This really drives home for me the need to have some kind of separation in algorithms.
I have a few musical moods. I like country music and I like electronic music but I don't like to mix them. Moreover, I listen to a lot of Latin music sometimes but u really don't want to mix that with the other types. It's not necessarily a genre specific separarion even though I've presented it here that way.
All this to say, the first song that popped up for me to rate was latin. I'm not in that mood and don't want to rate latin or country songs right now. I don't want algorithms recommending things in those buckets to me right now. I don't want to swipe it away either tho because i might really like it when I'm in the latin mood. So my only option is to quit.
Many thanks! And really appreciate your feedback. This is indeed a challenging proposition but I really like the idea.
If I've understood correctly:
1. Assuming, your filter is set to ["country", "latin"], you would like the application to "take the hint" (for lack of a better expression) as you are disliking a certain genre (let's say country) and provide results that do not include the genre you're not in the mood for.
2. Further, you would like to have the results mixed i.e. currently, it provides you 50 tracks of a specific genre at a time.
3. Finally, you would like an option for "keeping a track on ice", that is, I don't want to listen to this right now but might later so don't completely eliminate it from the poll of options.
> It's not necessarily a genre specific separarion even though I've presented it here that way.
> 1. Assuming, your filter is set to
My interpretation of OP is the following:
I have different situations for music. I might listen to certain tracks in the early morning, others when coding, and yet others at the gym. If I have song x, I can: like X in the morning, like X for coding, dislike for both, like it for both. Of course, expand the two options to n.
Just like Netflix has different recommendation profiles for different users in a family, music recommendation algorithms should have the ability to make distinct profiles for different moods of one user.
Didn't expect much from it, but i was wrong. It works really really good. Login worked without a problem. Selected some categories and the from/to release years. Switched to the player and immediately a good song begins to play. I played this "song tinder" for 10 minutes or so and already found 25 unknown songs i liked.
But what doesn't seem to be working is the "go back" circle, at least it doesn't do anything for me, or is it meant to do something else?
Also very intuitive design, that isn't very common today!
Thank you so much for your feedback! I really appreciate it!
And thanks for letting me know about the go back button. It should "rewind"/show you the last disliked/skipped track assuming at least one track has been disliked/skip.
Can you please let me know what browser you are using and if you are coming in on mobile/web. I will take a look and see what the issue could be.
I did use Firefox Nightly on Android. Seems to be working now. I think my "problem" was that it only works once and doesn't go back further in history. (Probably did already rewind and then tried to click on it again).
If you like a song and then go back it goes to the song before the liked song, is this supposed to be like that?
Oh and on firefox on android the buttons go underneath the playing song card, if you click fore example the X the songcard is overlapping the X button after that.
My partner and I built out a web app (mobile responsive) that integrates with Spotify API. They designed the UI, I built out the front end. Long story short, just as the title suggests, it's a tool that helps you discover new tracks on Spotify by way of swipe cards.
Built using the good ole' front end fundamentals: HTML, CSS, JS using Next.js deployed on Vercel.
Please feel free to provide feedback, happy to take anything on board. Happy discovering y'all!
Note this app will always be free (i.e. no ads, no user data collection, no payment walls) and I'm just happy my partner and I were able to deliver something that hopefully will help others discover music and have fun along the way .
I gave it a try, cool concept. Had some hits in the first two batches but it just didn't seem to get the message about when I disliked something. I had an entire batch of 'reggaeton', which I disliked every one of. Surely you can early out if its clear you don't like something, and better hone in on the things where there's more of a match?
Thank you! Appreciate your feedback. I agree, the filtering piece is non-existant in its current state. Ultimately it's a client side application and there's no server component except to handle the OAuth flow. Per Spotify T&C I can't store user data so won't be able to track your likes/dislikes per se.
That said, I will file an issue and definitely look to add some "smarter" filtering client side within the current session so you're not seeing the same tracks again, as this issue has popped up on an other thread.
good idea at the same time sad because you will miss the opportunity to learn new music styles which are not based on your preferences, what you would archieve with something like radio. same concept in social media or netflix vs television or print media.
or is there a mode for a 1 in 5 tracks are random to your preferences just so you can have the option to learn something completely new?
Thank you for your feedback! Will take this on board! If I've understood correctly:
1) The tracks that are presented to you should be a mixed bag based on your preferences? Currently the application showcases 50 random tracks for ONE genre at a time.
2) The mixed bag of tracks should also provide you with some tracks that are outside of the genre/year preferences you have specified?
Thanks for your feedback! Hmm, that is odd that it's making those suggestions for you. We received the opposite feedback which we were looking to include in a latter release - currently, the search doesn't leverage locale to carry out the search.
Agreed. The rational was not to "pollute" Spotify with a bunch of public Trackafind playlists. But I've received this feedback elsewhere. Plan is to add flexibility to make the playlist private/public cause I've received feedback for wanting that ability. Alternatively, drop the private scope altogether and have all playlists public by default.
Edit: it needs the create, edit permissions to create the Trackafind playlist and add tracks to it. The follow just comes as part of that auth scope.
This is why there is a play button for each track, I'm afraid :(
Should that change, I will push an update. That said, autoplay is enabled and works for Desktop web as well as non-ios devices (as far as I'm aware see https://caniuse.com/?search=autoplay).
the concept is very cool but im a bit wary of giving access to my account when I dont see the source being open!
Btw for some reason, scrolling down on your site doesn't work, when I opened it I couldnt see the login button at all and i had to zoom out to find it. I'm running librewolf and nixos!
Thank you for your comment. And I understand the concern. Will be opening up the code once I finish refactoring and do some code clean up. I'll take a look at the login button issue. Do you happen to know what your devices' screen dimensions are?
This really drives home for me the need to have some kind of separation in algorithms.
I have a few musical moods. I like country music and I like electronic music but I don't like to mix them. Moreover, I listen to a lot of Latin music sometimes but u really don't want to mix that with the other types. It's not necessarily a genre specific separarion even though I've presented it here that way.
All this to say, the first song that popped up for me to rate was latin. I'm not in that mood and don't want to rate latin or country songs right now. I don't want algorithms recommending things in those buckets to me right now. I don't want to swipe it away either tho because i might really like it when I'm in the latin mood. So my only option is to quit.
The product that solves this will be very useful