New answers tagged compiling
You need to set up your build scripts so that object files for different architectures are written into different directories. There will naturally be no conflicts between Windows objects and Unix objects because they use different file names (*.obj and *.exe vs. *.o and *), but Unix systems use the same file names. Alternatively, unpack the sources in ...
Try ./prog to run prog in the current working directory, as . is typically not (nor should be) in PATH. Also, a Makefile is perhaps much more sensible, as then you can simply type make test and have the program built (if necessary) and tested: prog: prog.c test: prog echo blah de blah | ./prog A Makefile can also integrate with emacs or vim ...
If the build was successful, once it is completed, there will be a foo.pkg.tar.xz in the build directory and you can use pacman to install the built package: sudo pacman -U foo.pkg.tar.xz
You need to install the appropriate linux-headers package. apt-get install linux-headers-$(uname -r) should do the trick.
You should find everything you need to rebuild in the src.deb package that matches you .deb package. Look at apt-src(8).
I have successfully built the library using the command line instruction given here. I have had difficulties to build a sample in Xcode. To get it done, I have defined the search paths for the headers files and the libraries ("Build Settings" pane). Than I have defined the library itself in the "Build Phases" than "Link Binary With Libraries" pane. After ...
From the looks of your screenshot, it looks like you downloaded the pre-compiled binaries. If that's the case, you won't have to compile anything. You should find an executable nodejs file inside the bin folder. If you want to compile your own, you should download the source tarball here instead of the binary tarball.
There is the UNIX linker that is available freely in source (under the CDDL). The sourcecode was made available with the OpenSolaris project. Due to several deficits and bugs the GNU linker is no alternative to the UNIX ld on Solaris. On the other side, it may be that linking on Linux depends on the gld deviations (e.g. on different command line options and ...
Why don't you try unpacking, but not fully installing, the binary package for your OS. For example, I looked in the releases dir and downloaded (with e.g. firefox, curl, or wget) the 64 bit rpm for my Fedora 22, into a newly created directory, myatom, then unpacked it with cd myatom rpm2cpio x86_64.rpm | cpio -cid and ran ~/myatom/usr/bin/atom and ...
If the package happens to be in a debian repo somewhere, you can use build-dep to install the build dependencies and mark those as 'automatically' installed. You can then use autoremove to cleanup those build deps. apt-get build-dep -o APT::Get::Build-Dep-Automatic=true WhatImBuilding apt-get autoremove If whatever you're building doesn't already have a ...
You can manually edit the configure.ac file, look up the atomic section, and replace the line AC_RUN_IFELSE([AC_LANG_PROGRAM([[ with AC_COMPILE_IFELSE([AC_LANG_PROGRAM([[ You then need to run bootstrap.sh to recreate the configure script. This should let you configure, but compilation/linking might fail because your toolchain may not, in fact, support ...
Top 50 recent answers are included