Re: Compiled files without sources????

[Stephen J. Turnbull]

 > On Sun, Jul 31, 2011 at 9:03 PM, David Kastrup <address@hidden> wrote:

 > > Anyway, I find it bad taste to drag an internal discussion from
 > > the list into the limelight like that.

It's not an "internal discussion".  It is infringement of copyright
for anyone but the FSF (technically, including members of the Emacs
project publishing private branches!) to redistribute a complete Emacs
of those versions verbatim.  There are probably millions of copies in
the wild, and in theory a for-profit distributor could be sued for
statutory damages and subjected to criminal prosecution[1], and *that*
is *no joke*.  The public deserves to be told about this, just as they
are told about other major defects such as security holes in
webservers.

Since this is the FSF's mistake, I can't really see it doing more than
informing people that their distributions are infringing and politely
requesting them to cease and desist until upstream is fixed, of
course.

 > > The mud-slinging is detrimental to making a solid and timely fix.

I don't see why.  Making the fix *is* a purely internal matter.  What
LWN says should (and I expect will) have no effect on the fix (which
is already in progress).

True, the fact that even the FSF can mess up this way is going to make
a few people wonder whether the downsides of copyleft are more serious
than had previously been apparent.  So make a solid and timely fix and
those who have open minds will dismiss the whole thing as a rather
large typo.

Tim Cross writes:

 > Personally, I was surprised when I read the lead-in. I also thought it
 > was an 'internal' matter being dealt with, but thought it worth
 > pointing out, regardless of its accuracy. I also suspect it is really
 > only getting such attention because of emacs being a well known FSF
 > package  and Richard's direct involvement.

Of course.  This is *news*.  Prurient and unimportant, but that's what
sells newspapers, unfortunately.

 > In some ways emacs' and other key FSF projects need to set the
 > standards, which I think was one of Richard's main points.

This isn't really a question of "setting (high) standards."  It's a
matter of (minimal) due diligence (which is required by the license of
licensees), so that the licensees can legally exercise all rights the
license purports to grant.  It's unfortunate that in this case the due
diligence is cleaning up, rather than preventing, a mistake.


Footnotes: 
[1]  Conviction is highly unlikely in the circumstances, of course.
But the law would likely be on the FSF's side as far as forcing the
distributor to try to recover copies, etc., even though the
distribution was 100% in good faith.  That could be rather expensive
for the likes of Red Hat or Canonical!