bug-fileutils
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [RFC] First batch of changes for ls to support GNU extensions.


From: Alfred M. Szmidt
Subject: Re: [RFC] First batch of changes for ls to support GNU extensions.
Date: Thu, 28 Mar 2002 22:02:57 +0100
User-agent: Gnus/5.090004 (Oort Gnus v0.04) Emacs/21.2.50

* Roland McGrath writes:
[snip]
> Put two spaces after the end of a sentence in a texinfo file.
> Typo/spelling error: "extension" Use complete sentences, i.e. a
> period after "extension."  (The same errors are repeated in the
> other documentation addition.)

Any way of making Emacs do that automaticly?

[snip].

All the things you noted above have been fixed.

> Now as to the features.  I tend to think the switches should be
> added unconditionally, and just marked as either no-op or an error
> at run time.  A switch like "append translator info on the end" can
> just be a no-op on systems without translators.  All the switches
> having to do with the nobody bits can just always act as if all the
> bits are clear when on a system where, in fact, they always are.

> As to what switches we want, I don't think a single "Hurd
> extensions" switch is right at all.  There should be
> function-specific switches about nobody bits and about translators.

Maybe.. The reason why I made it one switch was that people would probably
want to see the nobody bits and translator information at the same time.

At least that is what I would like to have...

> First, a simple switch to print the nobody bits.  It should print
> something other than "---" or showing the other bits when S_IUSEUNK
> is clear, e.g. "..."  or "***" or something.  That way it's never
> ambiguous whether the nobody bits happen to be set the same as the
> other bits or whether S_IUSEUNK is not set.

Well, when S_IUSEUNK is not set the default is to use the group bits (unless
there are plans on changing this). So setting them to "---" would give the
user a false impression, as the group bits will be used.

> We might also want to have a switch that does something like print
> the nobody bits iff they are set on any file in the listing.  That
> behavior could conceivably be the default in the absence of
> POSIXLY_CORRECT.

> As to translators, I am not sure what all options we want.
> Certainly we need an option to always use O_NOTRANS, analogous to
> -d.  Under that option, ls would use
> fd=open(,O_NORW|O_NOTRANS);fstat(fd) instead of plain lstat.  I'm
> not sure exactly which other options we need.

Any ideas on a name for that switch (--translator?)?
No idea what options we need for ls.. Anyone got any bright ideas to share?

-- 
Alfred M. Szmidt



reply via email to

[Prev in Thread] Current Thread [Next in Thread]