From: David Garamond <lists@...6.isreserved.com>
Date: 2004-01-12T22:52:45+09:00
Subject: Unixisms (Re: [PATCH] File.readable_world? and File.writable_world?)

Yukihiro Matsumoto wrote:
> |> I think it would be source of confusion.  They are too similar.  How
> |> about writable_world? returns its permission bits if the file is
> |> writable from others, otherwise nil?
> 
> |That's an interesting idea. The only problem is that world_writable?
> |would return nil when checking a file that has accidentally acquired,
> |for example, mode 446, which is very insecure. Returning nil in such a
> |case from a method with a name like world_writable? might also be
> |confusing.
> 
> The mode 446 is -r--r--rw-, which is writable from others, so that
> world_writable? should not return nil, right?
> 
> |What about Nobu's suggestion to have the methods in File::Stat, too? Do
> |you want me to add them there, also?
> 
> Sounds nice.

[moved to ruby-talk]

Funny that yesterday I was reading some comments on Slashdot about 
Ruby's slant towards Unixism and against Unicode integration (especially 
compared to Python). Even Perl doesn't have a -w for world fields :-)

I don't mind Unixisms in Ruby though, since I work mainly in Linux 
anyway :-) But I wonder, can this method make sense on Windows' NTFS 
permission styles, or with ACLs?

-- 
dave