On Tue, Mar 21, 2006 at 10:23:40PM +0100, Christian E. Boehme wrote: > On Mon, Mar 20, 2006 at 08:36:05PM -0800, Steve Langasek wrote: > > Huh? Are you really suggesting that the standard C++ compiler has been > > unable to find any of its own header files for over a week in unstable, and > > no one noticed? > That's what the report might lead to assume. I _definitely_ wondered > for myself. I have currently three g++ incarnations (3.3, 3.4, 4.0.3) > installed and none of these actually let me transform my sources into > binaries. The 3.3 and 3.4 versions _do_ find their headers (as indicated > in my report) but fail at the linking stage. Apparently, they have no > idea which C++ library to link against as all the unresolved symbols > come from ``std''. Then you're obviously doing something wrong, but you haven't actually told us what you're doing, which makes it difficult to help you debug it. Please show us the commands you're running to try to build/link these programs, as well as the simplest C++ program for which you're seeing this problem. > > g++-4.0 is certainly not unable to find its header files in the general > > case, so this bug doesn't qualify as "rendering the package unusable". > Over here it does. What I find noteworthy is the fact that ``apt-get > install g++-4.0'' did _not_ lead to installing the headers. I had to > explicitly invoke apt-get on libstdc++6-4.0-dev to have these installed > yet still g++-4.0 does not find them. What's the point in having a C++ > compiler without it's headers on a system ? AFAIK, it's a perfectly valid use case to use g++ for compiling C++ code which doesn't reference the standard C++ library. -- Steve Langasek Give me a lever long enough and a Free OS Debian Developer to set it on, and I can move the world. vorlon@debian.org http://www.debian.org/
Attachment:
signature.asc
Description: Digital signature