Re: [Rhythmbox-devel] music on external drives

["Andrew Conkling" <andrew conkling gmail com>]

On Thu, Mar 6, 2008 at 1:43 AM, Jonathan Matthew <notverysmart gmail com> wrote:

On Thu, Mar 6, 2008 at 9:37 AM, Andrew Conkling
<andrew conkling gmail com> wrote:
> On Tue, Mar 4, 2008 at 6:06 PM, Emmanuel Beffara <manu beffara org> wrote:
> > Hence my first question: Is it possible to
> > tell Rhythmbox to look for audio files only in one folder? Maybe this
> > has to do with HAL? And is the big CPU usage at the beginning normal or
> > is there a problem somewhere?
>
> I've noticed the same with my iAUDIO X5. I will actually close Rhythmbox
> before connecting it to the computer. HAL is configured to only look for
> music in the Music folder of the device, so I assume Rhythmbox is doing that
> too, but I notice the same behavior. (I've been meaning to reach out and
> send a message like yours. :)

This was broken until shortly after 0.11.4. Regardless of the
audio_folders setting or the information that came from HAL, it was
still searching the whole device due to an incredibly dumb mistake on
my part.

Is this already fixed in SVN, or should I file a bug about it?

> > Second problem: assume I have this external drive mounted an Rhythmbox
> > running, then exit RB, then unmount the external drive, the launch RB.
> > It does notice that the files are not there anymore, and happily removes
> > them from the media library. Problem is, it removes one file at a time,
> > removing about 30 per second, updating its display each time. If I exit
> > RB, remount the external drive and restart RB, the same happens in
> reverse.
>
> That's normal; RB re-scans the library in the background when it starts up.
> If you turn on library monitoring in the Preferences, it will actually do
> this real-time.

What's happening here is that whether the entries are initially
visible is based purely on whether they were present last time, rather
than some combination of that and whether the filesystem they're on is
mounted. It probably wouldn't be hard to eliminate at least one of
these cases by making the initial visibility a bit smarter.

And should I file a bug on this?

Thanks,
Andrew