Date: Wed, 11 Nov 2020 16:40:01 +0000 From: bugzilla-noreply@freebsd.org To: ports-bugs@FreeBSD.org Subject: [Bug 250740] mail/mailsync 12.2-RELEASE breaks package build Message-ID: <bug-250740-7788-z1Wd3C7cSA@https.bugs.freebsd.org/bugzilla/> In-Reply-To: <bug-250740-7788@https.bugs.freebsd.org/bugzilla/> References: <bug-250740-7788@https.bugs.freebsd.org/bugzilla/>
next in thread | previous in thread | raw e-mail | index | archive | help
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=3D250740 --- Comment #7 from Dimitry Andric <dim@FreeBSD.org> --- (In reply to Fernando Apestegu=C3=ADa from comment #4) First of all, the c-client software was clearly written before C++ existed, since it uses C++ keywords as struct member names, so that is why they have= to hack around it with #defining those keywords to something non-conflicting. However, those defines should *not* be active when including any system headers, and that is clearly violated by the c-client.h file. It is simply totally broken, and you were lucky that it worked at all. In any case, it is likely not about clang or clang++, but about the used li= bc++ headers. If you use C++, libc++ overrides certain standard C headers such as stdio.h, string.h, etc. Since the c-client headers are including multiple system headers with 'private' redefined, this wreaks havoc when it eventual= ly goes into libc++'s own headers. I guess g++ doesn't work that way, but I haven't investigated. In my opinion, c-client is obsolete software which is clearly incompatible = with C++, and should either be patched up to relatively non-ancient standards, or ditched altogether. Note that mailsync itself does not look much better. :) --=20 You are receiving this mail because: You are the assignee for the bug.=
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?bug-250740-7788-z1Wd3C7cSA>