Re: pg_rewind, a tool for resynchronizing an old master after failover

On Thu, May 23, 2013 at 01:48:24PM -0400, Heikki Linnakangas wrote:
> On 23.05.2013 08:03, Simon Riggs wrote:
> >On 23 May 2013 12:10, Heikki Linnakangas<hlinnakangas(at)vmware(dot)com> wrote:
> >
> >>Please take a look: https://github.com/vmware/pg_rewind
> >
> >The COPYRIGHT file shows that VMware is claiming copyright on unstated
> >parts of the code for this. As such, its not a normal submission to
> >the PostgreSQL project, which involves placing copyright with the
> >PGDG.
>
> We have a lot of code in PostgreSQL source tree with different
> copyright notices, and there's no problem with that as long as the
> coe is licensed under the PostgreSQL license. For patches that add

Really? Where? I think we have removed them all, as far as I know.
A quick grep shows:

$ grep -r 'Portions Copyright'|egrep -v 'Global|Regents'
./src/backend/regex/regexport.c: * Portions Copyright (c) 1998, 1999 Henry Spencer
./src/backend/regex/regprefix.c: * Portions Copyright (c) 1998, 1999 Henry Spencer
./src/include/regex/regexport.h: * Portions Copyright (c) 1998, 1999 Henry Spencer
./src/include/getopt_long.h: * Portions Copyright (c) 1987, 1993, 1994
./src/bin/pg_dump/pg_backup_directory.c: * Portions Copyright (c) 2000, Philip Warner
./src/port/getopt_long.c: * Portions Copyright (c) 1987, 1993, 1994
./src/port/getopt_long.c: * Portions Copyright (c) 2003

Can someone comment on the "Philip Warner" item? Would someone contact
him to clarify we can remove the mention? CC'ing him.

> or modify code in PostgreSQL, we generally have copyright notices
> with just PGDG, to avoid having a long list of copyright notices of
> a lot of companies and individuals on every file. I'm no lawyer, but
> I believe there's no difference from the legal point of view.

Probably, but some mentions can cause concern when our code is reviewed
by companies, so simplicity is good.

> >As a result, while it sounds interesting, people should be aware of
> >that and I suggest we shouldn't discuss that code on this list, to
> >avoid any disputes should we decide to include a similar facility in
> >core Postgres in the future.
>
> That's just paranoia. There are a lot of tools out there on
> pgfoundry, with various copyright holders and even difference
> licenses, and it's fine to talk about all those on this list.
> Besides, the code is licensed under the PostgreSQL license, so if
> someone decides we should have this e.g in contrib, you can just
> grab the sources and commit. Thirdly, there's no reason to refrain
> from even discussing this, even if someone would include a similar
> facility in core Postgres - this is about copyrights, not patents
> (and yes, this contribution has been cleared by VMware legal
> department; VMware doesn't hold any patents on this)

I think Simon has a good point, as VMWare has asserted patents on some
changes to their version of Postgres in the past, so if the copyright
mentions VMWare, we can't assume it is patent-free. Just the fact you
had to check with the VMware legal department verifies there is cause
for concern about things coming from VMWare. In fact, I am curious what
level of contribution requires a legal check, but I am not sure you can
even share that information.

Anyway, I would love to think we don't need to worry about this, but I
think we do --- not in this case, but in general. I acknowledge that
VMWare has been disciplined in share only patent-free information, at
the community's request.

--
Bruce Momjian <bruce(at)momjian(dot)us> http://momjian.us
EnterpriseDB http://enterprisedb.com

+ It's impossible for everything to be true. +