Re: <OK> Re: [Groff] moving TOC to start

[M Bianchi]

On Fri, Sep 30, 2005 at 12:20:40PM +0200, Tadziu Hoffmann wrote:
> >     :
> > BTW, I *never* have *any* user writeable directory before the
> > system binary directories, in *my* PATH;  but, I guess it would
> > be naive to expect everybody to follow that piece of simple
> > security advice.
> 
> Obviously you're doing the Right Thing in this regard, but I
> find it sometimes convenient to "replace" some system programs
> with other versions or with wrappers with added functionality,
> and these must come before the "normal" programs in PATH if you
> don't always want to type the complete path to the executable
> (or remeber a new name for each).

This idea that _MY_ version of  ls  is better than the one everyone else uses
is something that goes back to the dawn of UNIX.  ((Insert long involved story
where customer complains that a Makefile is broken because it doesn't work with
his version of  ls  here.))

But it is _OH_ so convenient!

So what I _try_ to do and usually do ((insert lecture on human failings here))
is to name my personal versions of commands, and personal commands, starting
with captical letters;  a sampling:

        A             Lpr           Ntrigue       Rlogin        WhoIs
        Acroread      M             Ooffice       Ssh           XTERM
        CI            Make          Pr            T             Xanim
        Date          Man           Ps            Vnoai         Xclock
        Locate        Netscape      Psmm          Whitelist     Xfig
        Lpq           Nslookup

Two advantages:
        CI  doesn't interfere with  ci

        When I use  CI  in my other shell scripts I can see that I'm probably
        using features contained there.  When it is  ci  I _know_ I'm sticking
        to the standard.

Does it guarantee security?  No.  But all those files marked unwritable, and if
I was really smart  ~/bin  would be too.

-- 
 Mike Bianchi