Different control points seeing different things

[DavidHB]

I have been reading this thread with interest, albeit as something of an outsider, as the Control Points I use predominantly are Bubble DS Next on Android and Linn Kazoo on Windows. Both of these CPs provide near instantaneous display of album art, and the content of the browse tree generated by MinimServer (almost 600 albums, perhaps half of which are multi-disc) is identical in both CPs. When I open either app, there is typically no waiting for the browse tree to populate once the CP has achieved network connectivity, which only takes a couple of seconds when the network is behaving itself.

I have been wondering why my experience is so different from that of the users who have reported results with other CPs. The size of the album art may be part of the issue; most of my album art has been downloaded from the web, and the image sizes are typically smaller than those quoted by other users. But the obvious difference is that other CPs apparently cache the browse tree on the device running the CP, whereas the CPs I am using don't seem to do much caching at all (or do it in the background, unseen by the user).

Caching the browse tree at the CP is AFAIK allowed in the UPnP specification, but opinion is divided as to whether it is a good thing. I am firmly of the view that, for MinimServer users, caching at the CP is generally not a good thing, especially when it is used to override MinimServer settings. If you use several different players and several CP apps, it makes sense to keep the user interface as consistent as possible between platforms, and this means that as much of the browsing experience as possible should be managed by the server. Also caching generates a significant network and processing overhead, which has the results described in previous posts.

It would be interesting if the OP and others who have reported results could try out, say, Kazoo on Windows or Bubble DS (Linn and compatible players only)/Bubble UPnP (other players) on Android. This would at least serve to show which performance issues are related to the CP and which have other causes. Where CP caching appears to create performance and other problems, app authors should be encouraged to make the caching optional, or to dispense with it altogether.

David