Incorporating patches into GNU Emacs

Incorporating patches into GNU Emacs

[Bill Wohler]

So someone sends you a cool patch to put into Emacs. You've signed the
copyright assignment papers and can check code into Emacs; they have not
and can not.

Can you simply incorporate their patch and give them credit in the
ChangeLog?

--
Bill Wohler <wohler@newt.com>  http://www.newt.com/wohler/  GnuPG ID:610BD9AD
Maintainer of comp.mail.mh FAQ and MH-E. Vote Libertarian!
If you're passed on the right, you're in the wrong lane.

Re: Incorporating patches into GNU Emacs

[Steve Youngs]

|--==> "BW" == Bill Wohler <wohler@newt.com> writes:

  BW> So someone sends you a cool patch to put into Emacs. You've
  BW> signed the copyright assignment papers and can check code into
  BW> Emacs; they have not and can not.

  BW> Can you simply incorporate their patch and give them credit in
  BW> the ChangeLog?

The way I understand it, no.  

The author of the original code needs to assign the copyright to the
FSF.  For various reasons, the FSF _must_ hold the copyright on every
single line of code it distributes.  You can't assign a copyright that
you have no legal right to.  Only the guy with the cool patch, or
possibly his employer can do that.  And just because you incorporate
his code into your own doesn't alter the fact that you don't hold the
copyright to that code.

You could also quite possibly be breaching cool patch author's
copyright.  But that's highly unlikely if he has sent you the patch,
although his employer might have a claim on it.

-- 
|---<Steve Youngs>---------------<GnuPG KeyID: 10D5C9C5>---|
|        XEmacs - The only _______ you'll ever need.       |
|          Fill in the blank, yes, it's THAT good!         |
|------------------------------------<youngs@xemacs.org>---|

Re: Incorporating patches into GNU Emacs

[Bill Wohler]

Steve Youngs <youngs@xemacs.org> writes:

> |--==> "BW" == Bill Wohler <wohler@newt.com> writes:
>
>   BW> So someone sends you a cool patch to put into Emacs. You've
>   BW> signed the copyright assignment papers and can check code into
>   BW> Emacs; they have not and can not.
>
>   BW> Can you simply incorporate their patch and give them credit in
>   BW> the ChangeLog?
>
> The way I understand it, no.  
>
> The author of the original code needs to assign the copyright to the
> FSF.  For various reasons, the FSF _must_ hold the copyright on every
> single line of code it distributes.  You can't assign a copyright that
> you have no legal right to.  Only the guy with the cool patch, or
> possibly his employer can do that.  And just because you incorporate
> his code into your own doesn't alter the fact that you don't hold the
> copyright to that code.
>
> You could also quite possibly be breaching cool patch author's
> copyright.  But that's highly unlikely if he has sent you the patch,
> although his employer might have a claim on it.

Strictly speaking, you're probably right.

Still, I find it hard to believe that a) every Emacs user on the planet
has submitted papers to the FSF or b) no patches have been incorporated
into Emacs from folks who haven't submitted papers. Where's the middle
ground?

--
Bill Wohler <wohler@newt.com>  http://www.newt.com/wohler/  GnuPG ID:610BD9AD
Maintainer of comp.mail.mh FAQ and MH-E. Vote Libertarian!
If you're passed on the right, you're in the wrong lane.

Re: Incorporating patches into GNU Emacs

[Steve Youngs]

|--==> "BW" == Bill Wohler <wohler@newt.com> writes:

  BW> Can you simply incorporate their patch and give them credit in
  BW> the ChangeLog?

  BW> Steve Youngs <youngs@xemacs.org> writes:
  >>The way I understand it, no.  

  BW> Strictly speaking, you're probably right.

  BW> Still, I find it hard to believe that a) every Emacs user on the planet
  BW> has submitted papers to the FSF or 

Of course not.  There's no reason for it.  You don't need to assign
anything to just use Emacs.

  BW> b) no patches have been incorporated into Emacs from folks who
  BW>    haven't submitted papers. Where's the middle ground?

The FSF, at its discretion, can waive the requirement of an
assignment.  What the FSF takes into account when making such a
decision, I couldn't say.  They'd more than likely have some sort of
"rule of thumb" that balances aggregated code already accepted plus
the current patch against the risk.

And that's just a long-winded way of saying "it depends". :-)

-- 
|---<Steve Youngs>---------------<GnuPG KeyID: 10D5C9C5>---|
|        XEmacs - The only _______ you'll ever need.       |
|          Fill in the blank, yes, it's THAT good!         |
|------------------------------------<youngs@xemacs.org>---|

Re: Incorporating patches into GNU Emacs

[Bill Wohler]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 2436 bytes --]

Here's a specific example from bug.gnu.emacs: A patch from an Emacs user
who hasn't signed papers. Are you telling me that a) you'll ignore it,
or b) you'll really go through all the trouble of getting Gustav to sign
papers in order to use it?

p.s. Is there an Emacs Maintainers Manual where all this stuff is neatly
summarized?


From: Gustav Hållberg <gustav@virtutech.se>
Subject: vc-annotate does not copy truncate-lines from the origin buffer
If you have set truncate-lines in a buffer, it would make a lot of sense for
the vc-annotate command to use that setting in the newly created annotation
buffer too (the same could probably be said for other variables too, but this
is the one I was being annoyed by right now).

Suggested patch attached.

- Gustav

Newsgroups: gnu.emacs.bug
Date: 17 Apr 2003 11:34:18 +0200
X-Boundary: _______________________________________________________________________________

[2. text/x-patch; vc-annotate-truncate-lines.patch]

--- /disk1/sources/emacs-21.3/lisp/vc.el	Mon Jan 20 12:03:30 2003
+++ vc.el	Thu Apr 17 11:27:14 2003
@@ -2941,6 +2941,7 @@
   (vc-ensure-vc-buffer)
   (let* ((temp-buffer-name (concat "*Annotate " (buffer-name) "*"))
          (temp-buffer-show-function 'vc-annotate-display)
+	 (temp-truncate-lines truncate-lines)
          (rev (vc-workfile-version (buffer-file-name)))
          (vc-annotate-version 
           (if prefix (read-string 
@@ -2953,7 +2954,7 @@
                                    nil nil "1.0"))
             1.0))
          (vc-annotate-backend (vc-backend (buffer-file-name))))
-    (message "Annotating...")
+    (message "Annotating... %s" temp-truncate-lines)
     (if (not (vc-find-backend-function vc-annotate-backend 'annotate-command))
 	(error "Sorry, annotating is not implemented for %s"
 	       vc-annotate-backend))
@@ -2962,6 +2963,9 @@
 		       (file-name-nondirectory (buffer-file-name))
 		       (get-buffer temp-buffer-name)
                        vc-annotate-version))
+    (save-excursion
+      (set-buffer temp-buffer-name)
+      (setq truncate-lines temp-truncate-lines))
     ;; Don't use the temp-buffer-name until the buffer is created
     ;; (only after `with-output-to-temp-buffer'.)
     (setq vc-annotate-buffers

--
Bill Wohler <wohler@newt.com>  http://www.newt.com/wohler/  GnuPG ID:610BD9AD
Maintainer of comp.mail.mh FAQ and MH-E. Vote Libertarian!
If you're passed on the right, you're in the wrong lane.

Re: Incorporating patches into GNU Emacs

[Stefan Monnier]

> Here's a specific example from bug.gnu.emacs: A patch from an Emacs user
> who hasn't signed papers. Are you telling me that a) you'll ignore it,
> or b) you'll really go through all the trouble of getting Gustav to sign
> papers in order to use it?

For such trivial changes, we don't need papers.

> p.s. Is there an Emacs Maintainers Manual where all this stuff is neatly
> summarized?

I think that the standards.texi document tries to explain a bit
of it, but I can't remember the details.

Also, in many cases, it's easier to rewrite the thing than to get
the paperwork done.  After all, in most cases writing the code is the
easy part.  And also you generally have a better knowledge of Emacs
so you get to write a better solution (e.g. using with-current-buffer
rather than save-excursion+set-buffer).
Also I'd argue in the specific case that truncate-lines should actually
always be set to t for vc-annotate buffers.


	Stefan

Re: Incorporating patches into GNU Emacs

[Kim F. Storm]

"Stefan Monnier" <monnier+gnu/emacs@rum.cs.yale.edu> writes:

> > Here's a specific example from bug.gnu.emacs: A patch from an Emacs user
> > who hasn't signed papers. Are you telling me that a) you'll ignore it,
> > or b) you'll really go through all the trouble of getting Gustav to sign
> > papers in order to use it?
> 
> For such trivial changes, we don't need papers.

You can still attribute such changes to the author by writing (tiny change)
after his name in the change log.  See the ChangeLog files for examples.

-- 
Kim F. Storm <storm@cua.dk> http://www.cua.dk

Re: Incorporating patches into GNU Emacs

[Bill Wohler]

Aha! Now we're getting somewhere.

"Stefan Monnier" <monnier+gnu/emacs@rum.cs.yale.edu> writes:

>> Here's a specific example from bug.gnu.emacs: A patch from an Emacs user
>> who hasn't signed papers. Are you telling me that a) you'll ignore it,
>> or b) you'll really go through all the trouble of getting Gustav to sign
>> papers in order to use it?
>
> For such trivial changes, we don't need papers.

That's what I thought. At which point do the changes become nontrivial?

> Also, in many cases, it's easier to rewrite the thing than to get
> the paperwork done.

OK, so *that's* what folks do in practice. Simply take the ideas and
incorporate them.

Still curious about what constitutes a non-trivial patch though.

--
Bill Wohler <wohler@newt.com>  http://www.newt.com/wohler/  GnuPG ID:610BD9AD
Maintainer of comp.mail.mh FAQ and MH-E. Vote Libertarian!
If you're passed on the right, you're in the wrong lane.

Re: Incorporating patches into GNU Emacs

[Richard Stallman]

    The author of the original code needs to assign the copyright to the
    FSF.  For various reasons, the FSF _must_ hold the copyright on every
    single line of code it distributes.  You can't assign a copyright that
    you have no legal right to.  Only the guy with the cool patch, or
    possibly his employer can do that.  And just because you incorporate
    his code into your own doesn't alter the fact that you don't hold the
    copyright to that code.

That is correct.  Only the person who wrote the code can
assign or disclaim the copyright on it.

    The FSF, at its discretion, can waive the requirement of an
    assignment.  What the FSF takes into account when making such a
    decision, I couldn't say.

Our lawyer says we should do this rarely, only when it is very
important.  And we want to get some other sort of contract with
the author in lieu of the normal assignment.

Re: Incorporating patches into GNU Emacs

[Richard Stallman]

    That's what I thought. At which point do the changes become nontrivial?

We can use up to 10 to 20 lines from one person before we
need papers.