Re: nodes have authors

[Thomas Bushnell, BSG]

nisse@lysator.liu.se (Niels Möller) writes:

> tb@becket.net (Thomas Bushnell, BSG) writes:
> 
> > Normally (without special flags) you should always follow the
> > translator, and you would never see the => syntax.
> 
> What is "special"? ls normally follows symlinks silently, but ls -F
> and ls -l shows some information about symlinks. Are these particular
> flags special enough?

I think our rule is that ls should continue to use exactly the stat
calls it does now, which follow translators, but sometimes don't
follow symlinks.  Note the difference between O_NOLINK and O_NOTRANS
in this regard.  The former (but not the latter) is supposed to be
used by lstat.

> I think it will be easier to understand the system if symlinks and
> translators are treated in the same way by programs walking the file
> system, in the absence of any hurd-specific flags.

Well, the reason against this is that translators are really just
mount points, and existing ls does *not* treat them differently.

> By the way, what's a good "ls -F"-indicator character for translators?
> `&' (process), `!' (action), or `?' (oddity)? Perhaps active and
> passive translators should be displayed differently.

Hrm, don't know. :)