From: Dudy K. <ko...@gm...> - 2009-04-17 22:28:00
|
On Sat, Apr 18, 2009 at 1:10 AM, Todd Zullinger <tm...@po...> wrote: > Dudy Kohen wrote: > > It doesn't need the taglib sources, just the library (and some of > > the includes, but they are all in the dev package of taglib)... > > Cool. That's good to know. > > > The question is not "Can it be done?" it can be, I believe it's even > > very simple, I never installed taglib from sources, just the > > bindings... The question is "should we do it?" > > I'd say that depends on how easily we can add it to the build system > weighed against the benefit of adding taglib support now. Is the main > benefit that taglib is better maintained than libid3tag? > > I thought I saw a comment in the code or on the list about USLT tags > still not being liked by the iPod? I think that may well be an iPod > bug, one that only seems to be tickled when using id3v2.4 tags? I've > got plenty of id3v2.3 tags that my iPod reads just fine. :/ iPod needs a null terminator, as far as I understand, the specs mention it as well, but no answer has been received from the mailing list in id3.org. It's not a big deal, if it doesn't exist it simply displays the next frame code. It has nothing to do with tag version, yet it turns out some taggers add it. > > > I started to try and get the bindings building as part of the gtkpod > build but I got sidetracked. I was thinking that the tag_c.* files > could go into a taglib subdir to be built if we find that the system > taglib is too old. Then we can link against our local copies. The > autoconf logic would go something like this: > > if pkg-config taglib_c >= 1.6 or contains the required functions > use system taglib > else > if pkg-config taglib > build and use local bindings > fi > fi > if no taglib found or buildable > use libid3tag > fi > I like this. > > And maybe in the beginning we'd be better off making taglib require > explicit enabling via a --with-taglib option. That way unsuspecting > users shouldn't notice any change at all. More adventurous folks can > enable taglib and help shake out any bugs. When it's been tested and > solid, we can make taglib the default (and remove libid3tag sometime > after). Sound like a plan? > Since taglib's main advantage is better tag write support and I think that 3200 test cases is more than the avarage, I think we should make it default, it's SVN, people are adventurous to try it... If the default will be without it, I would prefer to leave the warning. > > -- > Todd OpenPGP -> KeyID: 0xBEAF0CE3 | URL: www.pobox.com/~tmz/pgp<http://www.pobox.com/%7Etmz/pgp> > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ > It is not enough to be busy, so are the ants. The question is: what > are we busy about? > -- Henry David Thoreau > > > > ------------------------------------------------------------------------------ > Stay on top of everything new and different, both inside and > around Java (TM) technology - register by April 22, and save > $200 on the JavaOne (SM) conference, June 2-5, 2009, San Francisco. > 300 plus technical and hands-on sessions. Register today. > Use priority code J9JMT32. http://p.sf.net/sfu/p > _______________________________________________ > Gtkpod-devel mailing list > Gtk...@li... > https://lists.sourceforge.net/lists/listinfo/gtkpod-devel > > |