From mboxrd@z Thu Jan 1 00:00:00 1970 From: Thomas Petazzoni Date: Thu, 23 Jul 2015 09:31:09 +0200 Subject: [Buildroot] [PATCH 1/1] package/boost: boost-filesystem needs wchar In-Reply-To: <1437635893.2784.4.camel@embedded.rocks> References: <1437142516-12378-1-git-send-email-joerg.krause@embedded.rocks> <20150722085507.4f78d81f@free-electrons.com> <1437635893.2784.4.camel@embedded.rocks> Message-ID: <20150723093109.38aa581e@free-electrons.com> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: buildroot@busybox.net Dear J?rg Krause, On Thu, 23 Jul 2015 09:18:13 +0200, J?rg Krause wrote: > There are a lot more then five boost modules needing wchar. If we make > boost depending on wchar we add reverse dependencies to some packages > which could be build without wchar before. > > What do you think? * cc-tool -> already depends on wchar * gnuradio -> already depends on wchar * kodi -> already depends on wchar * libftdi1 -> does not depend on wchar yet, but only the C++ bindings of libftdi1 need boost, so I think it's OK * mpd -> already depends on wchar * pulseview -> already depends on wchar * python-libconfig -> already depends on wchar since it depends on python or python3, and both depend on wchar * thrift -> already depends on wchar * yaml-cpp -> would add a wchar dependency * zmpqq -> already depends on wchar So the only affected packages are the C++ bindings of libftdi1 and yaml-cpp. Conclusion: I believe it's reasonable to add a wchar dependency globally to boost. It's a pain to keep track which sub-option of boost needs or does not need wchar, and boost is a "big" thing, so having an additional toolchain dependency that is nevertheless already very common, is not a big issue. Thanks! Thomas -- Thomas Petazzoni, CTO, Free Electrons Embedded Linux, Kernel and Android engineering http://free-electrons.com