Dear Paul, this message from April 2020 didn't have any reaction from you, or any other gitk maintainer. Is this intentional? I thought translation contributions would be regarded as helpful improvements, so if possible, please check whether you could merge this in? Original message: I think my email program has messed up the encoding of the characters in the inline patches I sent to you. I would kindly ask whether you could accept the patches as attached files in this exception case. Attached you find them. In the patch files, the Umlaut characters are two bytes respectively, i.e in utf-8 encoding, just as the files in the repository. I'm pretty sure my email program had the grandious idea of encoding that into latin-1 which means one byte per Umlaut, effectively breaking the patch file. I hope the attached files are fine, otherwise we would have to come up with other means of getting my patches from my local git repo to your computer - maybe a github clone or something similar? I just try the attached files first. Thanks! Best Regards, Christian Am Montag, 13. April 2020, 09:34:56 CEST schrieb Paul Mackerras: > On Mon, Feb 10, 2020 at 10:57:05PM +0100, Christian Stimming wrote: > > So that subsequent translation updates have an up-to-date > > starting point. > > > > Signed-off-by: Christian Stimming > > I tried to apply this and the following patch, but this patch had a > > lot of rejects. For example, in this hunk: > > @@ -21,11 +21,11 @@ msgstr "" > > > > msgid "Couldn't get list of unmerged files:" > > msgstr "Liste der nicht zusammengeführten Dateien nicht gefunden:" > > the u with umlaut appears as a single 0xfc byte, but in the file > po/de.po, it appears as a sequence of two bytes: 0xc3 0xbc, and so > patch considers that the lines don't match. I have my locale set to > en_AU.UTF8. Any ideas why the context in your patch would be > different to what I see in the checked-out file? > > Paul.