On 21/11/2012, at 9:32 AM, James Peach <jpe...@apache.org> wrote: > On 21/11/2012, at 9:27 AM, Igor Galić <i.ga...@brainsware.org> wrote: > >> >> Hey folks, >> >> regarding this MASSIVE change: >> >>> TS-1582 space between string and format specifiers >>> >>> GCC 4.7 (and newer versions of clang) REQUIRE a space between >>> the format the string and a format specifier (a # defined "string") >>> in an string concatenation to be C++11 conformant. >> [snip] >>> ++++++++++---------- >>> 35 files changed, 126 insertions(+), 126 deletions(-) >>> ---------------------------------------------------------------------- >> >> Which essentially boils down to: >> >> sed s|%"PRId64"|%" PRId64"|g >> >> Because that's what the C++11 spec says, and that's what GCC 4.7, >> and clang now, thanks to -std=c++11 enforce. TS-1582 is only the >> first sign of this breakage, and TS-1853 is what follows. >> >> Zhao is quite unhappy with such a massive change. Understandably >> because he is still working, in a branch, on SSD optimizations. >> >> This change makes merging "upstream" changes for him much more >> difficult. Before I revert this commit again, I'd like to >> bring this to attention, because: >> >> trunk is CTR, and I'd like to keep it that way. >> >> There's a number of options we have from here, and I'd really >> like we figure this out now. > > Obviously I did not test this on GCC 4.7 like I should have.
Oh, and when people (like me) commit breaking changes like this we should feel comfortable with just reverting them IMHO ... > If I had I would have realized that this change has more impact that I was > expecting. I'm ok with reverting these changes and rescheduling at a more > convenient time. I can commit a change to disable C++11 mode for now. > > >> >> i >> >> -- >> Igor Galić >> >> Tel: +43 (0) 664 886 22 883 >> Mail: i.ga...@brainsware.org >> URL: http://brainsware.org/ >> GPG: 6880 4155 74BD FD7C B515 2EA5 4B1D 9E08 A097 C9AE >> >
