Open David-UPCT opened 10 years ago
OK, I'll try to isolate the issue with a new test, then I'll try it out using the main mutagen code. Will let you know if I did anything interesting (but probably won't be for a few days yet).
Could you email me a copy of the file and the original image (ben.sput@gmail.com)? Thanks for reporting!
Email sent. Thanks.
Hi. I'm not sure if the problem came from your mutagen version, so I'll tell you the full story.
I use archlinux with rhythmbox and the plugin rhythmbox-coverart-browser, which uses your package (named as python-mutagenx in AUR repository). All my files are ogg with coverart embedded. I used to insert the cover with easytag, but now I use kid3-qt instead of easytag.
Everything was fine up to python-mutagenx 1.22.1, I attach a picture. You can see 3 images of the cover (the one with a big star), a small one in up-left, and another two in the right.
However in 1.23 and 1.24 I can't see the covers generated with kid3 in rhythmbox (note that easytag's are OK). The first time that I open a new ogg file with embedded coverart inserted with kid3 only one of the 3 avaliable images appears (the small one in the upper-left part of rhythmbox is sligthly visible). I attach a picture.
Even worse, after the first time, each time I open rhythmbox even the middle image is gone:
The first time that rhythmbox read a embedded coverart file it saves the embedded image in a cache directory. I post a link with two of these 'cached images' generated by rhythmbox, 00000034 was automatically created when python-mutagenx version is 1.22.1, and the one with 1.24.1 version is named 000000c8. None of the files has extension, but my image viewer can read '34' without problem, but it doesn't recognize 'c8' as an image file. Both should be the same file, but the bad image is only 88kb and the good image is 202kb . I think that's the origin of the problem.
http://www.speedyshare.com/Bt9y5/download/00000034 http://www.speedyshare.com/7VR7k/download/000000c8
I know it's a weird bug, and maybe it's not related with mutagen, but downgrading to python-mutagenx to 1.22.1 solve the problem. Any idea? Thanks