В сообщении от 17 января 2009 Detlef Riekenberg написал(a):
On Sa, 2009-01-17 at 14:24 +0300, Vitaly Lipatov wrote:
В сообщении от 16 января 2009 Detlef Riekenberg написал(a):
On Fr, 2009-01-16 at 20:51 +0300, Vitaly Lipatov wrote:
version=`(GIT_DIR=$(TOPSRCDIR)/.git git describe HEAD2>/dev/null || echo "wine-@PACKAGE_VERSION@") | sed -n -e '$$s/(.* )/const char wine_build[] = "\1";/p'` && (echo $$version | cmp -s - $@) || echo $$version >$@ || ($(RM) $@ && exit 1)
version=`(GIT_DIR=$(TOPSRCDIR)/.git git describe HEAD2>/dev/null || echo "@PACKAGE_NAME@ @PACKAGE_VERSION@")
Your patch change also the minus to a space
Wine 1.1.13 looks more correctly than wine-1.1.13, or not?
No. Even Wine-1.1.13 is dangerous, because unix is case dependent.
Try "git describe HEAD" for reference The echo is used only as fallback, when "git describe HEAD" failed
Well, you are accepted echo is used only as fallback.
You change the output of wine --version.
- A very high risk, that you break scripts, because of the space
- A high risk, that you break scripts, because of the different case.
What script can check wine version for wine if this name is not changed for ages? Are you know an example of such script? I know only one wine related script, it is winetricks :)
You change the output of winetest
- Did you checked, that the parse script on winehq still works?
I guess parse scripts works with wine built from git, and echo fallback is not used in any case.
For changing "Wine" to wine", you can pipe through 'tr "W" "w"', when every supported system has a working "tr".
PACKAGE_NAME can be some another name, no Wine nor wine.