[ID3 Dev] 'Extending' ID3 V2.4

[Ion Todirel]

I don't think that dependency is a problem, why? Because many (very) people are using .NET and J2EE platforms (the most), here you have xml support. Others problems we can solve.
  
Scott Wheeler <wheeler at kde.org> wrote:
  On Friday 10 February 2006 23:32, Dan O'Neill wrote:
> Anyone want to start a draft set of requirements for XML tags?
>   - Backwards compatibility w/existing tags
>   - Adoptable
>   - Must have a reference implementation for writing and reading
>   - ?

I didn't actually send the responce I typed to the XML tags bit because I 
didn't take it seriously, but well, since it seems that you're on that track:

There are many problems with using XML as a generalized tagging format. The 
largest three are (I'm not anti-XML; it certainly has many useful 
applications, but this is in my opinion clearly not one of them.):

- This would require implementers to include an XML parsing engine. For all 
of the complaints here about the complexity of ID3v2, XML is much harder to 
parse (but of course there are existing parsers, but then you face more 
external dependencies).

- XML is hierarchical rather than linear. All tagging formats that I know of 
can be parsed linearly treating the tag as a bitstream. Maintaining a DOM 
tree consumes not-insignificant amounts of memory and processing. While this 
would just be an annoyance on the desktop, currently my EUR 60 CD player can 
read ID3v2 and I assure you that it does not have an abundance of RAM. Also 
the overhead in processing speed becomes significant even on the desktop when 
you consider working with tens of thousands of files.

- XML is larger. This isn't a problem in terms of storage; you can write 
pretty huge tags before they're significant relative to the audio content -- 
but again, thinking deeper and of implementations is important. If you're 
indexing an MP3 collection you're likely trying to minimize disk access. If 
the average tag size went from 2 kb to 10 kb (I'd assume larger padding with 
XML tags), disk reads being the major bottleneck in music collection 
indexing, you're probably effectively making indexing 5x slower (Ok, probably 
a bit less than that since disk buffers are probably working on pages larger 
than a kilobyte, but you get the idea.).

Cheers,

-Scott

-- 
If the answer is "more lawyers" then the question shouldn't have been asked.

---------------------------------------------------------------------
To unsubscribe, e-mail: id3v2-unsubscribe at id3.org
For additional commands, e-mail: id3v2-help at id3.org



		
---------------------------------
Yahoo! Photos – NEW, now offering a quality print service from just 8p a photo.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.id3.org/pipermail/id3v2/attachments/20060211/7f50a41a/attachment.html>