Yammi switched to taglib

Oliver Noelle oli.noelle at web.de
Sat Jan 22 00:49:51 CET 2005

Hi Scott and others,

I'm the maintainer of the yammi project (http://yammi.sourceforge.net), and I 
finally switched to taglib for reading and writing tags.

First, thanks a lot for taglib and a really clean and easy API (just how you 
would expect it) that made switching to taglib easier than expected. If you 
want, you can add yammi to the list of projects using taglib.

Two suggestions for refinements in the docs:

1. class docs for FileRef: "Also note that it is probably a good idea to plug 
this into your mime type system rather than using the constructor that 
accepts a file name."
=> could you give a reason or example why it is better to use the mime types?
(eg. if I would rename an mp3 file to have an ogg suffix, would the 
constructor fail whereas the mime type recognition would still recognize the 
file as an mp3 file?) ...this was just not clear to me...

2. FileRef::isNull (or other methods?)
could you add some more details what happens if taglib cannot handle the 
file(type)? Eg. if I let taglib try to handle a "readme.txt" file, after 
constructing a FileRef, does the isNull() method return true?
What about broken mp3 or ogg files, where exactly in using taglib do I see 
which information could NOT be retrieved from the file (by checking 
tag()->isNull() and audioProperties()->isNull() ? or does this result in a 
FileRef->isNull() ???)
Some more info about error behaviour could be useful in handling files that 
are not correct (according to the specs) or not supported by taglib.

But as I said before:

Big thanks for a library with a very easy API and good documentation, I should 
have switched to it a lot earlier!



Concerning the future of yammi (in case you are interested):
As I don't have a lot time for further development of yammi, and as it is 
basically a one-man-project (with some help and patches from outside) with a 
grown codebase which gets more and more difficult to maintain, I plan to 
finish a stable version 1.2 of it, and only release another one or two bugfix 
versions of it in the near future without introducing a lot new features...

But as I am still interested in digital music management, I was considering 
contributing to juk or amarok (eg. by integrating yammi's fuzzy search into 
these projects). To be honest, I did not get used to the user interface of 
both projects so far (maybe I am too much used to yammi...), but amarok is my 
favourite at the moment. So I might offer a patch for amarok in the near 
future integrating the fuzzy search into that project.

...just to let you know what my plans are... Good luck with taglib and juk!

More information about the taglib-devel mailing list