Re: RFC: Test::Copyright

2011-06-13 Thread Andrew Suffield
On Tue, Jun 14, 2011 at 07:37:00AM +0200, Philip Newton wrote:
> Also, I'm not sure why the copyright statements in individual files
> need to match the general copyright statement - if a given submodule
> was last updated in 1997 and was stable since then, then I would
> expect it to have a copyright of (say) "Copyright (C) E. X. Ample
> 1995-1997", and not to find "2011" in the copyright line just because
> a new version of the module was released that modified other files in
> the package - hence giving a copyright of (say) "Copyright (C) E. X.
> Ample 1995-1999, 2005-2011" for the entire thing.

The whole thing is completely unnecessary. It's a historical habit
that refuses to go away, and it's not even ours. The presence and form
of copyright statements was necessary in the US prior to 1989, when
they signed the Berne convention after a century of dithering.

The Berne convention explicitly disallows these statements having any
significance - copyright is universal and automatic. The things stick
around because lawyers don't see "it's a pointless waste of time" as a
reason to stop doing something. They usually give a fairly vague
argument about it being a deterrent and discouraging defenses based on
"I didn't know", and then fall back on "Why not include it just to be
safe? You want to be safe don't you? Where's the harm?"

Double irony: the US notice requirement was for the word 'Copyright' or
the C-in-circle symbol, but (C) was never an acceptable substitute.

None of this has ever been relevant in the UK.

I leave you with a copy of /bin/true from Solaris:

8<--
#!/usr/bin/sh
#   Copyright (c) 1984, 1986, 1987, 1988, 1989 AT&T
# All Rights Reserved

#   THIS IS UNPUBLISHED PROPRIETARY SOURCE CODE OF AT&T
#   The copyright notice above does not evidence any
#   actual or intended publication of such source code.

#ident  "@(#)true.sh1.6 93/01/11 SMI"   /* SVr4.0 1.4   */
8<--


Re: RFC: Test::Copyright

2011-06-13 Thread Philip Newton
What's the point of checking for a copyright ending date that matches
the current year?

Or perhaps I should ask, whom do you envisage running such tests? The
developer, or end users?

It's good for the developer, I suppose, but useless for end users - if
they install something that was last updated three years ago (because
it's been stable since then and no bugs were found nor are any new
features needed), then there's no point in having a current copyright.

Also, I'm not sure why the copyright statements in individual files
need to match the general copyright statement - if a given submodule
was last updated in 1997 and was stable since then, then I would
expect it to have a copyright of (say) "Copyright (C) E. X. Ample
1995-1997", and not to find "2011" in the copyright line just because
a new version of the module was released that modified other files in
the package - hence giving a copyright of (say) "Copyright (C) E. X.
Ample 1995-1999, 2005-2011" for the entire thing.

Cheers,
Philip


RFC: Test::Copyright

2011-06-13 Thread Nicholas Bamber
Experience in packaging perl modules from CPAN , suggests that most CPAN
authors are woefully ignorant of copyright and licensing issues.
Unfortunately I have been provoked into attempting to do something about
this. I present Test::Copyright. I would appreciate feedback not least
on the idea.

-- 
Nicholas Bamber | http://www.periapt.co.uk/
PGP key 3BFFE73C from pgp.mit.edu