To be clear, we still plan to use their data at some point. It’s not off of the roadmap. But, especially in the metadata space where we have a lot of irons in the fire, we will always tend to prioritize things that deliver the most value/effort. The point I was making above is that the amount of value present is not huge–significantly smaller than I would have been led to expect based on the enthusiasm in the feature request.
This is in part because Discogs is a small data source, in part because the data is rather messy, in part because it lacks richness/depth and to some extent structure, and in part because it is so vinyl focused (which we are clearly not). The best thing they would do is cover some more releases of albums which we already have data for–this has some value for sure…it’s just not the panacea that Discogs fans sometimes make it out to be. So we are prioritizing it accordingly.
We did extensive exploratory work and prototyping with Discogs data before reaching this conclusion–I am not just shooting from the hip on this. If that exploration had yielded transformative results it would have been turned into finished product and released.