Re: [PATCH] Remove Bitkeeper documentation from Linux tree

Re: [PATCH] Remove Bitkeeper documentation from Linux tree

[Jonathan A. George]

Jeff,

The BK documentation constitutes an implicit advertisement and 
endorsement of a product with a license which to many developers 
violates the spirit of open source software.  This is not to say that BK 
is not an effective product, nor that the document in question is useful 
for people who choose the tool, but to me it seems comparable to 
including a closed source binary module in the kernel distribution. 
 Moving the document to the BK website would be nicer, and would 
certainly assauge bad feelings regarding such an integral implicit 
endorsement of a tool.

--Jonathan--

Re: [PATCH] Remove Bitkeeper documentation from Linux tree

[Jeff Garzik]

On Mon, Apr 22, 2002 at 01:37:19PM -0500, Jonathan A. George wrote:
> The BK documentation constitutes an implicit advertisement and 
> endorsement of a product with a license which to many developers 
> violates the spirit of open source software.

Agreed.  

And the simple fact of Linus using BitKeeper does the
_exact_ _same_ _thing_.

Talk Linus out of using BitKeeper, and sure, I'll remove the doc.


> This is not to say that BK 
> is not an effective product, nor that the document in question is useful 
> for people who choose the tool, but to me it seems comparable to 
> including a closed source binary module in the kernel distribution. 

No, it is not comparable at all with that.  There are no license
problems with the document -- it is GPL'd.

It describes the same thing as Documentation/SubmittingPatches does, and
is very relevant to kernel development.


>  Moving the document to the BK website would be nicer, and would 
> certainly assauge bad feelings regarding such an integral implicit 
> endorsement of a tool.

Removing the doc from the kernel sources would be a token gesture only.

Some developers disagree violently with the use of non-open-source tools
at all, and that is a fundamental disagreement.

The majority of the "silently seething" developers, I imagine, are only
gonna be satisfied when (a) BitKeeper is GPL'd or (b) Linus stops using
BitKeeper.  Both of these seem very remote possibilities at present.

	Jeff


P.S.  The doc is _not_ going on the BK website by my hand.  (though I
have given BitMover permission to post the doc whereever they wish)
I can maintain my own docs much better than Larry can :)

Re: [PATCH] Remove Bitkeeper documentation from Linux tree

[Jonathan A. George]

Jeff Garzik wrote:

>On Mon, Apr 22, 2002 at 01:37:19PM -0500, Jonathan A. George wrote:
>  
>
>>The BK documentation constitutes an implicit advertisement and 
>>endorsement of a product with a license which to many developers 
>>violates the spirit of open source software.
>>
>Agreed.  
>
>And the simple fact of Linus using BitKeeper does the
>_exact_ _same_ _thing_.
>
Not quite.  Linus uses BK as a tool to facilitate kernel development. 
 However, he has not made BK _part_ _of_ _the_ _kernel_. ;-)  Obviously 
anyone can use any tool ON the kernel, but integrating into the kernel 
is something else.

>Talk Linus out of using BitKeeper, and sure, I'll remove the doc.
>
No need.  His tools are his choice.  The kernel itself is ours not his; 
 thus the distinction.

>>This is not to say that BK 
>>is not an effective product, nor that the document in question is useful 
>>for people who choose the tool, but to me it seems comparable to 
>>including a closed source binary module in the kernel distribution. 
>>
>No, it is not comparable at all with that.  There are no license
>problems with the document -- it is GPL'd.
>
The GPL has a specific intent which (though use in the kernel is more 
like the LGPL because binary modules are permitted) which doesn't 
explicitly extend to such documentation.  The spirit of the GPL which 
causes us to exclude binary modules from the distribution very much does 
apply.  To a lawyer this might not matter, but I expect more from a top 
free software contributor like you.

>It describes the same thing as Documentation/SubmittingPatches does, and
>is very relevant to kernel development.
>
In fact I effectively noted its usefulness above.  Thus no argument here.

>>Moving the document to the BK website would be nicer, and would 
>>certainly assauge bad feelings regarding such an integral implicit 
>>endorsement of a tool.
>>
>Removing the doc from the kernel sources would be a token gesture only.
>
I strongly disagree for the reasons I have noted.

>Some developers disagree violently with the use of non-open-source tools
>at all, and that is a fundamental disagreement.
>
>The majority of the "silently seething" developers, I imagine, are only
>gonna be satisfied when (a) BitKeeper is GPL'd or (b) Linus stops using
>BitKeeper.  Both of these seem very remote possibilities at present.
>
Note:  you are not quoting me here ;-)  However, I completely agree.

>	Jeff
>
>P.S.  The doc is _not_ going on the BK website by my hand.  (though I
>have given BitMover permission to post the doc whereever they wish)
>I can maintain my own docs much better than Larry can :)
>
That is a minor detail (as you know).  My suggestion was an attempt to 
balance the value of your admittedly useful document with the nature of 
the endorsement.

--Jonathan--

Re: [PATCH] Remove Bitkeeper documentation from Linux tree

[Anton Altaparmakov]

At 19:37 22/04/02, Jonathan A. George wrote:
>Jeff,
>
>The BK documentation constitutes an implicit advertisement and endorsement 
>of a product with a license which to many developers violates the spirit 
>of open source software.  This is not to say that BK is not an effective 
>product, nor that the document in question is useful for people who choose 
>the tool, but to me it seems comparable to including a closed source 
>binary module in the kernel distribution. Moving the document to the BK 
>website would be nicer, and would certainly assauge bad feelings regarding 
>such an integral implicit endorsement of a tool.
>
>--Jonathan--

I hereby publically endorse the use of bitkeeper. I think it's a great tool 
and I think anyone who is not using it is missing out. Anyone who asks me 
what a good SCM to use is will hear "bitkeeper" from me. And I shall not 
change my mind until a better or at least as good tool becomes available.

Taking out the document only because it describes a non-free tool is 
ridiculous. I don't object to the removal per se. I could live with that. I 
object to the REASONS for removal.

And I am prepared to speak for that, in case you hadn't noticed. (-;

How about that?

Best regards,

Anton


-- 
   "I've not lost my mind. It's backed up on tape somewhere." - Unknown
-- 
Anton Altaparmakov <aia21 at cantab.net> (replace at with @)
Linux NTFS Maintainer / IRC: #ntfs on irc.openprojects.net
WWW: http://linux-ntfs.sf.net/ & http://www-stu.christs.cam.ac.uk/~aia21/

Re: [PATCH] Remove Bitkeeper documentation from Linux tree

[Alexander Viro]

On Mon, 22 Apr 2002, Jonathan A. George wrote:

> Not quite.  Linus uses BK as a tool to facilitate kernel development. 
>  However, he has not made BK _part_ _of_ _the_ _kernel_. ;-)  Obviously 
> anyone can use any tool ON the kernel, but integrating into the kernel 
> is something else.
> 
> >Talk Linus out of using BitKeeper, and sure, I'll remove the doc.
> >
> No need.  His tools are his choice.  The kernel itself is ours not his; 
>  thus the distinction.

Bullshit.  His copy is his.  Mine is mine.  Yours is yours.  Each of
us is perfectly within his rights when he adds whatever patches he
likes.

And that's _it_.  Linus has absolute control over his copy, as long
as he doesn't run afoul of copyright restrictions.  So do I.  So do
you.  GPL explicitly allows to modify and redistribute result of
modifications.  As the matter of fact, as soon as you attempt to
limit such right, you are losing all rights granted to you by GPL.

"Official" tree is the copy placed by Linus on ftp.kernel.org.  And
as long as ftp.kernel.org admins keep his account (and write permissions
on directory in question) that copy is controlled by Linus.  Period.
End of story.  Linus has exactly the same rights as anybody else and
_everyone_ has a right to modify his copy as he likes.

If you don't like it - take it with RMS and FSF, who happen to feel
very strongly about that right.  That's what GPL is about.  If you
don't like modifications done by somebody, you have only one 
recourse - you are allowed to back them off _in_ _your_ _copy_
and distribute that copy.

It's fscking amazing that self-proclaimed GPL advocates happily
ignore the main stated goal of GPL - to ensure that everybody will be
able to hack on his copy as he wants and share results with
everybody else.  The fact that your modifications are in there
does not allow you to stop anybody else from further modifications.

Re: [PATCH] Remove Bitkeeper documentation from Linux tree

[Jeff Garzik]

On Mon, Apr 22, 2002 at 02:19:13PM -0500, Jonathan A. George wrote:
> Jeff Garzik wrote:
> 
> >On Mon, Apr 22, 2002 at 01:37:19PM -0500, Jonathan A. George wrote:
> >  
> >
> >>The BK documentation constitutes an implicit advertisement and 
> >>endorsement of a product with a license which to many developers 
> >>violates the spirit of open source software.
> >>
> >Agreed.  
> >
> >And the simple fact of Linus using BitKeeper does the
> >_exact_ _same_ _thing_.
> >
> Not quite.  Linus uses BK as a tool to facilitate kernel development. 
>  However, he has not made BK _part_ _of_ _the_ _kernel_. ;-)  Obviously 
> anyone can use any tool ON the kernel, but integrating into the kernel 
> is something else.

That's a poor comparison, we are not integrating BitKeeper into the
kernel :)

Like or not, BK is used by several maintainers, including The Big
Penguin himself.  There are plenty of good reasons why this doc
should be in the kernel source, only a single, lone reason against
it:  some people don't like _talking_ about proprietary software in
the kernel sources.


> >Talk Linus out of using BitKeeper, and sure, I'll remove the doc.
> >
> No need.  His tools are his choice.  The kernel itself is ours not his; 
>  thus the distinction.

He is the gatekeeper.  The kernel can be considered his, too, because it
is on his judgement that changes are accepted and rejected.  If he uses
BitKeeper, and some maintainers find it easier to use BitKeeper, that's
a far stronger implicit advertisement for BitKeeper than some doc will
ever be.  If Linus stopped using BitKeeper, that would provide incentive
for other maintainers to stop using BitKeeper, which in turn means there
is technical merit in the removal of the BK doc.


> >>This is not to say that BK 
> >>is not an effective product, nor that the document in question is useful 
> >>for people who choose the tool, but to me it seems comparable to 
> >>including a closed source binary module in the kernel distribution. 
> >>
> >No, it is not comparable at all with that.  There are no license
> >problems with the document -- it is GPL'd.
> >
> The GPL has a specific intent which (though use in the kernel is more 
> like the LGPL because binary modules are permitted) which doesn't 
> explicitly extend to such documentation.  The spirit of the GPL which 
> causes us to exclude binary modules from the distribution very much does 
> apply.  To a lawyer this might not matter, but I expect more from a top 
> free software contributor like you.

The doc is free by any standard.  The _subject_ of the doc is not free.
There are plenty of documents talking about proprietary software
which are free, covered by the GNU FDL if not the GPL or some other
free license.  Separate the licensing of the _subject_ of the doc,
from the doc itself.

I simply don't see a comparison at all.  Closed source and binary
modules by definition hides IP from most.  The doc is hiding nothing.

I don't think it will accomplish anything, but if it satisfies you or
others, I can change the doc's license to GNU FDL.

What matters to me is that other free software contributors are trying
to dictate what we can and cannot _talk about_ in the kernel sources.


> That is a minor detail (as you know).  My suggestion was an attempt to 
> balance the value of your admittedly useful document with the nature of 
> the endorsement.

The nature of the endorsement is implicit when BK is used by kernel
developers, not when a doc is present.

I'm very open to suggestions about combatting the notion that BK
is required for kernel devel.  Removing this doc may satisfy a few
people, but (a) it does not address the fundamental disagreement,
and (b) it is a slippery slope when we start removing docs because
we disagree with their subject on ideological grounds.

Even if I was completely anti-BitKeeper, pro-100%-open-source,
I would find Daniel's patch troubling because of "(b)"

Kernel source is no place to start drawing strict notions of what is
acceptable to discuss.  I weigh that assertion above a token gesture to
the silently-seething developers that does nothing to change the real
status quo.

	Jeff

Re: [PATCH] Remove Bitkeeper documentation from Linux tree

[Jeff Garzik]

On Mon, Apr 22, 2002 at 08:47:39PM +0100, Anton Altaparmakov wrote:
> Taking out the document only because it describes a non-free tool is 
> ridiculous. I don't object to the removal per se. I could live with that. I 
> object to the REASONS for removal.

Likewise.  This is perhaps saying it better than I have been...

	Jeff

Re: [PATCH] Remove Bitkeeper documentation from Linux tree

[Jonathan A. George]

Alexander Viro wrote:

>On Mon, 22 Apr 2002, Jonathan A. George wrote:
>
>>Not quite.  Linus uses BK as a tool to facilitate kernel development. 
>> However, he has not made BK _part_ _of_ _the_ _kernel_. ;-)  Obviously 
>>anyone can use any tool ON the kernel, but integrating into the kernel 
>>is something else.
>>
>>>Talk Linus out of using BitKeeper, and sure, I'll remove the doc.
>>>
>>No need.  His tools are his choice.  The kernel itself is ours not his; 
>> thus the distinction.
>>    
>>
>
>Bullshit.  His copy is his.  Mine is mine.  Yours is yours.  Each of
>us is perfectly within his rights when he adds whatever patches he
>likes.
>
Both statements are true.  Imagine that.

--Jonathan--

Re: [PATCH] Remove Bitkeeper documentation from Linux tree

[Alexander Viro]

On Mon, 22 Apr 2002, Jonathan A. George wrote:

> Alexander Viro wrote:

> >On Mon, 22 Apr 2002, Jonathan A. George wrote:

> >>No need.  His tools are his choice.  The kernel itself is ours not his; 
                                                              ^^^^^^^^^^^^
> >> thus the distinction.
> >
> >Bullshit.  His copy is his.  Mine is mine.  Yours is yours.  Each of
> >us is perfectly within his rights when he adds whatever patches he
> >likes.
> >
> Both statements are true.  Imagine that.

So where the hell do you or anybody else get the right to prohibit him adding
GPLed file to his copy?

RTFGPL.  Again, if you don't like said additions - fucking tough,
they are explicitly allowed by the license.

Re: [PATCH] Remove Bitkeeper documentation from Linux tree

[Rik van Riel]

On Mon, 22 Apr 2002, Jonathan A. George wrote:

> The BK documentation constitutes an implicit advertisement and
> endorsement of a product with a license which to many developers
> violates the spirit of open source software.

Yeah right, just like having having drivers/bluetooth/ is
an implicit advertisement for bluetooth hardware ;)

Could you please get over the fact that people will always
include useful stuff into the kernel ?

regards,

Rik
-- 
Bravely reimplemented by the knights who say "NIH".

http://www.surriel.com/		http://distro.conectiva.com/

Re: [PATCH] Remove Bitkeeper documentation from Linux tree

[Jonathan A. George]

Jeff Garzik wrote:

>On Mon, Apr 22, 2002 at 01:37:19PM -0500, Jonathan A. George wrote:
>  
>
>>The BK documentation constitutes an implicit advertisement and 
>>endorsement of a product with a license which to many developers 
>>violates the spirit of open source software.
>>
<snip>

>>The majority of the "silently seething" developers, I imagine, are only
>>gonna be satisfied when (a) BitKeeper is GPL'd or (b) Linus stops using
>>BitKeeper.  Both of these seem very remote possibilities at present.
>>
>>	Jeff
>>
Pax.  I think that Rik made a good point RE bluetooth support, and 
obviously no one with cache  in the kernel community cares enough about 
the non-free implications of using BK (an admittedly good tool) to keep 
the kernel source free of such endorsements.  In conceding I appreciate 
but disagree with your position.  Oh, and of course Al showed his ass on 
your behalf thus proving that intelligence and politeness are not related.

Respectfully,
Jonathan

[PATCH] Re: Remove Bitkeeper documentation from Linux tree

[Thunder from the hill]

Hi,

Sorry, I haven't checked email for a while. Putting this into an
external readme file might still leave some people complaining,
but if you wish I could do that either.
The last version was more or less a cut and paste from Larry, this one is 
a little edited...

Regards,
Thunder

diff -u bk-kernel-howto.txt~ bk-kernel-howto.txt
--- bk-kernel-howto.txt~	Mon Apr 22 16:11:37 2002
+++ bk-kernel-howto.txt	Mon Apr 22 16:15:35 2002
@@ -15,6 +15,15 @@
 "bk help <command>" or in X "bk helptool <command>" for reference
 documentation.
 
+    Also notice that BitKeeper  is not free software. You may use it
+    for free, subject  to the licensing rules  you can download from
+    <URL:http://www.bitkeeper.com/manpages/bk-blk-1.html>, but it is
+    not open source.  If you feel  strongly about 100% free software
+    tool chain, then don't use BitKeeper.  Linus has made very clear
+    that he will continue to accept and produce traditional "diff -Nur"
+    style patches.  It is explicitly not a  requirement that you use
+    BitKeeper to do  kernel development,  people may choose whatever
+    tool works best for them.
 
 BitKeeper Concepts
 ------------------

-- 
                                                  Thunder from the hill.
Not a citizen of any town.                   Not a citizen of any state.
Not a citizen of any country.               Not a citizen of any planet.
                         Citizen of our universe.

Re: [PATCH] Re: Remove Bitkeeper documentation from Linux tree

[Jeff Garzik]

On Mon, Apr 22, 2002 at 04:31:15PM -0600, Thunder from the hill wrote:
> Hi,
> 
> Sorry, I haven't checked email for a while. Putting this into an
> external readme file might still leave some people complaining,
> but if you wish I could do that either.
> The last version was more or less a cut and paste from Larry, this one is 
> a little edited...

What I'll do is take this and some of stuff sent to me privately
(spelling correction, peer review stuff) and update that.  I will
probably reword your intro here.

A README in that dir might not be a bad idea, but it (like the other
docs) will actually be useful.  It will include the "BK disclaimer" as
well as other stuff.

	Jeff