Kenneth Goldman wrote:
-----Original Message----- From: Miro Hrončok mhroncok@redhat.com
On 16. 02. 23 14:50, Kenneth Goldman wrote:
Could not execute mockbuild: Could not query n-v-r of hello: 'utf-8' codec
can't decode byte 0x93 in position 259: invalid start byte
That indicates it's not actually UTF-8.
I can only report when I cut and paste the smart quotes from the web page to the .spec file (Firefox -> emacs), I get 0x93 and 0x94 and fedpkg reports that error.
So apparently Emacs saved the file in some Microsoft encoding, probably Windows code page 1252, but FedPKG tried to read it as UTF-8.
When I load the packaging tutorial I get UTF-8 from the web server, so it seems that something on your computer converted the text into a Microsoft encoding.
Programs should generally use the character encoding specified in your locale, but FedPKG might be hardcoded to use UTF-8, perhaps justified by the policy that Fedora spec files must be UTF-8 encoded: https://docs.fedoraproject.org/en-US/packaging-guidelines/#_spec_file_encodi...
If you're going to cooperate with the rest of us in the Fedora project, then you'll need to handle UTF-8 in spec files. Sooner or later you'll encounter some non-ASCII characters. You may need to tell Emacs to read and write spec files as UTF-8, or you may need to fix your locale. Run "locale" to check. Going by your email address, en_US.utf8 is probably the right locale for you.
Björn Persson
-
Björn Persson