Something that has bothered me for a long time is how slow Beatport is to load playlists in the macOS Djay app. I have a lot of large playlists and even loading one of them can take literally a minute or more until it shows the correct number of tracks.
I’m assuming this is a server side issue as Beatport has always been slower to download tracks compared to other services. My question is whether it’s a possibility to add some sort of local cache that saves the metadata of the tracks? So then these large playlists would show much quicker. Obviously not caching the audio just the text of the playlist. It could then check on launch if it changed or something.
Just asking as it annoys me every day and would love a solution!
Hi @Digital_Self, thanks for the detailed write-up. And thanks @PKtheDJ for pinning down the playlist size and connection details.
djay doesn’t currently cache Beatport playlist data locally between sessions, so there’s no setting I can point you to for this one. Before I pass this along to our devs, two quick questions so we can rule a few things out:
Which version of djay are you running?
Do you have tracks in your Beatport Offline Locker, and roughly how many?
I ask about the Locker because a large one has slowed down Beatport loading in the past. In the meantime, it’s worth splitting your biggest list (the ~4k one) into a few smaller playlists on the Beatport side to see if they come up faster. I haven’t benchmarked it, so let me know what you see.
Your metadata cache idea is a good one. I could move this thread over to Suggestions so other Beatport users can vote on it if you like. I hope that helps!
Also sorry for the double reply but just to add - yes making the playlist smaller will technically make it load faster, but it doesnt fix anything. now i just have 2 slow playlists instead of one. the issue appears to be that it loads in bursts of 20 tracks at a time, a few times a second. The number of tracks in the playlist is immaterial, thats just the speed it loads everything. This is also true for the My Beatport section. Tracks load in 20 at a time so it takes a long time to load them all.
Thanks for the additional info @Digital_Self - that’s really useful detail. You’re right that splitting the playlist just spreads the wait around, so scratch that suggestion.
With you on 5.6.10, no Offline Locker, and a wired 200 Mbps connection, I don’t see anything left on your side to change. The pattern you’re describing, steady batches of about 20 tracks no matter the playlist size, and in My Beatport too, is exactly what our testing team needs to look into.
If you can, grab a short screen recording of one of your large playlists loading, starting from the moment you click it that would be helpful for our engineering team.
Please also share your macOS version # and Mac hardware specs.
Rather than changing this topic into a suggestion, I think it would be better if you created a separate suggestion topic for: “Cache Beatport playlist data locally between sessions”. Thanks!
Like spotify is defining rate limits how developers can use the access to data
And note that for
“it loads in bursts of 20 tracks at a time”, why this?
If Dj pro detects with api that there was network error reading the 20 track burst,
dj pro can request spotify server like “send these 20 tracks again, there was error in this data”.
Rathre than “send these 5000 tracks again, I don’t know which of those has failed”.
Also it works like this:
Client send a request to server.
Server calculates a “checksum” for the data before sending it to client.
Server sends the checksum1 to client.
While the client is getting the data, client is also calculating a checksum 2.
If the checksum2 doesn’t match the checksum1 sended by the server, client knows that there was a network failure and then the client can request server to send the data again.
Djay pro, is probably reading pages of 20 items at a time. Most Api’s require paging with a max page size. So it looks like Dj pro is requesting a page, then add the page to the complete result and update the user interface, then the next page and so on.
Probably better to cache the playlist data in the Djay pro database. I don’t have experience with the beat port Api, but Spotify has a versionId for each playlist. If you update the playlist for example in the Spotify app the timestamp gets another value. So you could simplycheck if the playlist is updated. If it is updated you could refresh the cache in the background. I use this method in my own Music collection app. It regurlarly check the versionId and updates if needed. In my Ui I also have a force update button, which I could use if I know I updated the playlist and ‘need’ the new info