Hi, Ben,

On Wed, 01 Dec 1999, you wrote:
> >>>>> "Scott" == Scott Thomas Haug <[EMAIL PROTECTED]> writes:
>     Scott> The distribution is about 500k in size.  The actual source
>     Scott> code for the library is little under 300k uncompressed.
>     Scott> I'm not sure if this would be viable to include in the
>     Scott> freeamp source and change freeamp to use it, but I'd be
>     Scott> happy to help make it work.
> Wow. :) I don't know if this says that id3v2 is hugely bloated,
> C++ is hugely bloated, or what:
> [ben@gilgamesh:~/src/python/ID3]% ls -l ID3.py                          10:51PM
> -rwxr-xr-x   1 ben      users        7814 Nov 19 14:28 ID3.py
> That's more like it ;)

If ID3.py only handles id3v1 tags, that makes a lot of sense.  If it handles
id3v2 tags as well, I'd love to see that code!  id3v2 /is/ pretty, ahem,
"feature-rich", which is probably one of the reasons it never caught on as much
as it did (many complain it is "over designed" and too complicated to
implement easily).  Granted, the 300k id3lib source code is now mostly comments
and doc++ documentation (only 103k after running it through mangle), but that's
still a far cry from less than 8k.

However, one could easily argue that id3v2 is big because it handles pretty
much any kind of meta-data you'd want to throw at it, whereas id3v1 is pretty
inadequate for handling anything beyond the most basic (and short-lengthed) of
meta information.  Which is why I'm interested in hearing about the xml-based
tagging mechanism that ruaok mentioned a few weeks prior.  There's got to be a
better way...


PS C++ might also be bloated, but the id3lib source doesn't make much use of
advanced C++ features beyond classes, exception handling, and just a little bit
of inheritance.  Otherwise, it's mostly C in C++ clothing.

> Ben
> -- 
> Brought to you by the letters P and G and the number 5.
> "Porcoga daisuki!"
> Debian GNU/Linux maintainer of Gimp and GTK+ -- http://www.debian.org/
> _______________________________________________
> http://www.freeamp.org/mailman/listinfo/freeamp-dev


Reply via email to