On Sun, Mar 4, 2012 at 8:45 PM, Pete Batard <[email protected]> wrote:
> On 2012.03.04 22:53, Rocky Bernstein wrote: > >> Pete Batard's patches have now been merged into the git master branch. >> > > Wow, thanks! I've been trying to find time to get things going with the > integration branch this week, but that didn't materialize itself. > Sorry you had to go through the integration on your own. > > > Pete: it is possible there are some small residual changes I've missed. If >> that is the case send patches and we'll take care of. >> > > Well, the MSVC solution files and config are not there, but this can be > taken care of later on. Note that MSVC will require keeping a version.h in > the source tree, since we can't autogenerate one there. > > Also, we will very much need ecma_167.h to be patched further to bring it > in line with that currently exists in -pbatard, since, as previously > mentioned, we have a problem with union of empty arrays being one byte for > MS compilers vs 0 bytes for gcc, which throws any code that uses sizeof off > balance. Ok. Make a patch for that. > > > I've tested the changes using "make distcheck" on GNU/Linux, cygwin , >> Solaris and MinGW. >> >> On MinGW there seems to be a disagreement between what autoconf uses for >> filename paths (e.g. /c/cygwin...) and what fopen() of libc on MinGW >> understands (e.g. c:/cygwin). As a result there are a couple of failures >> here which I worked around by modifying config.status to insert c:/cygwin >> for /c/cygwin. >> > > That's because you didn't pick cdio_strdup_fixpath() from util.h/c / > _cdio_stdio.c / bincue.c / cdrdao.c. Ok. Definitely want to get this sooner as it will make *all* the tests work on MinGW. > I also have some very recent change in _cdio_stdio.c to support opening > images with a patch that contains extended characters. > Not sure if you synced recently, but I've been more or less continuously > applying patches as I picked them up in my app, and I see a few other > changes that you didn't pick. For instance, there were various problems > with Joliet support that I fixed recently. > I'll try to send patches for those when I get a chance. > Many thanks. > > On another note, I see there are still paranoia related files in > src/cd-paranoia/ as well as a test/testparanoia.c. Shouldn't these be > removed as well? > Yes. I've made another scan and hopefully have removed all remnants now in git. > Regards, > > /Pete > >
