On Sun, 15 Feb 2015 17:55:47 +0100 Markus Teich <markus.te...@stusta.mhn.de> wrote:
> > This way of writing usage() is common, because it is called when an error in > > arguments is detected or when the user interactively ask for it, and in this > > last case the return value is not important because it is not going to be > > executed in any script where the return value is important. > > This can easily be solved by not calling die() from usage() but the other way > around. usage() only prints the info and does not exit anything. die() calls > usage and then exits accordingly. I mostly see it done this way. What's the problem? k0ga imho said pretty clearly why the usage() has developed as we know it. The normal use is with eprintf() across the suckless-projects, but the non-zero exit-status is a common approach and unimportant if you use it in an interactive environment. Cheers FRIGN -- FRIGN <d...@frign.de>
