[Rhythmbox-devel] Some ideas for work with iPod, last.fm and non-utf encodings

[Pavel Shuvalov <shuvalov pavel gmail com>]

Hi,

I am user of Rhythmbox and i use it for managing tracks on my iPod. I 
have some ideas for make better work with iPod and tags encodings, and 
very interested in your opinion about this ideas. Also, i am a student, 
and i am planning to apply application for GSoC if somebody says that 
this ideas are needed for more then one man(and they actual and not 
implemented yet). So, ideas:

1.Almost all linux players support last.fm — Amarok, Banshee, Rhytmbox, 
xmms. But i listening a lot of music, using my iPod, and want to 
scrobble it too. In linux there are some small scripts/GUI programs for 
scrobbling played music to last.fm(i used lastpod written in Java). But 
it is very inconvenient — using gtkpod(sometimes i use gtkpod, the 
reason is described in chapter 2) for uploading music, lastpod for 
scrobbling and some another player for listening. So, i propose 
implement function for scrobbling listened tracks from iPod to last.fm. 
In UI it can be simple item in context menu for iPod icon. This can be 
done as new plugin, or improvements in existing module, i think.

2.Some time ago i faced problems in copying music in flac format from my 
library to iPod. I think, that on-fly encoding(using lame or sox) is a 
great feature for Rhytmbox.

3.In Russia there are some common historical problems in encodings (i am 
about this: http://en.wikipedia.org/wiki/Mojibake). This problems are 
especially acute, when Linux and ID3Tags are encountered. Widely 
distributed Windows program Winamp saves tags in CP1251 encoding, but in 
standard it must be in UTF8. This difference in standards interpretation 
makes scrobbling a bit problematic. Overriding default encoding for id3 
tags was discussed some time ago, but i want propose things, that i used 
in iTunes. In iTunes you can select item in track context menu named 
«Convert to UTF», and iTunes will convert tags in UTF8. In my current 
job i take part in creating some algorithms for tags encoding 
recognition for russian symbols — and it helps us to normalize tags in 
numerous tracks to one encoding. Btw, as far as i know,in Amarok you 
could specify base(not UTF8) encoding, but this feature was removed in 
versions >= 1.4. I think, that Mojibake in tags is not only russian 
problem, and it is possible to introduce some plugin, that will help 
users to normalize encodings of there files.

Thanks for attention,
Shuvalov Pavel.