[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: --enable-static in gconfmm2.0-- is it really needed?



James Strandboge wrote:
Hi,
Hi James,

Thank you very much for your backport. I use it on several production systems without any problems. Good work!


I am the maintainer of the gnome2.2 backport for woody, and have tried
unsuccessfully to get the c++ bindings to compile on woody.  With the
newly added gtk+2.2.2, I have actually gotten much further.  I use
CC=gcc-3.0 and CXX=g++-3.0 for everything 'mm' related, however when I
get to compiling gconfmm2.0, it dies with:

perhaps depending on (or redistributing) the excellent g++ 3.3 backport of Adrian Bunk might be another choice.

I chose that one for my (unpublished=internal) recompilation of libgtkmm2.0 on woody+gnome2.2 and never regretted it.

You should ask the C++ programmers you target whether 3.0 is sufficient for their needs - I would clearly avoid it [the dynamic_cast bug forced me to abandon 2.95 for my programs*, and if I had to use a non-default compiler I would not choose a no longer maintained version with known bugs, even if it's part of the standard distribution and slightly more easily available]

As a side-effect this would give you better binary compatibility with other distributions and even sid (rendering glibc 2.2/2.3 the only possible difference).

   Christof

*) non trivial glademm generated programs simply segfault when compiled with 2.95.



Reply to: