INTERACT FORUM

Please login or register.

Login with username, password and session length
Advanced search  
Pages: [1]   Go Down

Author Topic: Uber-complicated smartlist headaches (help!)  (Read 767 times)

ForsakenArdor

  • Regular Member
  • Junior Woodchuck
  • **
  • Posts: 94
  • I hate Jay Leno.
Uber-complicated smartlist headaches (help!)
« on: December 04, 2004, 06:25:44 am »

Okay, so, I've been at this for hours but I can't seem to figure it out.  I want to have a smartlist that has songs rated 4 and 5 stars, without duplicates both from the same artist as well as duplicates of songs that have been covered by a different artist AND I don't want any song that's been played within the past two weeks AND I don't want any version of a duplicate song to play if ONE of the versions has been played within the past two weeks.  Then sorted randomly.

Still with me?

The easiest part is get the 4 and 5 stars rated.  I just put: rating=4,5

Eliminating the recently played songs is also easy:
[Last Played]=>20160

Eliminating duplicates is simple too: ~nodup=[artist],[name]

I figured out how to take out song covers, too, but I'm sure there's a more elegant solution after my first couple steps below.  First, I do a search for duplicates of songs by [name] only and then I take the songs that are the same and fill in the composer field with the writer/composer of the song.  It's impractical to do this with all of the non-duplicate songs so I have to add a Not Composer=<empty> to go with the No Duplicates string.  And since these two strings don't really like each other, to prevent them from filtering out the majority of the songs, I have to stick another string in there at the beginning, followed by or, then the same string again.  The simpliest solution is to use a rating string, since that only eliminates songs I don't want to be played anyway.  Like this:
rating=4,5 or rating=4,5 ~nodup=[composer],[name] -composer=[]


What's proven beyond my comprehension is how to exclude the duplicates of recently played duplicates.  So far, I've created ... let me go slow with this.

First, a smartlist consisting of a normal duplicate group.
~dup=[artist],[name]

Then, a composer (covers/original) duplicates group.  It's easier to show these in their own smartlist than it is to exclude them from another.
~dup=[composer],[name] -composer=[]

A third group lists the songs from both these lists and adds the [last played] string.  I can't type it out, but you go to rule, then playlist and then you X the boxes for both of those lists and have 'Include' selected.
Playlist: list1,list2 [last played]=<20160

This is where I get stuck.  I've done some things before, but they only partially work and fail when you have an odd number of song duplicates (3, 5, 7, etc.)


While I can't work that mess out, I still have trouble combining the strings of what I can figure out.  They just don't mesh well and I don't know how to make them mesh.  This is what I got so far:

rating=4,5 ~nodup=[Artist],[Name] [last played]=>20160 or rating=4,5 ~nodup=[composer],[name] -[Composer]=[] [Last Played]=>20160 ~sort=[random]

This gives me the ratings, songs not played in two weeks (not counting duplicates) and, apparently, plays only one of each song duplicate per sync and then the random sorting.

Why do I want all this?  Simple.  Imagine a playlist with 20 songs.  10 of them are live versions of the other songs.  MC cuts out the duplicates, you play the 10 songs on day 1 and you're good.  Sync again and MC gives you the other songs, the same exact songs for day 2, only different versions.  Before I figured out the duplicates thing in smartlists, I would get sick of hearing the same song over and over again in the same day.

I know this is possible.  It's just really complicated math and logic.  If anyone could figure this out, JRiver should probably make it a preset.  Paging Stephen Hawking.
Logged
Pages: [1]   Go Up