Re: ISSUE-27: rel-ownership - Chairs Solicit Proposals

On Jan 22, 2010, at 11:29 AM, Larry Masinter wrote:

> It would be harmful to the community and the benefits of open standards
> to allow a proprietary organization to freely take the results of 
> an open standards process, make changes to the standard to advantage
> their proprietary technology, and then publish the result as if it 
> were the authoritative version of the open standard, in a way that
> would confuse users and developers into using the proprietary version
> when they were expecting the public one. It would allow and encourage
> "embrace and pollute". Whether the standards organization itself
> is in a position to pursue remedies is not the issue -- it's the
> question of whether this practice is condoned by the community.
> 
> "Free Software" is a great when technologies and software can
> compete on their technical merit and end-user benefit. But
> IANA allowing someone to publish a version of the actual registered
> values, but to change the meanings (oh, say, define "ISO-8859-1"
> to mean "actually Windows 1252") would be harmful.
> 
> So I think it is reasonable to disallow downstream recipients
> to make arbitrary changes to registries or registered values.


This is often the kind of reasoning used to justify restrictive copyright license terms on standards products - that someone might use permission to distribute modified copies in order to deviate from the standard or otherwise undermine it

 In the case of use of these registry lists inside a software product, I don't think the argument really works. It's trivial to have an external file with additional entries, blacklisted entries, or modifications to existing entries, or to make such changes in code, given a pristine list. So the hypothetical proprietary malefactor will not be stopped by restrictions on redistribution. However, forbidding changes (even if verbatim redistribution is allowed) is incompatible with many open source / free software licenses. This effectively prevents free software projects from embedding even verbatim copies. So it's harder for free software to directly use these registries to conform to the spec. Not only is validator.nu free software, but so are the engines in 3 of the 5 most popular browsers, so this is a real problem, not just theoretical. Meanwhile, I don't think anyone is eager to deliberately mislead about the contents of IANA registries.

In conclusion, allowing redistribution only of verbatim copies does essentially nothing to make it harder for a proprietary technology to deviate, while actually making it more work for an open source project to conform.

> Perhaps the restrictions go too far in terms of IANA registries,
> and should allow Henri to publish the "Henri-registry-set"
> which is derived from the IANA registry but includes Henri's
> favorite additional values, or which allows downstream users
> to do what they want with it (as long as they don't confuse
> anyone into think it is the IANA registry) that would be fine,
> but I think *some* restrictions are appropriate and reasonable.

If the restricted were just that you can't claim a modified version is the true IANA registry, or that you must indicate any changes, I believe that would be Free Software compatible. And it would probably be a more helpful rule than only allowing verbatim redistribution.

Regards,
Maciej

Received on Friday, 22 January 2010 23:58:51 UTC