Ar an seachtú lá de mí Bealtaine, scríobh It's me FKtPp ;): > ================================================================ > Dear Bug Team! > > I find that dired buffer decode setting will get confused whenever I > finished reading news with gnus. > > Step to reproduce this bug: > > 1. open a dired buffer on directory contains some file with Chinese in > their filename. > => Chinese characters get properly displayed (chinese-gb2312) in the dired buffer > > 2. kill the buffer, and run gnus by running `M-x gnus <RET>' > 3. quit gnus by typing `q' > 4. re-open dired on the same direcotry (with those chinese-files) > => Chinese characters get garbled (displayed in latin-iso8859-1) > > Maybe you can also reproduce this by using some file with > japanese/korean characters in their filename. > > ================================================================ > > [...] > > Coding system aliases: > > 'native is aliased to utf-8 > 'file-name is aliased to binary > 'mswindows-multibyte-system-default is not a coding system alias file-name should be an alias of utf-8, not of binary. Looks like Gnus is trashing it, or the related variable file-name-coding-system, but a quick grep doesn't point me to where--it's always bound, not set. Can you do a M-x gnus-bug RET, and CC xemacs-beta? The version info there would seem to be relevant.