On Sun, 25 Nov 2012, Jonas Sicking wrote:
> On Sun, Nov 25, 2012 at 12:38 PM, Ian Hickson <i...@hixie.ch> wrote:
> > On Sun, 25 Nov 2012, David Bruant wrote:
> >>
> >> The intent is clear: the WHATWG publishes documents in the public 
> >> domain for very good reason. Anyone (W3C included!) can reuse them 
> >> under close to no condition, not even credit.
> >
> > I can speak pretty authoritatively to the intent, if that's what you 
> > are interested in.
> >
> > The relevant philosophy in the WHATWG context is multi-pronged:
> >
> > 1: Specs should be reusable in software, documentation, tutorials, and 
> > the like, without any barrier, whether free software or proprietary 
> > software, whether in books printed for money or FAQs that are 
> > themselves free to copy, whether in online courses with $10,000 entry 
> > fees or demos on street corners that are organised by marketing 
> > departments.
> >
> > 2: A spec author can "go bad" without realising it, so it should be 
> > possible to fork a specification if that happens, without the author 
> > having any control over this.
> >
> > 3: Forking specifications, publishing multiple copies of 
> > specifications, and publishing easy-to-find-with-a-search-engine 
> > snapshots of specifications, are all things that hurt interoperability 
> > by making implementors reference different requirements. The only time 
> > that forking a specification is justified is #2 above.
> >
> > We use open licenses on our specifications because of #1 and #2. We 
> > can't legally prevent #3 while allowing #1 and #2, so we rely on 
> > common sense and good faith to achieve #3.
> 
> I'm not sure in what capacity you are writing this. [...] I forget 
> exactly what policies govern WHATWG, but I don't know if the above can 
> be considered an official WHATWG policy.

I don't know what "official" would mean here. I just meant the intent that 
is behind my (and Anne's, I believe) advocacy of open licensing for 
specifications.


> However I'll note that not everyone at least at Mozilla agree with #3.

#3 is actually the most empirically testable one, at least the first 
sentence of it. Given the number of e-mails I get from implementors asking 
me questions with links to outdated snapshots of specs they found via 
search engines, where their question was already answered by the editor's 
draft of that spec, I don't really see how it can be controversial. :-)

What do you think justifies having multiple copies of a spec?

-- 
Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'

Reply via email to