Hi there
Recently, I was thinking about if it was possible to replace identical songs on different albums in one's collection by hard- or softlinks that point to one common file instead of having the same song several times as different files. I was wondering about the following things and hoping you could give me some answers:

1. the AFT feature of amarok generates a unique ID for each song in your collection. This ID would be determined by the contents of the song, I suppose it is an md5 sum or something like that, right? That would mean that the unique ID would not be affected by the file name or its location? (If this is a stupid and obvious question please forgive me, I have not yet informed myself about the whole md5 thing very well)
2. let's say I delete one song and replace it by a link that points to another song which is identical. Would the statistics of the link become the same as the file that it's pointing to, i.e. would they get the same unique ID (maybe only after a rescan or a 'calculate new ID' request using DCOP, but eventually)? Also if I create new links to an already existing file.

thanks

interesting question - you might get a more indepth reply posting this to the mailing list, as I'm not sure the author of the AFT stuff reads this forum.

address is amarok@kde.org

I sent a mail amarok@kde.org but unfortunately never got an answer. Maybe they don't like mails from people who are not members of their mailing list.

It's good form to set a "reply-to" address for emails to mailing lists, so that when people respond they don't have to remember to CC: you in any replies. Mostly, people remember to do so anyway, but not always.
To check whether there have been any responses, look at the mailing list archive:
http://mail.kde.org/pipermail/amarok/20 ... 05851.html

edit: Also, check the May archives for more recent responses.

Thanks for your advice, there actually have been quite a few answers to my post, but they didn't make it to my mail address.
The result of a little discussion was, that the idea of marking songs as duplicates is now encouraged to be thought about by anyone who wishes to spend time doing so. If I had time I would do so myself (I wanted to learn Ruby anyway), but unfortunately I am very busy at the moment.