PATCH: Taglib 1.3.1 TAG extended
wheeler at kde.org
Sat Jan 22 14:58:47 CET 2005
On Saturday 08 January 2005 11:48, you wrote:
> > That said, as this value isn't standard with any tag type except ID3v2 it
> > doesn't make sense to have it in the generic interface. That class isn't
> > simply a convenience for TagLib users, it's meant to represent a
> > reasonable set of things that developers can expect to exist in most
> > tags.
> But thats the point. They DO exist in the tags AND are in the standard!
> That was the first thing I did. I read the standards of APE and
> Xiphcomment. ID3V2 supports it
> APE supports it
> XiphComment supports it.
I'm not sure which standard you're referring to -- I'm not saying that it's
illegal to store in the way that you've done, I'm just saying that it's not
something that extends beyond TagLib.
> The only thing with XiphComment is that there is not yet a standard but
> only proposals. So to be 100% standards conformant we mustn't even use
> TITLENAME= in Xiphcomment since its only proposed and not defined. Ok...
> we could add MOD support to taglib and then strip ALL but Titlename to
> be as generic as possible....
"Below is a proposed, minimal list of standard field names with a description
of intended use."
It then lists many fields. "MEDIA" is not among them. That's what I'm
> I don't want to be offensive. I just saw that TagLibs "smallest set in
> common" is even smaller than it had to be. Please tell me what you think
> of it. (I will subscribe to taglib-devel to keep track of it)
I've toyed around with some schemes for making the Tag interface more
extensible from a users perspective in 2.0, but fundamentally the abstract
interface is meant to be simple. It presently supports pretty much only
what's supported -- officially -- in every format. I'm not saying that such
won't ever change in the future, but that's how it got the way that it is and
I don't really think that was a mistake...
For a successful technology, reality must take precedence over public
relations, for nature cannot be fooled.
More information about the taglib-devel