Mailing-List: contact zsh-workers-help@zsh.org; run by ezmlm
Precedence: bulk
From: Bart Schaefer <schaefer@brasslantern.com>
Message-Id: <160705092948.ZM16622@torch.brasslantern.com>
Date: Tue, 5 Jul 2016 09:29:48 -0700
In-Reply-To: <20160705124832.342d9e48@pwslap01u.europe.root.pri>
Comments: In reply to Peter Stephenson <p.stephenson@samsung.com>
        "Re: Next release (5.3)" (Jul  5, 12:48pm)
References: <20160704114016.190f48be@pwslap01u.europe.root.pri> 
	<160704080424.ZM12848@torch.brasslantern.com> 
	<20160705093321.79d7c4bc@pwslap01u.europe.root.pri> 
	<20160705124832.342d9e48@pwslap01u.europe.root.pri>
To: "Zsh Hackers' List" <zsh-workers@zsh.org>
Subject: Re: Next release (5.3)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii

On Jul 5, 12:48pm, Peter Stephenson wrote:
} Subject: Re: Next release (5.3)
}
} > > (1) Daniel's suggested change to :A [care to offer an opinion?]
} 
} Thinking out loud, don't take any notice...
} 
} It would be possible to add a different letter for this.  However, I
} think you're still faced with asking what the user actually expects
} from the paths they use, which comes back to the option settings.

That's actually sort of the point -- the option settings are to have
CHASE_DOTS unset by default, but :A behaves as if CHASE_DOTS is set,
so it's actually inverted from what the options indicate the user is
expecting.

Sure, we could add another modifier [maybe :L mneme "ls -L"?  Except
confusion with :l and with (L) which has always been an oddball] but
then you have the seemingly unnecessary case that :a:L is :A and that
:L:a may mean something entirely different.

Having written that, though, it would seem that if :L might still
leave ".." in the path (? is that what happens when the symlink is
itself ".."-relative?) then maybe :L is actually needed and even the
proposed change to the :A default is inadequate for all cases.