koel
koel copied to clipboard
implement ampache XML api
Could we implement Ampache XML api, so users could use Ampache compatible players I know Ampache API seems quite bad, but hey, free players for everyone!
https://github.com/ampache/ampache/wiki/Clients
Api documentation https://github.com/ampache/ampache/wiki/XML-API
I'm not against the idea, but not a fan either. Quite frankly I'm not fond of the fact that Ampache uses XML for their API (maybe due to historical reasons). That its API became the de-factor standard thanks to its popularity, and now every other media service should implement the obsolete API in order to gain some player support, is even sadder. This isn't of course, by any means, to bash Ampache or its developers.
Yes, I agree. I am kind of forced to use Ampache right now because of player support(and I am not eager to write python to support my fav media player that also supports ampache quite well), despite the fact that Ampache is too complicated and ugly :). There is nothing wrong with XML Api (take a look at SOAP), but Ampache-s way is messed up. Can't fix it though. If only there was pretty standard that works...
Ampache XML api is indeed for historical reason. I'm not a big fan of it neither even if I'm the current Ampache maintainer (this was created on the early age of the project). But it is, and will be, keep for backward compatibility. About being complicated/ugly, that's partially a matter of taste, a lot of effort was done with Ampache 3.7 and 3.8 on feature simplification and better UI. Ampache API itself is still updated on latest version and except that it is based on xml, there is no real issue with the API. More a matter of taste here again. For other stuffs, I will say https://github.com/phanan/koel/issues/104#issuecomment-165973301 => I don't care about the project, having a consolidated open-source media community is more important here for me, and the API has a driving role.
I don't care about the project, having a consolidated open-source media community is more important here for me, and the API has a driving role.
I'd agree, hence "not against the idea." Will need to buy some time for this though. Unless, of course, someone can send a PR over.
Why not just create a new API standard and players? The beetbox project has a solid proposal.