Having trouble with sourcetree-bitbucket-git, and charactersets.
Default in sourcetree is utf-8, under "repo settings"
if I try to commit a file that is windows-1252 format (latin1, iso-8859-1 etc), which contains norwegian characters (æ, ø, å) then in the view where you select "hunk" to stage:
image2016-8-16 11:9:30.png
those black dots should be å, and I get this error message:
---------------------------
Error encountered
---------------------------
'git apply' failed with code 1: Checking patch xxx/file.sql...
error: while searching for:
xxx uniqueidentifier,
xx uniqueidentifier,
xxx int NULL,
DRGGrupperingAvslutningsm�te varchar(1) NOT NULL,
xxx int NULL,
xxx bit NOT NULL,
xxx bit NOT NULL,
error: patch failed: xxx/file.sql:15
error: xxx/file.sql: patch does not apply
See C:\Users\xxxx\AppData\Local\Atlassian\SourceTree\LastFailedPatch.txt for patch details (D:\xxx)
---------------------------
OK
---------------------------
If I change to one of the other charactersets in the sourcetree config, then it looks ok in the window:
image2016-8-16 11:12:31.png
but I still get an error message, but now it seems like sourcetree is toying around with utf-16?
image2016-8-16 11:13:13.png
(you get two garbled characters where it should be one, typical for utf-16 rubbish)
If I convert the file to utf-8, then I get lots more errors on the other lines with text in the file that have these characters, so that is obviously no-go.
and now it starts to garble other text in sourcetree gui also, in the list of staged files:
image2016-8-16 11:15:31.png
the the tekst there should be NpkKjører, and in the commit log list it also does the same:
image2016-8-16 11:16:41.png
should read: "prøver å fikse.."
What on earth is going on here?
I had no problem uploading the whole repository with sourcetree, now It's impossible to commit code ?
Running the latest version, 1.9.6.1
Tested with system git, and embedded git, no difference.
