support for user-definable labels

[Angus Prune]

On 07/11/06, Rich <rich at hq.vsaa.lv> wrote:
> Angus Prune wrote:
> ...
> > The problem I see with the score at the moment, is that if I have my
> > whole collection on shuffle I end up skipping lots of tracks I like
> > but aren't in the mood for.  This adversely affects the track's score,
> > although clearly this matters less with a more mature library and
> > scores.
>
> as your suggestions seem pretty complex and even if they are
> implemented, it would take quite some time, here are some tips to make
> scores slightly more useful :
>
> 1. you can temporarily disable scores
>         a) dcop amarok collection disableAutoScoring true
>         of course, to enable, use the same with 'false' :)
>         b) by stopping the scoring script
>
> note that first way also disables updating of the playcount, which is
> more desirable if you would like to keep your score data (default score
> script takes into account playcount when calculating scores).
>
> 2. if a track is paused, pressing 'next' will not update it's playcount
> and score
>
> 3. you could use even less-impulsive scoring script. these scripts use
> pretty simple calculations, even i was able to grasp them, so maybe you
> can modify default one and save it as "myscores" or something :)


None of those quite achieve what I want, although they are useful
tips,  I guess it would be straight forward fro me to create a button
which automatically pauses then skips.


> > Perhaps my idea shouldn't be applied to labels, but to a new concept
> > called moods, of which a person might have half a dozen or so. (happy,
> > sad, party, mellow etc etc)  although I feel this is needlessly
> > complicating things.
>
> we currently have scores and ratings, and labels are coming in next.
> adding 'moods'... might be too much for most users :)
> and it would be confusing with already existing 'moodbar' :)

I thought that it probably would be.



I'm not a programmer so I can't judge the feasibility of my ideas,
however, it doesn't feel like this should not be beyond our current
capabilities to implement.  Surely it would just be a matter of
storing a score next along with the label.  This would be updated in a
similar fashion to the current score, except using the label's score
instead of the track's global score.

Where does the technical problem arise?  Is it the storing of the
scores?  The updating of the scores?  The size of database that this
would result in?

I had, in my ignorance that modern systems where fully capable of
handling all of the above.
I'm not ready to give up on this idea, but am willing to put what work
I'm able to into refining the idea.

Any advice you can give me would be appreciated.


It is starting to look like this idea is one I file with the robot
chef, self driving car and cure for spam.  Either that or I've got the
next Google in my hands.


James