Yes i would also like this feature to be implemented :)

*** Bug 117464 has been marked as a duplicate of this bug. ***

*** This bug has been confirmed by popular vote. ***

does anyone works on implementation?
unfortunately i have not much time - all i got at the moment is inserting mulptiple items to playlist when .flac file with embedded cuesheet is dropped.

if anyone has more time, i can email the diff to reduce effort

also have a look at #121519 where i posted a cuefile track-change bugfix.

fyi: got some progress, but i implemented it in engine controller class, while it is better to do this in enginebase (for session resume etc), so i'll be movin my code to enginebase class. i hope i'll finish patch soon after 1.4 :)

*** Bug 118580 has been marked as a duplicate of this bug. ***

Does there exist a diff or branch somewhere?

Created attachment 15639 [details]
!!Incomplete!! Patch for proper cue sheet support

here it is, i'm stuck with non-immediate reaction to stop call.
it stops after playing 0.3-1 second of the next track :(

maybe the reaction will be better if we call pause first?

Hmm, I applyed your patch, added a .flac file with cue sheet inside and it resulted only in one element in the playlist.

$ metaflac --list *.flac
...
METADATA block #1
  type: 4 (VORBIS_COMMENT)
  is last: false
  length: 950
  vendor string: reference libFLAC 1.1.1-beta1 20040731
  comments: 1
    comment[0]: CUESHEET=
...

insertion of several items works for me on current svn, but with xine it hangs (so i kill it), and gstreamer has stopped working even on usual mp3 after update

metaflac --list *flac gives
METADATA block #2
  type: 5 (CUESHEET)
..
  number of tracks: 4
...
METADATA block #3
  type: 4 (VORBIS_COMMENT)
...
  comments: 1
    comment[0]: CUESHEET=
...

It adds only one item :/ in CueFile::loadFromTag the sheet string list is empty. Also, I'm unable to find the specifications how to add title and author data tags. According to http://flac.sourceforge.net/faq.html#general__no_cuesheet_tags there isn't one really.

maybe the problem is w/ taglib?
i have 1.4-3 (debian)

Version 1.4 (gentoo) here.

Got it to work by manually adding the whole content of a .cue-file into the CUESHEET comment. Is this really a correct way of doing this? There is something called CUESHEET METABLOCK in flac files which I generated when encoding to flac (flac --cuesheet=cdda.cue cdda.wav)

>Is this really a correct way of doing this?
Oh yes! other formats, like ape or wavpack, dont have such metablock.

anyways, the cue handling should be done as in foobar2000.org

Created attachment 15666 [details]
Patch for adding length of the last track in CUESHEET in a flac file.

About the non-immediate; it stops immediate for me (using gst10 engine).

btw, attaching patch that'll add the length of the last track (might not be the best way).

did you try mixed cds?
(i use 'Moving Shadow 04.1 - Mixed by Timecode':)

What do you mean with mixed cds? I tried with just ripping a song from a CD i had and added:
PERFORMER "My CD"
TITLE "My Title"
FILE "cdda.wav" WAVE
  TRACK 01 AUDIO
    TITLE "Track 1"
    PERFORMER "My TiTle"
    INDEX 01 00:00:00
  TRACK 02 AUDIO
    TITLE "Track 2"
    PERFORMER "My TITLE"
    INDEX 01 02:37:12
  TRACK 03 AUDIO
    TITLE "Track 3"
    PERFORMER "Title is mine"
    INDEX 01 03:10:04
as cuesheet.

i mean how can you be sure that it stops immediately.
eg, play it with foobar2000 (under wine) and compare

I play the song. Press stop. The music stops immediately.
amarok: END__: void GstEngine::destroyPipeline() - Took 0.0095s

i'll try to implement proper stopping for gst this weekend - i was pointed to http://bugzilla.gnome.org/show_bug.cgi?id=338290

when trying to skip to the next song the whole cuesheet is skipped and the next _file_ gets played ... 

hi, 
there is still an issue when a cuesheet ends and a radiostream/lastfm stream is the next item in the playlist - the title(metadata) isn't updated properly:

for a second or so the correct title of the stream is displayed but then a metarefresh reverts the metadata to the first item of the (previous) cuesheet.

btw. what about making it configurable if a "skip" command skips to the next item in the cuesheet or the next item in the playlist?

_imho_ this part is loosely broken: i just got started to implement ending track at exact position when they excluded gstreamer from distro.

the foobar2000 is develped by 1 person from scratch, and amarok uses 3rd party libs like xine with poor interface.

i'd rather begin looking into kbabel code then trying to work on such mess

i converted all ape.cue images to mp3 (usin foobar), burned images to dvd and deleted them from hd

Replying to comment #25 from Simon Bühler:
The cue sheet metadata problem is reported in bug #127683.

A nice little improvement would be the display of the starttime of each track next to it's title in the display of a cue file. Nothing important or necessary but i could see the rest time of the current track.
It is know anyway to seek to the corresponding position. 

Created attachment 19146 [details]
Shows the length of the CUE track in Context Browser

@Simon's request, this patch makes the Cue File table in Conext Browser showing
the length of each track. This is pretty useful and simple hack, although not
even close to what was asked originally by Ian P. Christian. Please note that
this will only work with http://bugs.kde.org/show_bug.cgi?id=139707 applied.

SVN commit 620706 by aoliveira:

Show the length of the CUE tracks in Context Browser.
Patch by Dawid Wróbel <dawid@klej.net>
CCBUG: 112477


 M  +1 -0      contextbrowser.cpp  


--- trunk/extragear/multimedia/amarok/src/contextbrowser.cpp #620705:620706
@@ -2339,6 +2339,7 @@
                     + i18n("&#xa0;&#8211; ") +
                     "</span>\n"
                     "<span class='album-song-title'>\n" + escapeHTML( it.data().getArtist() ) + "</span>\n"
+                    "<span class='album-song-time'>&nbsp(" + MetaBundle::prettyTime( it.data().getLength()/1000, false ) + ")</span>\n"
                     "</a>\n"
                     "</td>\n"
                     ""

Looks like there's a semi-colon missing at the end of the   entity in that patch.

SVN commit 620771 by markey:

Fix missing semi colon after nbsp entity. Thanks Ted Percival.

CCBUG: 112477


 M  +1 -1      contextbrowser.cpp  


--- trunk/extragear/multimedia/amarok/src/contextbrowser.cpp #620770:620771
@@ -2339,7 +2339,7 @@
                     + i18n(" – ") +
                     "</span>\n"
                     "<span class='album-song-title'>\n" + escapeHTML( it.data().getArtist() ) + "</span>\n"
-                    "<span class='album-song-time'>&nbsp(" + MetaBundle::prettyTime( it.data().getLength()/1000, false ) + ")</span>\n"
+                    "<span class='album-song-time'>&nbsp;(" + MetaBundle::prettyTime( it.data().getLength()/1000, false ) + ")</span>\n"
                     "</a>\n"
                     "</td>\n"
                     ""

Is there any progress being made on Ian's original request: to have tracks of a .cue file individually added to the collection for searching, etc?

Nope. At this point we're *really* busy with getting the foundations of Amarok2 right. No time for new features :)

*** Bug 144469 has been marked as a duplicate of this bug. ***

*** Bug 112805 has been marked as a duplicate of this bug. ***

To save a request i made in http://bugs.kde.org/show_bug.cgi?id=144469 I just wanted to add that when you play the whole CD, the preaudio gaps should be played normally, while they should be ignored when jumping to a single track or when continuing to play with a different song than the one that would normally be on the CD. Think those intermissions you have on mix CDs. Also, crossfading and any gaps should be switched off when playing tracks that follow each other on the original CD.

*** Bug 137918 has been marked as a duplicate of this bug. ***

*** Bug 132564 has been marked as a duplicate of this bug. ***

In light of the impeding KDE4 (and thus Amarok2) release, I am going through all bug reports I am involved in. To the best of my knowledge, this issue is still open.

yeah, i hope that with the new multimedia subsystem design (phonon) we'll be able to implement cue support in a nice way...

Oh, will Phonon take care of this, in the future? If yes, please reassign this bug. And thanks :)

Phonon has nothing to do with this.

Setting this to 2.0-SVN for house-keeping.

Taking this one over. I plan to complete this for Amarok 2.1.

as it doesn't work in 2.1 (check bug 187587), a chance that cue sheets support will be done in 2.1.1 ?

This bug is old and all the patches are based on Amarok 1.4

See the bug 187587 which is up to date.