Re: [Gendispatch] I-D Action: draft-eggert-ietf-and-trust-00.txt

[Jay Daley <[e-mail]>]

> On 25 Oct 2022, at 07:19, John Levine <[e-mail]> wrote:
> 
> It appears that Jay Daley  <[e-mail]> said:
>> Maybe I’m missing something obvious, but how is it a “tremendous amount of work” to have a database (or even a spreadsheet) that lists every RFC and the
>> known licensing of each one, even if that is just ‘unknown’?  
>> 
>> I’m sure it would be straightforward to add that into Datatracker. 
> 
> I suppose we could create a database that says UNKNOWN for every RFC
> up to 3977 and TLP for everything after that, but more detail than
> that would indeed be a huge amount of work, providing no benefit
> that I can see.
> 
> Determining the copyright status of each pre-3978 RFC is a fact
> specific labor intensive project. Who are the authors? What if any
> copyright notice does the RFC have? Have the authors licensed rights?
> (in most cases, no.) If there were multiple authors and one licensed
> his rights, did he have authority to act for the rest of the authors?
> 
> Did the author(s) write the RFC as part of their job in which case the
> rights belong to whoever they were working for, or as individuals? If
> the former, do we have a license from the employer (unless it was ISI,
> probably not.) Were they working for the US government which does not
> assert copyright on its works? Etc., etc.
> 
> The vast majority of early RFCs are of purely historical interest.  No
> new RFC will use text from them, so their copyright status does not matter.
> 
> Occasionally someone does want to use old material, at which point the
> new authors look at all of those questions and try to figure out who
> has the rights and whether they can use it. This has been happening
> occasionally ever since we published RFC 3978.

So that I understand what you’re saying - if someone wanted to update an early RFC in a way that requires licensing of the content, they would have to take legal advice first, which would go into some depth, to determine if they could do so without infringing copyright.  Let’s say they did that by contacting the original authors (or their employers) who were happy to license that RFC to the Trust - you still don’t think it would be useful to record that somewhere so that others know that?  

You also mentioned previously that al RFCs published pre-1976 are in the public domain because they did not include a copyright notice - which is approximately the first 700 RFCs.  Would it not be helpful to have that information noted against each RFC as it’s not apparent from reading the text?

As a different example, take RFC 3978 itself.  If you read the text then at the top that’s simply copyright the Internet Society, and at the bottom is the full copyright statement that refers to BCP 78, which takes you to RFC 5378, which obliquely points to the Trust’s Legal Provision and then you find out what you can do with that RFC.  Would it not be so much simpler to have a list of answers to basic reuse questions alongside each RFC:
	- Can I do X … [Yes (subject to …), No, Unknown]

It does feel to me as if some people are so familiar with all of the complex details, they don’t recognise that from a less informed perspective there is a lot more information that could be made much more easily accessible.

Jay

> 
> I do not see why that would now be the trust's responsibility -- its
> job is to hold whatever IP assets the IETF owns, not to be its legal
> research department.
> 
> R's,
> John

-- 
Jay Daley
IETF Executive Director
[e-mail]