I think it's important to keep in mind that, with Linux, you cannot tell the user community how they should program, or use the system: LSB, IMHO, must reflect what people actually do and expect of Linux, rather than telling them how you or I think they ought to use it.
So, I would be opposed to specifying anything on the grounds that one "should" or "should not" do things a certain way. System V was a great example of what happens when a committee of sorts _decides_ what is good and bad style. -s Daniel Quinlan wrote: > > Daniel Quinlan <[EMAIL PROTECTED]> writes: > > >> Do we really want applications relying on mail or mailx? Isn't it > >> better practice to use the local copy of /usr/sbin/sendmail? > > Jakob 'sparky' Kaivo <[EMAIL PROTECTED]> writes: > > > Yes, it is better practice to use sendmail. But, there are a large > > number of existing scripts and programs that rely on /bin/mail (and a > > good number of those rely on /bin/mail acting like mailx). We need to > > accomodate those. Also, UNIX98 mandates mailx and their is no real > > compelling reason not to. > > Accomodating broken programs isn't a requirement. /bin/mail isn't > standardized, so you can get different programs on different systems. > mailx is standardized, but is designed to be an *interactive* mail > system. Unix98 mandates it, but we need to be careful. Our focus is > on programs used by applications, not interactive users. > > I think it would be better to document a subset of the sendmail API > for LSB applications. > > Dan > > -- > To UNSUBSCRIBE, email to [EMAIL PROTECTED] > with subject of "unsubscribe". Trouble? Email [EMAIL PROTECTED]
