Gapless mp3''s:-
Have you ever wondered how to solve this problem?

Easy!

Change the binary of LAME so that the Comments Tag holds the necessary information for the decoder to understand a track's beginning and ending (time wise) to compensate accurately for each tracks length. A NEW, special LAME decoder is all that is needed to then make sure all future decodes of whole albums are perfect with regards to being gapless (as long as the general mp3 tag contains the original CD's beginning and end time data.

The lenght is already stored in the LAME tag and LAME enc/dec would be gapless if the one that were supose to make it work had checked his work. I suppose the next LAME version will work correctly....
Also I understand that foobar found a way to make it gapless even though there is this bug.

edit: it seems I was too fast and that LAME actually is working correctly (the bug was caused by a frontend it seems).
So the lenght is stored in the LAME tag but only foobar reads it ATM.

The delay and padding values are stored in the Info tag.
But when decoding, Lame does not remove padding at the end. (you could consider that decoding with Lame is partly incomplete regarding features).

Several people seem to imply that the values are not stored properly. Is this a confirmed thing? (I mean a bug is possible)

Isn't this what the latest foobar2000 betas are doing?

BTW Jan - which frontend was reponsible for the bug?

Case told me the problem was using mpc and foo_clienc since mpc doesn't report the lenght accurate enough.
Both Peter and Case tested this way and therefore concluded that LAME was buggy which seems not to have been the case anyway.

I think that foobar does exactly as ment, removing also the silence at the end. The transitions are perfect, one can decode using diskwriter and see it for himself