On Sat, Mar 26, 2016 at 12:11 PM, Andrea Pescetti <pesce...@apache.org> wrote: > On 22/03/2016 Damjan Jovanovic wrote: >> >> For the past month I've been working on preparing the gbuild branch >> for merge into trunk, and I believe it to be essentially ready. > > > Thank you Damjan, this is one of the best news about project development > we've had in recent months! And I won't insist again that these news deserve > a blog post... > >> My work can be found in branches/gbuild-reintegration > > > I've just built it correctly on one of my environments (Linux64). OK to > merge it from my side, with the aim of ironing out any further issues by > 4.2.0. For the record, a parallel build (-P4) worked great too. > >> * Fixes bugs 117685, 117218, 117340, 116880, 116755, 117350, 117340, >> 117610, 116959, 117845, 117687, and 117171. Note how at least 116755 >> is a user-visible bug in OOXML parsing. > > > This is really impressive. I've taken a look all of them (117340 is listed > twice, I don't know if you wanted to list something else). May I know why > this branch development fixes > https://bz.apache.org/ooo/show_bug.cgi?id=116880 and > https://bz.apache.org/ooo/show_bug.cgi?id=116755 ? I mean, you fixed them > "while at it" or are they side effects of the build system changes? I am > actually interested in the opposite situation, i.e., can build system > changes reach so deep within the user experience? Or, in other words, can > they introduce new bugs at a level this deep?
Thank you. Yes several are repeated, and no I didn't want to list something else. Those bug fixes were present in the original gbuild branch - they're not something I developed. I don't know why. Build changes can't really break code. There were some code changes, for example controlling function visibility using code instead of using .map files, but if that broke something, linking would fail and the build would presumably fail; only run-time dynamic linking can break at run-time. > I don't get why > https://bz.apache.org/ooo/show_bug.cgi?id=117171 > is addressed too. Me neither. >> I don't have the ability to test Mac OS and it would be appreciated if >> someone could or remote access could be given to me so I can. > > > You received access to an ASF-owned Mac back in October 2015. That Mac was > then taken over by Infra and transformed in an ASF-wide service to optimize > resources. You can surely get access to it again if it doesn't work for you. > I can reactivate the discussion with Infra, just ask in case. Please do. I was told I received access back then, but AFAIK never given its IP address, login credentials or protocol to access it. >> Can I merge branches/gbuild-reintegration to trunk now? > > > +1 from me. Marcus' concern about merging trunk into it beforehand was > correct, but the set of unrelated changes is minimal and under control in > any merge strategy. Let's get Mac working first, then see about a merge. > Regards, > Andrea. Regards Damjan --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org For additional commands, e-mail: dev-h...@openoffice.apache.org
