This is what the Paul said about the issue concerning uname -s. ------- Start of forwarded message ------- Date: Thu, 24 Oct 2002 11:01:37 -0700 (PDT) From: Paul Eggert <[EMAIL PROTECTED]> To: [EMAIL PROTECTED] CC: [EMAIL PROTECTED] In-reply-to: <[EMAIL PROTECTED]> ([EMAIL PROTECTED]) Subject: Re: uname -s
> From: [EMAIL PROTECTED] (Alfred M. Szmidt) > Date: 24 Oct 2002 13:33:29 +0200 > > According to POSIX, uname -s should print the "implementation of the > operating system" [1], but according to GNU uname it should print the > kernel name. Is there a reason for this? The coreutils maintainer and I discussed this with Richard Stallman in September 2001, and we came up with the approach currently used in GNU coreutils uname. POSIX does not have the notion of "kernel", so it cannot provide much guidance as to when to output the name of the operating system, or to output the name of the kernel. Historically, "uname -s" prints the sysname component of what the `uname' system call returns, and we didn't want to change that due to backward compatibility concerns. So we loosely interpreted the POSIX phrase "implementation of the operating system" to mean "kernel", and we added the "-o" option to GNU uname to print the operating system name. The operating system name is currently determined by a heuristic at coreutils build time, since there's no portable system call to determine it. ------- End of forwarded message ------- _______________________________________________ Bug-hurd mailing list [EMAIL PROTECTED] http://mail.gnu.org/mailman/listinfo/bug-hurd
