I thought of one other use case that would be really nice, from a server utilisation perspective at least, is if jriver acted as a cluster when it came to serving clients (for sharing transcoding load etc). I have no idea whether the internal architecture of jriver is amenable to this but it would be great if it did it
thanks for the info, I didn't realise that was the case for a client, quite annoying (well more annoying that windows is so resolutely single user but still). If you have time to layout your workflow then that would be much appreciated.
Sure, here's how I import new media:
Machines:
1) Fileserver Machine: Linux with all media files on samba shares
2) MC Library Server Machine: Old Windows laptop sitting next to the Fileserver. All the samba shares from the fileserver are mapped as network drives (W: Y: and Z:)
3) Client HTPC/Workstation: Windows Desktop in the living room. All samba shares from the fileserver are mapped as network drives (same letters as Library Server)
(There are one or two other machines in play but they aren't relevant to the workflow).
When I want to rip CDs, import media, or do any other file-based library maintenance I use machine 3). Machine 3) is normally a client on machine 2)'s library server, but when I'm processing media I switch the library on machine 3) to a local library. Once I'm in the local library I rip or import media onto a local drive. Then I rename, tag, get movie and TV info, etc. Once all the files are appropriately tagged and processed, I use the rename, move and copy function to move the files and sidecars from a local drive on machine 3) to the shared drives on machine 1). I have auto-import enabled on machine 2), watching the shared drives on machine 1). So once the files land on the drives in machine 1), the library server on machine 2) picks up the files and the sidecars and imports them. When I'm done importing/processing I switch machine 3)'s library back to being a client on machine 2)'s library server. 9 out of 10 times by the time I get the library switched back to client mode, the files have already imported and are waiting for me. One tip: disable autoimport on the client PCs or strange results sometimes follow.
That method works for anything but a) MC updates on the library server and b) edits to view schemes. For those I still need to remote into the server using a VNC or teamviewer. I also use that method to do minor changes to cover art once the files are on the server.
In an ideal world, if MC linux had total feature parity, I could dispense with machine 2) entirely and just have machine 1) be the library server. We're getting closer, but it might be a little bit yet for the reasons I noted above.