Credit for vsprintk work

Credit for vsprintk work

[Steven Rostedt]

Frederic,

 $ git blame lib/vsprintf.c |grep 'Frederic Weisbecker' | wc -l
542

You did a hell of a lot of work on vsprintf.c, and you are still too 
modest to put in a copyright at the top.

I know some people think that git is good enough for ownership. But 
really, the amount of changes that you did in vsprintf.c deserves a bit of 
credit in the source file.

Is everyone OK if a copyright goes at the top of this file giving credit 
to Frederic?  If for anything else, it will facilitate people to find who 
broke it if it breaks ;-)

I think Frederic should send one more patch to add his copyright. That is, 
of course, if Frederic wants too.

Note, Frederic did not ask me to post this. In fact, this is the first 
time he's seen this complaint. I'm posting this because I was showing this 
code to some students at a University and when I wanted to point out 
who wrote it, I was surprised that Frederic's name was not mentioned.
I was looking at the code via LXR so git was not helpful here.

-- Steve

Re: Credit for vsprintk work

[Steven Rostedt]

On Wed, 25 Mar 2009, Steven Rostedt wrote:

> 
> Frederic,
> 
>  $ git blame lib/vsprintf.c |grep 'Frederic Weisbecker' | wc -l
> 542
> 
> You did a hell of a lot of work on vsprintf.c, and you are still too 
> modest to put in a copyright at the top.
> 
> I know some people think that git is good enough for ownership. But 
> really, the amount of changes that you did in vsprintf.c deserves a bit of 
> credit in the source file.
> 
> Is everyone OK if a copyright goes at the top of this file giving credit 
> to Frederic?  If for anything else, it will facilitate people to find who 
> broke it if it breaks ;-)
> 
> I think Frederic should send one more patch to add his copyright. That is, 
> of course, if Frederic wants too.
> 
> Note, Frederic did not ask me to post this. In fact, this is the first 
> time he's seen this complaint. I'm posting this because I was showing this 
> code to some students at a University and when I wanted to point out 
> who wrote it, I was surprised that Frederic's name was not mentioned.
> I was looking at the code via LXR so git was not helpful here.

Lai also did this work too, so I recommend adding something like:

 * Binary printk modifications, copyright 2009:
 *  Frederic Weisbecker <fweisbec@gmail.com>
 *  Lai Jiangshan <laijs@cn.fujitsu.com>

Or something similar.

-- Steve

Re: Credit for vsprintk work

[Ingo Molnar]

* Steven Rostedt <rostedt@goodmis.org> wrote:

> 
> 
> 
> On Wed, 25 Mar 2009, Steven Rostedt wrote:
> 
> > 
> > Frederic,
> > 
> >  $ git blame lib/vsprintf.c |grep 'Frederic Weisbecker' | wc -l
> > 542
> > 
> > You did a hell of a lot of work on vsprintf.c, and you are still too 
> > modest to put in a copyright at the top.
> > 
> > I know some people think that git is good enough for ownership. But 
> > really, the amount of changes that you did in vsprintf.c deserves a bit of 
> > credit in the source file.
> > 
> > Is everyone OK if a copyright goes at the top of this file giving credit 
> > to Frederic?  If for anything else, it will facilitate people to find who 
> > broke it if it breaks ;-)
> > 
> > I think Frederic should send one more patch to add his copyright. That is, 
> > of course, if Frederic wants too.
> > 
> > Note, Frederic did not ask me to post this. In fact, this is the first 
> > time he's seen this complaint. I'm posting this because I was showing this 
> > code to some students at a University and when I wanted to point out 
> > who wrote it, I was surprised that Frederic's name was not mentioned.
> > I was looking at the code via LXR so git was not helpful here.
> 
> Lai also did this work too, so I recommend adding something like:
> 
>  * Binary printk modifications, copyright 2009:
>  *  Frederic Weisbecker <fweisbec@gmail.com>
>  *  Lai Jiangshan <laijs@cn.fujitsu.com>
> 
> Or something similar.

I'd suggest the standard copyright form:

  *
  * Binary printk modifications and generalizations:
  *
  *    Copyright (C) 2009  Frederic Weisbecker <fweisbec@gmail.com>
  *    Copyright (C) 2009  Lai Jiangshan <laijs@cn.fujitsu.com>
  * 

	Ingo

Re: Credit for vsprintk work

[Linus Torvalds]

On Wed, 25 Mar 2009, Steven Rostedt wrote:
> 
>  $ git blame lib/vsprintf.c |grep 'Frederic Weisbecker' | wc -l
> 542
> 
> You did a hell of a lot of work on vsprintf.c, and you are still too 
> modest to put in a copyright at the top.

I don't disagree, but "git blame" is _not_ a good way to do this.

At a minimum, use "git blame -w -M" to ignore stuff that was just moved 
and re-indented. but even then, quite frankly, "git blame" doesn't really 
make much sense. If you have

	if (!(flags & LEFT)) {
		while (len < field_width--) {

that turns into

	if (!(spec.flags & LEFT)) {
		while (len < spec.field_width--) {

then blame will count the new lines to the person who did the change, but 
was that really a code change? It was a small re-org.

I picked that example because those particular lines actually go back to 
before even the bitkeeper days - their original author is lost in history. 
It used to look different (field_width was called 'size' back in the 
days), but the logic of the code may well go back to the original code by 
Lasu in 1991.

The only point here? Don't use "git blame" as a copyright assigner.

			Linus

Re: Credit for vsprintk work

[Steven Rostedt]

On Wed, 25 Mar 2009, Linus Torvalds wrote:

> 
> 
> On Wed, 25 Mar 2009, Steven Rostedt wrote:
> > 
> >  $ git blame lib/vsprintf.c |grep 'Frederic Weisbecker' | wc -l
> > 542
> > 
> > You did a hell of a lot of work on vsprintf.c, and you are still too 
> > modest to put in a copyright at the top.
> 
> I don't disagree, but "git blame" is _not_ a good way to do this.
> 
> At a minimum, use "git blame -w -M" to ignore stuff that was just moved 
> and re-indented. but even then, quite frankly, "git blame" doesn't really 
> make much sense. If you have
> 
> 	if (!(flags & LEFT)) {
> 		while (len < field_width--) {
> 
> that turns into
> 
> 	if (!(spec.flags & LEFT)) {
> 		while (len < spec.field_width--) {
> 
> then blame will count the new lines to the person who did the change, but 
> was that really a code change? It was a small re-org.
> 
> I picked that example because those particular lines actually go back to 
> before even the bitkeeper days - their original author is lost in history. 
> It used to look different (field_width was called 'size' back in the 
> days), but the logic of the code may well go back to the original code by 
> Lasu in 1991.
> 
> The only point here? Don't use "git blame" as a copyright assigner.

I totally agree. I only used the git blame as a prelude to my argument. In 
fact, your explanation here verifies my point even more. That is, some 
people do not like the change log type notifications at the top of files.
You even point out that the original author was lost in history.

I agree that trivial changes to files or even simple bug fixes do not 
qualify a name at the top of the file. But any time there is a design 
change, that should be credited. git blame was not my reason I posted this 
email. I was Cc'd on all those emails that went back and forth for this 
improvement. As I was showing students the work that was done, I noticed 
that there was no copyright statement for it.

But I understand what you are saying. I should have been clearer in my 
intent, otherwise you'll get a bunch of people crying for copyrights just 
because they fixed all the whitespace in a file ;-)

-- Steve

Re: Credit for vsprintk work

[Frederic Weisbecker]

On Wed, Mar 25, 2009 at 09:58:52AM -0400, Steven Rostedt wrote:
> 
> Frederic,
> 
>  $ git blame lib/vsprintf.c |grep 'Frederic Weisbecker' | wc -l
> 542
> 
> You did a hell of a lot of work on vsprintf.c, and you are still too 
> modest to put in a copyright at the top.
> 
> I know some people think that git is good enough for ownership. But 
> really, the amount of changes that you did in vsprintf.c deserves a bit of 
> credit in the source file.
> 
> Is everyone OK if a copyright goes at the top of this file giving credit 
> to Frederic?  If for anything else, it will facilitate people to find who 
> broke it if it breaks ;-)
> 
> I think Frederic should send one more patch to add his copyright. That is, 
> of course, if Frederic wants too.
> 
> Note, Frederic did not ask me to post this. In fact, this is the first 
> time he's seen this complaint. I'm posting this because I was showing this 
> code to some students at a University and when I wanted to point out 
> who wrote it, I was surprised that Frederic's name was not mentioned.
> I was looking at the code via LXR so git was not helpful here.
> 
> -- Steve
> 


:-)

Ok, so if nobody is opposite to, I will send a patch for that.
Lai, no problem with me adding your copyright too?

Thanks!

Re: Credit for vsprintk work

[Lai Jiangshan]

Frederic Weisbecker wrote:
> 
> Ok, so if nobody is opposite to, I will send a patch for that.
> Lai, no problem with me adding your copyright too?
> 

It's fine for me.

Thanks.
Lai