On Sun, May 20, 2012 at 03:42:31PM +0100, Mark Brown wrote: > Dear Whatever,
Hello Mark, (Just using the snippet from reportbug, what do you prefer as address?) > If you're sending stuff like this please just send a patch that can > actually be applied to the package, it's much easier than having to faff > around reassembling a set of changes distributed in several different > forms. I thought it would be better if you have an overview of the proposed changes in debian/rules - and btw. piping the mail to `patch -p1` works just fine. On Sun, May 20, 2012 at 04:05:44PM +0100, Mark Brown wrote: > No, this makes no sense. What makes no sense? >> Description: Use build flags from environment (dpkg-buildflags). >> Necessary for hardening flags. >> >> example$(EXE): example.o $(STATICLIB) >> - $(CC) $(CFLAGS) -o $@ example.o $(TEST_LDFLAGS) >> + $(CC) $(CFLAGS) $(LDFLAGS) -o $@ example.o $(TEST_LDFLAGS) > > Why on earth are you patching Makefile.in for this? There's already a > perfectly good way of configuring all this stuff provided upstream which > you're just ignoring here for no reason. Because there is no consistent way to apply the LDFLAGS, it's necessary to pass them to TEST_LDFLAGS and SHAREDLIBV so I thought it's better to just use the normal LDFLAGS to pass those flags. Just modify the patch if you don't like my approach. But as always patches are just suggestions, you can either use them or write a better fix. Regards, Simon -- + privacy is necessary + using gnupg http://gnupg.org + public key id: 0x92FEFDB7E44C32F9
pgpImnQkpwoBx.pgp
Description: PGP signature
