Intelligent browsing: inconsistent behavior?

[nbpf]

(03-04-2019 12:21)bubbleguuum Wrote:  

(03-04-2019 12:12)nbpf Wrote:  So why does Linn Kazoo in this specific case present a consistent view of the two subsets of tracks (Britten and Schubert) while BubbleUPnP does not?

Because (unlike BubbleUPnP) it doesn't try to infer that a generic container is an album container based on the presence of "Album Artist" metadata associated to the container.
BubbleUPnP do this to workaround some media servers not using the proper class id on containers holding an album, so it can properly display them as albums. I will probably blacklist MinimServer from this behavior so you'd have consistent views (generic list of tracks). Other solution would be for MinimServer to not set an Album Artist on these generic folders.

But Linn Kazoo does offer "Play now", "Play later" and "More" options both for the Britten tracks and for the Schubert tracks while BubbleUPnP does so for the Schubert tracks but not for the Britten tracks!

Thus, no matter what they do, Linn Kazoo displays the leafs of a search tree consistently whereas BubbleUPnP displays logically equivalent results in different ways.

I do not think that blacklisting MinimServer would be a good solution. This would fix the consistency problem but one would be missing the standard controls ("Play now", "Play later" and "More") for sets of tracks that are legitimate results of a search query! Why shouldn't one be able to just play (enqueue, etc.) these tracks in much the same way as one is able to play (enqueue, etc.) a whole album?

From a user's perspective, there should not be first-class and second-class outcomes of a search query. The way a leaf is reached on the search tree should not have any impact on the way that leaf is displayed, in my view.