Re: Oct and Nov 2012

On 21/10/12 15:05, Earnie Boyd wrote:
>> I still have my Mercurial sandbox, with master branch recording the
>> state of CVS HEAD at the time, and this development on a feature branch,
>> so I can easily create a patch relative to that old CVS state, if you'd
>> like that?
>
> This is what I expected.

Okay, but I'd like to review it anyway, before I post it.

>> If you'd like me to rework it, relative to your current WSL state, that
>> will take me a bit longer.
>
> So that I can do this.

Well, I can certainly leave you to complete the final integration, but 
I'd also like to assess the probable impact, as part of my review.  I do 
have some reservations, (mostly cosmetic, but some possibly legal), 
concerning the changes you've made to date:

1) What benefit accrues from changing '* filename.c', as an initial
    comment identifying each source file, to '* @file filename.c'?  IMO,
    the latter is no more than a cosmetic (i.e. ugly) distraction.

2) Somewhat similarly, but with the usual IANAL disclaimer, likely more
    serious: I'm fairly sure that '@copy 2012 MinGW.org project' lacks
    any legal connotation, as would be attributable to a formal statement
    of copyright:

      Copyright (C) 2012, MinGW.org Project

    I vaguely recall[*] some guidance offered by Eben Moglen, to FSF
    developers, a few years ago; IIRC, the word 'Copyright' needs to
    appear *verbatim*, followed by a list of the individual years in
    which a copyrighted version was published, and finally the name of
    the copyright holder; the international copyright symbol, or (C) to
    represent it, is commonly included, but is not a legal requirement;
    (the UK copyright service, however, advises that it be included),
    At the time, I believe Eben also directed us to list each year to
    which a copyright notice applies *individually*; apparently, year
    ranges lack the legal standing of such a list, (but again, the UK
    copyright service advice contradicts this).

3) When a developer has gone to the trouble of identifying him/herself,
    even in code donated to the public domain, you should *not* remove
    it; in doing so, you are infringing the moral rights of those
    developers to be identified as the authors of that code, which, if
    I recall from our own previous discussions, is one of the objections,
    (possibly even the principal objection), to the concept of public
    domain in those jurisdictions which decline to recognise it.  With
    our migration to a more formal statement of copyright, surely it
    becomes even more important that this moral right be protected?

[*] And indeed, this recollection is supported by, and enshrined within 
the FSF's guidance notes on using the various forms of GNU licences:
http://www.gnu.org/licenses/gpl-howto.html

-- 
Regards,
Keith.

------------------------------------------------------------------------------
Everyone hates slow websites. So do we.
Make your web apps faster with AppDynamics
Download AppDynamics Lite for free today:
http://p.sf.net/sfu/appdyn_sfd2d_oct