Re: [ECOS] Re: eCosCentric copyright hold in headers

Re: [ECOS] Re: eCosCentric copyright hold in headers

[Jonathan Larmour]

[ Moved to maintainers list ]

Alex Schuilenburg wrote:
> Andrew Lunn wrote:
> 
> [...]
> 
>>> Yep. It should all get blasted away when the FSF assignment happens 
>>> (don't ask).
>>
>>
>> Which is my i asked. We should keep track of this just in case...
> 
> 
> and for the benefit of all...
> 
> When I asked them face-to-face last week at ESCW, Red Hat were unable to 
> provide a date when the actual assignment to the FSF would happen. 
> Apparently everything has passed through legal and it is now a matter of 
> waiting for their engineer to become available to actually make the 
> assignment and header changes.
> 
> We could offer to do the changes for them but would need legal-sign off 
> from someone in Red Hat for all the files that would have their 
> copyright changed.  If we are going to help make this happen, I suggest 
> we provide Red Hat with a list of all the eCos files for which they hold 
> copyright and have them approve the list (in writing).  We can then make 
> the copyright changes and assignments to the FSF (including eCosCentric 
> and other maintainer held copyrights) in one go, which would make a lot 
> of sense.

I'd have hoped we wouldn't have to be responsible for that. While we could 
search for the copyright banner in files, I can't guarantee every file 
contains a Red Hat copyright that should have (even from the days when 
we^H^HRed Hat were working on eCos). And for the files without a Red Hat 
copyright banner, it would need careful identification to work out whether 
they are Red Hat's or someone elses (or indeed are mostly someone elses but 
may contain portions of RH code thus making it a derived work). The 
consequences of us making a mistake with the identification is painful; but 
if Red Hat makes the mistake it's nowhere near as bad and I believe intent 
does matter a lot in these circumstances.

It would be much much nicer if Red Hat could arrange some sort of blanket 
assignment, perhaps just by reference to the contents of the entire eCos 
CVS repository at ecos.sourceware.org. Or perhaps just list every 
repository file, irrespective of copyright and finetune the wording of the 
assignment so that it assigns any right and title that _may_ belong to Red 
Hat in the listed files. I'm no lawyer though.  I'd be more than willing to 
talk to them about ways it could be done though..... if they'll talk to me!

Jifl
-- 
eCosCentric    http://www.eCosCentric.com/    The eCos and RedBoot experts
--["No sense being pessimistic, it wouldn't work anyway"]-- Opinions==mine

Re: [ECOS] Re: eCosCentric copyright hold in headers

[Andrew Lunn]

Should we propose what the new banner looks like?

Is the basic format of the current banner  OK?

--- ChangeLog   2004-04-08 16:12:54.000000000 +0200
+++ ChangeLog.new       2004-04-08 16:17:45.000000000 +0200
@@ -3,6 +3,7 @@
 // -------------------------------------------
 // This file is part of eCos, the Embedded Configurable Operating System.
 // Copyright (C) 1998, 1999, 2000, 2001, 2002 Red Hat, Inc.
+// Copyright (C) 2004 Free Software Foundation, Inc.
 //
 // eCos is free software; you can redistribute it and/or modify it under
 // the terms of the GNU General Public License as published by the Free
@@ -26,9 +27,6 @@
 //
 // This exception does not invalidate any other reasons why a work based on
 // this file might be covered by the GNU General Public License.
-//
-// Alternative licenses for eCos may be arranged by contacting Red Hat, Inc.
-// at http://sources.redhat.com/ecos/ecos-license/
 // -------------------------------------------
 //####ECOSGPLCOPYRIGHTEND####
 //===========================================================================

The last bit clearly wants to go since its not been true for a long
while.

Should the original RedHat Copyright line be deleted as well? I guess
leaving it could confuse people into thinking the files are still
copyright RedHat as well as being FSF.. 

          Andrew

Re: [ECOS] Re: eCosCentric copyright hold in headers

[Alex Schuilenburg]

Jonathan Larmour wrote:

[...]
>> We could offer to do the changes for them but would need legal-sign 
>> off from someone in Red Hat for all the files that would have their 
>> copyright changed.  If we are going to help make this happen, I 
>> suggest we provide Red Hat with a list of all the eCos files for which 
>> they hold copyright and have them approve the list (in writing).  We 
>> can then make the copyright changes and assignments to the FSF 
>> (including eCosCentric and other maintainer held copyrights) in one 
>> go, which would make a lot of sense.
> 
> 
> I'd have hoped we wouldn't have to be responsible for that. While we 
> could search for the copyright banner in files, I can't guarantee every 
> file contains a Red Hat copyright that should have (even from the days 
> when we^H^HRed Hat were working on eCos). And for the files without a 
> Red Hat copyright banner, it would need careful identification to work 
> out whether they are Red Hat's or someone elses (or indeed are mostly 
> someone elses but may contain portions of RH code thus making it a 
> derived work). The consequences of us making a mistake with the 
> identification is painful; but if Red Hat makes the mistake it's nowhere 
> near as bad and I believe intent does matter a lot in these circumstances.

Hmmm, this is of concern. I thought all of eCos could be attributed to 
either Red Hat, eCosCentric or one of the maintainers.  Are you saying 
that there are files that are not copyright one of the above?

If there are files that do not have a copyright banner from their owner 
(Red Hat or otherwise), then would you also not have to get consent from 
the owners also on the assignment to the FSF.


> It would be much much nicer if Red Hat could arrange some sort of 
> blanket assignment, perhaps just by reference to the contents of the 
> entire eCos CVS repository at ecos.sourceware.org. Or perhaps just list 
> every repository file, irrespective of copyright and finetune the 
> wording of the assignment so that it assigns any right and title that 
> _may_ belong to Red Hat in the listed files. I'm no lawyer though.  I'd 
> be more than willing to talk to them about ways it could be done 
> though..... if they'll talk to me!

IMHO this is not going to happen unless somebody outside Red Hat drives 
this.  I suggest getting a draft together for all maintainers, Red Hat 
and eCosCentric representatives to sign that simply states that they 
assign copyright to all the files for which they may hold copyright in 
the ecos CVS repository to the FSF.

Come to think of it, my previous suggestion of listing the files is not 
a good idea as there is no guarantee that the list will not change while 
we are waiting for signature.  You will have to freeze contribs while 
waiting for signatures which will be could be forever in the case of Red 
Hat.

We could approach Red Hat with a formal written offer to do this so that 
the assignment to the FSF which they announced can actually take place. 
  If you want help from me, let me know.

-- Alex


> 
> Jifl

Re: [ECOS] Re: eCosCentric copyright hold in headers

[Jonathan Larmour]

Andrew Lunn wrote:
> Should we propose what the new banner looks like?
> 
> Is the basic format of the current banner  OK?

I'd like it to be changed actually. Some of the information parts of the 
banner are _very_ frequently poorly maintained and/or inaccurate and do us 
a disservice. I think the Authors/Contributors bits are ambiguous; and the 
Purpose/Description ditto. It's never really been clear what Date the Date 
is for. I think I fill these in more than most people, but still rarely. 
Contributions very frequently have inaccurate fields here, most annoyingly 
the author.

I think something like the following fields would be better/clearer:

Maintained by: <person responsible for _this_ file, irrespective of what it 
was derived from>

Contributors: <anyone else who's worked on it and wants a mention>
Derived from: <what file/package it was derived from if relevant
Description:

I propose dropping the date entirely partly due to the ambiguity and partly 
because it doesn't tell us anything we can't find out more accurately from 
the changelogs.

I propose dropping the mini-description underneath the filename at the top, 
as it is usually no better than the description further down.

It would require some cunning script-fu to try and squeeze the content 
there now into this style as best as possible, but it would be good to do 
these types of sweeping changes that touch all files once. Of course plenty 
of the existing content is wrong too, but we can fix that piecemeal. For 
example most of the "Maintained by" headers would be filled in on a package 
-wide basis.


> --- ChangeLog   2004-04-08 16:12:54.000000000 +0200
> +++ ChangeLog.new       2004-04-08 16:17:45.000000000 +0200
> @@ -3,6 +3,7 @@
>  // -------------------------------------------
>  // This file is part of eCos, the Embedded Configurable Operating System.
>  // Copyright (C) 1998, 1999, 2000, 2001, 2002 Red Hat, Inc.
> +// Copyright (C) 2004 Free Software Foundation, Inc.
>  //
>  // eCos is free software; you can redistribute it and/or modify it under
>  // the terms of the GNU General Public License as published by the Free
> @@ -26,9 +27,6 @@
>  //
>  // This exception does not invalidate any other reasons why a work based on
>  // this file might be covered by the GNU General Public License.
> -//
> -// Alternative licenses for eCos may be arranged by contacting Red Hat, Inc.
> -// at http://sources.redhat.com/ecos/ecos-license/
>  // -------------------------------------------
>  //####ECOSGPLCOPYRIGHTEND####
>  //===========================================================================
> 
> The last bit clearly wants to go since its not been true for a long
> while.

Absolutely. We just didn't want to touch every file multiple times. People 
doing "cvs update" won't like us otherwise :-).

> Should the original RedHat Copyright line be deleted as well? I guess
> leaving it could confuse people into thinking the files are still
> copyright RedHat as well as being FSF.. 

When the assignment is properly made the RH copyright _must_ disappear in fact.

I have a script that munges file headers, that was used in the RHEPL->GPL 
change and other changes before that. It's horrible as it has evolved 
piecemeal over time so it's a real hacky mess so I don't want to publicise 
it here :-). It deals with all the various comment characters ( /* */ 
versus // versus # versus ; versus dnl versus <!-- -->) and differing 
licenses though (BSD stack, host tools, etc.) IIRC.

A perl wizard may be able to do a better job quite quickly though, 
especially if we're to munge the other header fields too - this is one of 
the things perl is good at.

Jifl
-- 
eCosCentric    http://www.eCosCentric.com/    The eCos and RedBoot experts
--["No sense being pessimistic, it wouldn't work anyway"]-- Opinions==mine

Re: [ECOS] Re: eCosCentric copyright hold in headers

[Andrew Lunn]

On Thu, Apr 08, 2004 at 03:54:26PM +0100, Jonathan Larmour wrote:
> Andrew Lunn wrote:
> >Should we propose what the new banner looks like?
> >
> >Is the basic format of the current banner  OK?
> 
> I'd like it to be changed actually. Some of the information parts of the 
> banner are _very_ frequently poorly maintained and/or inaccurate and do us 
> a disservice. I think the Authors/Contributors bits are ambiguous; and the 
> Purpose/Description ditto. It's never really been clear what Date the Date 
> is for. I think I fill these in more than most people, but still rarely. 
> Contributions very frequently have inaccurate fields here, most annoyingly 
> the author.
> 
> I think something like the following fields would be better/clearer:
> 
> Maintained by: <person responsible for _this_ file, irrespective of what it 
> was derived from>
> 
> Contributors: <anyone else who's worked on it and wants a mention>
> Derived from: <what file/package it was derived from if relevant
> Description:
> 
> I propose dropping the date entirely partly due to the ambiguity and partly 
> because it doesn't tell us anything we can't find out more accurately from 
> the changelogs.
> 
> I propose dropping the mini-description underneath the filename at the top, 
> as it is usually no better than the description further down.

Just to make that a bit more concrete i've made an example. 

//==========================================================================
//
//      common/kapi.cxx
//
//==========================================================================
//####ECOSGPLCOPYRIGHTBEGIN####
// -------------------------------------------
// This file is part of eCos, the Embedded Configurable Operating System.
// Copyright (C) 2004 Free Software Foundation, Inc.
//
// eCos is free software; you can redistribute it and/or modify it under
// the terms of the GNU General Public License as published by the Free
// Software Foundation; either version 2 or (at your option) any later version.
//
// eCos is distributed in the hope that it will be useful, but WITHOUT ANY
// WARRANTY; without even the implied warranty of MERCHANTABILITY or
// FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
// for more details.
//
// You should have received a copy of the GNU General Public License along
// with eCos; if not, write to the Free Software Foundation, Inc.,
// 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA.
//
// As a special exception, if other files instantiate templates or use macros
// or inline functions from this file, or you compile this file and link it
// with other works to produce a work based on this file, this file does not
// by itself cause the resulting work to be covered by the GNU General Public
// License. However the source code for this file must still be made available
// in accordance with section (3) of the GNU General Public License.
//
// This exception does not invalidate any other reasons why a work based on
// this file might be covered by the GNU General Public License.
//
// -------------------------------------------
//####ECOSGPLCOPYRIGHTEND####
//==========================================================================
//#####DESCRIPTIONBEGIN####
//
// Maintained by:   nickg, dsm
// Contributors:        nickg
// Derived from: 
// Description: C++ implementation of the C API
//
//####DESCRIPTIONEND####
//
//==========================================================================

I don't think this is going to be easy. Does dsm maintain this file?

     Andrew

Re: [ECOS] Re: eCosCentric copyright hold in headers

[Jonathan Larmour]

Andrew Lunn wrote:
> //
> // Maintained by:   nickg, dsm
> // Contributors:        nickg
> // Derived from: 
> // Description: C++ implementation of the C API
> //
> //####DESCRIPTIONEND####
> //
> //==========================================================================
> 
> I don't think this is going to be easy. Does dsm maintain this file?

My idea was no he wouldn't... "Maintained by" would be how maintains it 
now. Historical info is something for ChangeLogs and cvs logs, not banners, 
and not when it's no practical use.

Of course in your example someone would eventually have to clean up that 
file (unless the script explicitly watched for user dsm - there aren't that 
many names mentioned). There would probably want to also be a NO-MAINTAINER 
thing for something which no-one is claiming. That of course will help 
people realise where bit-rot is, so this in itself is useful too.

Jifl
-- 
eCosCentric    http://www.eCosCentric.com/    The eCos and RedBoot experts
--["No sense being pessimistic, it wouldn't work anyway"]-- Opinions==mine

Re: [ECOS] Re: eCosCentric copyright hold in headers

[Jonathan Larmour]

Alex Schuilenburg wrote:
> Jonathan Larmour wrote:
> 
> [...]
> 
>>> We could offer to do the changes for them but would need legal-sign 
>>> off from someone in Red Hat for all the files that would have their 
>>> copyright changed.  If we are going to help make this happen, I 
>>> suggest we provide Red Hat with a list of all the eCos files for 
>>> which they hold copyright and have them approve the list (in 
>>> writing).  We can then make the copyright changes and assignments to 
>>> the FSF (including eCosCentric and other maintainer held copyrights) 
>>> in one go, which would make a lot of sense.
>>
>>
>>
>> I'd have hoped we wouldn't have to be responsible for that. While we 
>> could search for the copyright banner in files, I can't guarantee 
>> every file contains a Red Hat copyright that should have (even from 
>> the days when we^H^HRed Hat were working on eCos). And for the files 
>> without a Red Hat copyright banner, it would need careful 
>> identification to work out whether they are Red Hat's or someone elses 
>> (or indeed are mostly someone elses but may contain portions of RH 
>> code thus making it a derived work). The consequences of us making a 
>> mistake with the identification is painful; but if Red Hat makes the 
>> mistake it's nowhere near as bad and I believe intent does matter a 
>> lot in these circumstances.
> 
> 
> Hmmm, this is of concern. I thought all of eCos could be attributed to 
> either Red Hat, eCosCentric or one of the maintainers.  Are you saying 
> that there are files that are not copyright one of the above?

No, I mean there are files which don't explicitly have a banner saying who 
owns the copyright, so it makes it difficult to use the copyright banners 
as a method of determining fully which files Red Hat needs to assign.

> If there are files that do not have a copyright banner from their owner 
> (Red Hat or otherwise), then would you also not have to get consent from 
> the owners also on the assignment to the FSF.

Maybe, but it becomes silly if the sole point of the exercise is that it 
all ends up as Copyright FSF and nothing else. Since we know that all the 
copyright is held by RH, eCosCentric and the maintainers it's naff if 
things have to be held up to work out some bureaucratic list; hence my 
suggested possibly viable alternatives.

>> It would be much much nicer if Red Hat could arrange some sort of 
>> blanket assignment, perhaps just by reference to the contents of the 
>> entire eCos CVS repository at ecos.sourceware.org. Or perhaps just 
>> list every repository file, irrespective of copyright and finetune the 
>> wording of the assignment so that it assigns any right and title that 
>> _may_ belong to Red Hat in the listed files. I'm no lawyer though.  
>> I'd be more than willing to talk to them about ways it could be done 
>> though..... if they'll talk to me!
> 
> 
> IMHO this is not going to happen unless somebody outside Red Hat drives 
> this.  I suggest getting a draft together for all maintainers, Red Hat 
> and eCosCentric representatives to sign that simply states that they 
> assign copyright to all the files for which they may hold copyright in 
> the ecos CVS repository to the FSF.

As soon as the FSF starts replying to my mails! There's still the issue of 
the publically stated "guarantee" we wanted from them not to immediately 
strip away the GPL exception. It looked like we were very close to agreeing 
this - it just all went very silent that's all. Right now I know that the 
FSF is not accepting assignments for eCos. Any agreement will have to be 
reviewed by either or both of the RH and FSF legal counsels I'm sure, and 
probably constructed by one of them.

> Come to think of it, my previous suggestion of listing the files is not 
> a good idea as there is no guarantee that the list will not change while 
> we are waiting for signature.  You will have to freeze contribs while 
> waiting for signatures which will be could be forever in the case of Red 
> Hat.
> 
> We could approach Red Hat with a formal written offer to do this so that 
> the assignment to the FSF which they announced can actually take place. 
>  If you want help from me, let me know.

It's more a question of getting a dialogue with anyone. I'm sure whatever 
the issues are we can work them out; but if we can't even start discussing 
them we're no further on.

I wonder, in all seriousness, if we (or some of us) chipped in to a 
donation to the FSF so we can say "Talk to us and we'll give you a 
donation.". I wouldn't have thought it would be too difficult to rustle up 
$150 or so given the current exchange rate. Maybe eCosCentric would like to 
chip in too :). I'm willing to contribute personally anyway, just to give 
myself some peace from this palaver!

Jifl
-- 
eCosCentric    http://www.eCosCentric.com/    The eCos and RedBoot experts
--["No sense being pessimistic, it wouldn't work anyway"]-- Opinions==mine

Re: [ECOS] Re: eCosCentric copyright hold in headers

[Alex Schuilenburg]

Jonathan Larmour wrote:

> I wonder, in all seriousness, if we (or some of us) chipped in to a 
> donation to the FSF so we can say "Talk to us and we'll give you a 
> donation.". I wouldn't have thought it would be too difficult to rustle 
> up $150 or so given the current exchange rate. Maybe eCosCentric would 
> like to chip in too :). I'm willing to contribute personally anyway, 
> just to give myself some peace from this palaver!

Yup, gladly, anything to move this along as we waste a lot of time with 
the problems of them not doing the assignments and getting the RH 
assignment sorted out.

When the books for 2003/2004 close we can see what is left for 
donations.  $150 should be doable, if only we could guarantee that it 
would make the FSF listen...

-- Alex

Re: [ECOS] Re: eCosCentric copyright hold in headers

[Frank Ch. Eigler]

[-- Attachment #1: Type: text/plain, Size: 272 bytes --]

Hi -

Sorry to bud into this old thread.  I gather that one reason for
the Red Hat assignment delay is to identify those files related
to RedBoot, and to exclude them from the process.  I don't know
this plan's novelty or rationale or degree of official blessing.

- FChE

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]