Naoki INADA recently provided nice patch for QBzr which added handy encoding selector to QBzr windows where content of user files is shown, including diff viewer (qdiff command), annotation viewer (qannotate command) and file content viewer (qcat and qviewer commands). This feature is in the trunk now and will be released with QBzr 0.17.
This is very useful feature for all who works with non-ascii documents. By default QBzr assumes that encoding of your document is UTF-8, and this is safe default for Linux users. Unfortunately on Windows documents typically created in ANSI encoding (e.g. CP1251 for Russian, CP1252 for Western Europe, and so on). So having the way to specify encoding to see your actual text instead of $%^&#@&* garbage is very important.
So now you can control the encoding not only from command-line (option --encoding) but from GUI as well. Encoding selector(s) available at the bottom of dialogs as combobox. Such encoding selector is very important for users of TortoiseBzr and Bazaar-Explorer, because they’re using QBzr dialogs from corresponding GUI applications.
Selected encoding is remembered and saved in branch configuration file branch.conf as encoding = xxx.
Hint: if you most of the time working with files of one encoding, you can put the encoding setting in main configuration file bazaar.conf:
email = Joe Random <email@example.com>
encoding = cp1251
And this settings will be used by default for all branches where no encoding specified in command-line or branch.conf.