On Mon, Jan 14, 2008 at 10:09:18AM +0100, Gerfried Fuchs <[EMAIL PROTECTED]> was heard to say: > On Fri, Feb 02, 2007 at 05:38:07PM -0800, Daniel Burrows wrote: > > On Tue, Jan 09, 2007 at 04:35:36PM +0200, Eddy Petrişor <[EMAIL PROTECTED]> > > was heard to say: > > > Some of the debtags categories shown in aptitude are duplicated (and > > > suffixed with a curly brace). This was not the case in older releases > > > (not sure when this changed and I am not sure if is releated/caused by > > > debtags upgrades and dropped dependencies). > > > > This is not a regression in the aptitude code; instead, it's due to > > the fact that the debtags people decided to change the format of their > > tags (I'm not sure when, but this is the first time I've noticed it). > > That's nice to know - but now even almost a year after your response > this is still unfixed. :/ Is there severe problems with debtags format > parsing? Have you tried to talk with Enrico about it? I don't think that > ignoring the fact that aptitude is misparsing the tags is something that > should be ignored. Either the format should be reverted to something > parseable or the parsing should get fixed (with maybe even disabling it > and just displaying the tag line verbatim).
All my discussions with Enrico in the past have indicated that the right thing to do is to use libdebtag/libapt-front/libept, which will keep up with the debtags format as it changes and also not suck speed-wise. I added support for libept to the source tree recently, but I haven't decided yet whether to enable support in the next Debian upload. There are a few ugly things about this, like that apparently it forces the user to run "debtags update" manually, rather than just doing apt updates as usual. Daniel
