actually listening to the music it recommends, so I don't really need another service to handle that part. Even if I did have a reason for wanting to separate out the discovery from the playback, I don't think it would be possible to do without significant friction, and I've found that any suggestion to the contrary usually just ignores or minimizes any use cases that the suggester doesn't personally value.
Of course they are connected. You may be right that Apple would be largely unaffected by a small group of people dropping them, but you also need to look at the other side of the market. If a small-but-meaningful minority puts a strong signal that they refuse to pay for corporate-controlled services but do want to pay for "independent" alternatives, then you have the right incentive for some player to show up and satisfy this demand.
You'll be "degrading your experience" only in the beginning, but with continued support the alternatives would be surpass the incumbent. It could be seen as an investment.
It's not that different from the "Linux on the desktop" case. Yes, using Linux over Windows was "difficult" in 2008-2012, but if it were not for the people willing to drop Windows then we would never had companies like Canonical and Valve willing to make a bet on it. And Linux might have ~4% of the desktop share, but you'd have a hard time to argue that Windows is objectively "better" and today it's more likely that people will migrating from Windows to Linux than the opposite direction.
It's not perfect, but a YouTube Music subscription also gives you ad-free regular YouTube. That's a pretty nice bonus if you're going to be paying for music streaming anyway.
This sounds like it shares a lot of the UX of piracy, so I'd be interested to know the bounds of it's legality. What happens when I join the largest instance and upload Universal Media Group's catalog?
What would happen? Nothing until it becomes a thing than it gets sued into oblivion.
They didn't find a magic loophole.
And those that really want to circumvent the law could just set this up in a place which offers more protections. If someone sets up an instance on 1984.hosting and uses a domain from njal.la... how would anyone be able to take it down?
They don't need to understand the fediverse to sue everyone and than everyone will disable the feature or adds some filter
People moved to Usenet providers but that traffic is much less than it was 5 or 10 years ago.
As soon as the industry gets wind about this practice people or servers or companies will get sued (there is after all nothing else these companies can do) and as soon as this happens some people in the line will try to protect them.
Could be server hosting company, the server hosters, source code Providers etc etc.
Btw that doesn't change the law: dependimg on your country there are already clear legal rules regarding lending digital copies to your friends.
They might fake users and start collecting data through those etc.
But only if it becomes a thing.
I think the main reason I'm not seeing anythinguch about lawsuits today is due to streaming services and the amount of legal media traffic compared to things like torrent.
If may try to be coy about it and abuse the "invite-only" mechanism to share with lots of people, but you are still risking getting reported.
There are also moderation tools in place that can let the instance admin see beyond the "privacy sharing" settings, so a more zealous admin would be the first to want abusers out of their servers.