Comment 3 for bug 34667

Sorry for the poorly worded bug report. Here's a step-by-step way to
reproduce the issue.

Convert a filename to iso-8859-1. Archive it. Extract it.

This will simulate the scenario where a user archives an iso-8859-1 file,
sends it to their friend, who is using a utf-8 encoded filesystem. The
friend extracts the file and is unable to view the filename correctly.

The ubuntu utf8-migration-tool is supposed to fix this problem, but it is a
one-use tool, and does not fix filenames that a user comes across during
their everyday work.

------------------------------------
Add the universe repository:
http://easylinux.info/wiki/Ubuntu#How_to_add_extra_repositories

sudo apt-get install convmv
cd ~/Desktop
mkdir testdir
cd testdir
gedit björk.txt

Save the file, and close gedit.

convmv -f utf-8 -t iso-8859-1 -r --notest björk.txt
cd ..
tar -czf fûnkyarchive.tar.gz testdir
convmv -f utf-8 -t iso-8859-1 -r --notest fûnkyarchive.tar.gz
rm -r testdir
------------------------------------

Use Nautilus and File-Roller to extract the file. The result is that the
archive and the archived file both have iso-8859-1 encoded filenames, which
are unviewable in the default ubuntu utf-8 encoded filesystem, and must be
manually converted to view correctly.

File-roller and Nautilus should be able to view alternate encoded filenames
and convert extracted filenames on the fly. Note that Rox-Filer is currently
able to view multiple filename encodings.

I've created a Nautilus-script that performs the conversion, but this is a
cumbersome tool for most users.

Thank you for looking into this!

Adam Kane

On 4/6/06, KarlGoetz <email address hidden> wrote:
>
> I was unable to reproduce the problem with a test archives i created.
> tar.bz2 and tar.gz (file inside called 07 - Zerstören.mp3) and the last
> had a file called "björk.ogg" inside.
> I'm not sure what my encodings are, so i cant be sure if the test is
> relevent or not.
> kk
>
> --
> ISO-8859-1 Support Missing
> https://launchpad.net/malone/bugs/34667
>