BIP 2 revival and rework

I've revived BIP 2 (from Deferred Status) and given it some updates. Most notably, I have reworked it to be a *replacement* for BIP 1 rather than an addendum.

https://github.com/luke-jr/bips/blob/bip0002_squash/bip-0002.mediawiki

Please review it. If things go well, hopefully we can get this done by Christmas. ;)

Other recent changes include: * OPL will no longer be an acceptable license. Many in the community feel that prohibiting publication is unacceptable for BIPs, and I haven't heard any arguments in favour of allowing it. * Accepted Status has been renamed to Proposed. The name "Accepted" seems a constant source of confusion since it requires only action from the author. * Non-image auxiliary files are permitted in the bip-XXXX subdirectory. This was already the norm despite BIP 1. * Email addresses are now required for authors. The Travis script has been enforcing this for months now already. * The Post-History header may be provided as a link instead of a simple date. A few BIPs were already doing this. * Markdown format is no longer permitted for BIPs. I don't see the point in allowing multiple formats, and so far we've been fine with just MediaWiki. * The Resolution header has been dropped, as it is not applicable to a decentralised system where no authority exists to make final decisions.

Other changes already in the previous draft of BIP 2: * An implementation is now required (when applicable) before BIPs can proceed to Proposed Status. * BIP Comments are newly introduced. * The License preamble headers have been added.

Thanks,

Luke

On Saturday, 24 September 2016 06:36:00 CEST Luke Dashjr via bitcoin-dev wrote:

> * OPL will no longer be an acceptable license. Many in the community feel
> that prohibiting publication is unacceptable for BIPs, and I haven't
> heard any arguments in favour of allowing it.

My suggestion would be that we replace OPL as an allowed license with one or two Creative Commons licenses. Following the suggestion from the OPL creators themselves. According to Wikipedia;

> Open Publication License was created by the Open Content Project in 1999
> as public copyright license for documents. The license was superseded
> in 2003/2007 by the Creative commons licenses.

I'd suggest saying that "Share alike" is required and "Attribution" is optional.

Executive summary; give the user the choice (next to public domain) between CCO and BY-SA see; https://en.wikipedia.org/wiki/ Creative_Commons_license#Seven_regularly_used_licenses

On Sat, Sep 24, 2016 at 11:41 AM, Tom via bitcoin-dev <bitcoin-dev@lists•linuxfoundation•org> wrote:

> I'd suggest saying that "Share alike" is required and "Attribution" is
> optional.

Please note there is no CC license that requires SA and at the same time has BY as an option.

Generally, I think CC0 is best suited as license for BIPs. If authors are scared that they won't get proper attribution, they can choose MIT/BSD or CC-BY. Other than that I don't think that more restrictive licenses are suitable for BIPs. The BIP repo seems like the wrong place to promote Open Access (e.g. by choosing a CC-BY-SA license). BIP 2 allows such licenses, but does not recommend them, which is fine.

I think that BIP 2 in its current form ( https://github.com/bitcoin/bips/blob/master/bip-0002.mediawiki @6e47447b ) looks good and addressed the feedback which was accumulated last year. If there are no objections I'd suggest to move forward with BIP 2 in the next couple of days/weeks.

Marco

On Saturday, 15 October 2016 12:11:02 CEST Marco Falke wrote:

> On Sat, Sep 24, 2016 at 11:41 AM, Tom via bitcoin-dev
> <bitcoin-dev@lists•linuxfoundation•org> wrote:
> > I'd suggest saying that "Share alike" is required and "Attribution" is
> > optional.
>
> Please note there is no CC license that requires SA and at the same
> time has BY as an option.
>
> Generally, I think CC0 is best suited as license for BIPs. If authors
> are scared that they won't get proper attribution, they can choose
> MIT/BSD or CC-BY.

My suggestion (sorry for not explaining it better) was that for BIPS to be a public domain (aka CC0) and a CC-BY option and nothing else.

I like you agree with that part, but I see you added two licenses. Do you have a good reason to add MIT/BSD to that list? Otherwise I think we agree. Using code-specific licenses (including the GPL) for documentation and specifically a specification is a really poor fit and doens't make much sense.

> Other than that I don't think that more restrictive
> licenses are suitable for BIPs. The BIP repo seems like the wrong
> place to promote Open Access (e.g. by choosing a CC-BY-SA license).
> BIP 2 allows such licenses, but does not recommend them, which is
> fine.
>
> I think that BIP 2 in its current form (
> https://github.com/bitcoin/bips/blob/master/bip-0002.mediawiki
> @6e47447b )

Well, it has this sentence;

> This BIP is dual-licensed under the Open Publication License and
> BSD 2-clause license.

Which is a bit odd in light of the initial email from Luke that suggested we drop the Open Publication License and we use the CC ones instead in addition to the public domain one.

Marco:

> looks good and addressed the feedback which was
> accumulated last year. If there are no objections I'd suggest to move
> forward with BIP 2 in the next couple of days/weeks.

Thats odd, you just stated you like the public domain (aka CC0) license, yet
you encourage the BIP2 that states we can no longer use public domain for
BIPs... Did you read it?
It says;
 «Public domain is not universally recognised as a legitimate action, thus
  it is inadvisable.» [1]

Also; This list has not seen a lot of traffic, if you want to make sure people keep using the BIP process, I think you need to reach out to the rest of the community and make sure this has been heard and discussed. Moving forward the way it is now will likely deminish the importance of the BIP process.

I strongly suggest people make very clear any and all changes that are proposed and defend each of them with reasons why you want to change things.

1) if you write as a rationale "In some jurisdictions, public domain is not recognised as a legitimate legal action" then you can at least name those jurisdictions and explain how they *do* support things like GPL. Burden of proof is on the man who wants to change things. It looks fishy when lawyers disagree. See the CC wikipedia page; "public domain: cc0 Freeing content globally without restrictions"

On Sat, Oct 15, 2016 at 1:00 PM, Tom Zander via bitcoin-dev <bitcoin-dev@lists•linuxfoundation•org> wrote:

> My suggestion (sorry for not explaining it better) was that for BIPS to be a
> public domain (aka CC0) and a CC-BY option and nothing else.

Indeed, we agree that BIPs should be licensed as permissive as possible. Still, I wonder why you chose otherwise with BIP 134. (Currently OPL and CC-BY-SA)

> I like you agree with that part, but I see you added two licenses.
> Do you have a good reason to add MIT/BSD to that list? Otherwise I think we
> agree.

Licenses that only require attribution are generally compatible with each other. I don't think we should pick one and only promote/endorse this one. Let's just leave the decision to the BIP author.

> Well, it has this sentence;
>
>> This BIP is dual-licensed under the Open Publication License and
>> BSD 2-clause license.
>
> Which is a bit odd in light of the initial email from Luke that suggested we
> drop the Open Publication License and we use the CC ones instead in addition
> to the public domain one.

I am pretty sure this is required to host the current text of BIP 2 in the repo, as currently BIP 1 still applies and still requires for all BIPs either OPL or PD, which is one of the reasons I think we should move forward with BIP 2 or amending BIP 1.

> Marco:
>> looks good and addressed the feedback which was
>> accumulated last year. If there are no objections I'd suggest to move
>> forward with BIP 2 in the next couple of days/weeks.
>
> Thats odd, you just stated you like the public domain (aka CC0) license, yet
> you encourage the BIP2 that states we can no longer use public domain for
> BIPs... Did you read it?
> It says;
>  «Public domain is not universally recognised as a legitimate action, thus
>   it is inadvisable.» [1]

BIP 2 does not forbid you to release your work under PD in legislations where this is possible. None of the licenses mentioned in BIP 2 is exclusive, so you can choose as many options as you like. One of the goals of BIP 2 is to no longer allow PD as the only copyright option.

>
> Also;
> This list has not seen a lot of traffic, if you want to make sure people keep
> using the BIP process, I think you need to reach out to the rest of the
> community and make sure this has been heard and discussed.
> Moving forward the way it is now will likely deminish the importance of the
> BIP process.
>
> I strongly suggest people make very clear any and all changes that are
> proposed and defend each of them with reasons why you want to change things.
>
>
> 1) if you write as a rationale "In some jurisdictions, public domain is not
> recognised as a legitimate legal action" then you can at least name those
> jurisdictions and explain how they *do* support things like GPL. Burden of
> proof is on the man who wants to change things.
> It looks fishy when lawyers disagree. See the CC wikipedia page;
>  "public domain: cc0 Freeing content globally without restrictions"

Luke is the BIP champion of BIP 2, so please cc him if you have suggestions on how to improve the process of gathering community consensus.

Marco

On Saturday, October 15, 2016 11:00:35 AM Tom Zander via bitcoin-dev wrote:

> My suggestion (sorry for not explaining it better) was that for BIPS to be
> a public domain (aka CC0) and a CC-BY option and nothing else.
>
> I like you agree with that part, but I see you added two licenses.
> Do you have a good reason to add MIT/BSD to that list? Otherwise I think we
> agree.

BIPs often should include code.

> Well, it has this sentence;
>
> > This BIP is dual-licensed under the Open Publication License and
> > BSD 2-clause license.
>
> Which is a bit odd in light of the initial email from Luke that suggested
> we drop the Open Publication License and we use the CC ones instead in
> addition to the public domain one.

The "real" license in this case is the BSD 2-clause. However, BIP 1 only allows OPL and public domain, so BIP 2 is available under OPL as well so that it is acceptable before/until it activates also.

> Thats odd, you just stated you like the public domain (aka CC0) license,
> yet you encourage the BIP2 that states we can no longer use public domain
> for BIPs... Did you read it?

CC0 and public domain are two different things.

> This list has not seen a lot of traffic, if you want to make sure people
> keep using the BIP process, I think you need to reach out to the rest of
> the community and make sure this has been heard and discussed.
> Moving forward the way it is now will likely deminish the importance of the
> BIP process.

Yes, you're right. I'll post to Lightning-dev and libbitcoin's list about BIP 2. If you're aware of any other Bitcoin development discussion groups, could you please bring BIP 2 to their attention so it gets wider review?

> 1) if you write as a rationale "In some jurisdictions, public domain is not
> recognised as a legitimate legal action" then you can at least name those
> jurisdictions and explain how they *do* support things like GPL. Burden of
> proof is on the man who wants to change things.

As I understand it, presently France and Germany do not recognise public domain as a possible status. GPL is merely a copyright license, so it should be valid anywhere copyright laws exist.

Luke

On Saturday, 15 October 2016 14:12:09 CEST Marco Falke wrote:

> On Sat, Oct 15, 2016 at 1:00 PM, Tom Zander via bitcoin-dev
>
> <bitcoin-dev@lists•linuxfoundation•org> wrote:
> > My suggestion (sorry for not explaining it better) was that for BIPS to
> > be a public domain (aka CC0) and a CC-BY option and nothing else.
>
> Indeed, we agree that BIPs should be licensed as permissive as
> possible. Still, I wonder why you chose otherwise with BIP 134.
> (Currently OPL and CC-BY-SA)

OPL was the only allowed option apart from CC0.

I dual licensed it so future acceptance of the CC-BY-SA one may mean someone can just remove the OPL from the BIP and no futher action or permission is needed from all the authors.

> BIP 2 does not forbid you to release your work under PD in
> legislations where this is possible

It does, actually.

> One
> of the goals of BIP 2 is to no longer allow PD as the only copyright
> option.

That's odd as PD was never the only copyright option.

On Sat, Oct 15, 2016 at 4:21 PM, Tom Zander via bitcoin-dev <bitcoin-dev@lists•linuxfoundation•org> wrote:

>> > My suggestion (sorry for not explaining it better) was that for BIPS to
>> > be a public domain (aka CC0) and a CC-BY option and nothing else.
>>
>> Indeed, we agree that BIPs should be licensed as permissive as
>> possible. Still, I wonder why you chose otherwise with BIP 134.
>> (Currently OPL and CC-BY-SA)
>
> OPL was the only allowed option apart from CC0.

I think you are misunderstanding what is allowed and what is required...

BIP1: "Each BIP must either be explicitly labelled as placed in the public domain (see this BIP as an example) or licensed under the Open Publication License"

So BIP1 *requires* PD or OPL but does not forbid other licenses. For example, you are free to multi license OPL (and additionally: BSD, MIT, CC0, ...)

BIP2: "Each new BIP must identify at least one acceptable license in its preamble."

So BIP2 *requires* an acceptable license but does not forbid other choices. For example, you are free to choose: BSD (and additionally: PD, CC-BY-SA, WTFPL, BEER, ...)

>> BIP 2 does not forbid you to release your work under PD in
>> legislations where this is possible
>
> It does, actually.

Huh, I can't find it in the text I read. The text mentions "not acceptable", but I don't read that as "forbidden".

>
>> One
>> of the goals of BIP 2 is to no longer allow PD as the only copyright
>> option.
>
> That's odd as PD was never the only copyright option.

Right. Though, up to now the majority of the BIP authors chose PD as the only option.

Marco

On Saturday, 15 October 2016 17:02:30 CEST Marco Falke wrote:

> >> BIP 2 does not forbid you to release your work under PD in
> >> legislations where this is possible
> >
> > It does, actually.
>
> Huh, I can't find it in the text I read. The text mentions "not
> acceptable", but I don't read that as "forbidden".

You suggest that a person can dual license something under both CC-BY-SA as well as under public domain. That means you don't understand copyright,

See, all licenses are based on you having copyright. In contrast; public domain is not a license, it means a certain text does not have copyright. Public domain is the lack of copyright.

One text can not at the same time have copyright and not have copyright, making your assumption impossible.

Hence, with PD not explicitly being allowed, you can't use PD.

Personally I prefer copyleft licenses, so the lack of PD is fine with me. The lack of a good copyleft we can use in BIPs is what got me involved in this discussion in the first place.