Re: Quick update on WebIDL "Level 1"

[marcos]

> On 11 Jul 2016, at 10:45 PM, Yves Lafon  wrote:
> 
> The goal of publishing this as a REC is not to have a final document nor to 
> please only
> the lawyers. The goal is to provide a document that contains the parts of the 
> WebIDL
> syntax that are implemented, and the associated implemented ES-binding, as a 
> guide
> for spec authors that are not following the main WebIDL spec evolutions (as 
> not everybody 
> has your knowledge of what is or is not usable in WebIDL).

Yes, but that's precisely the point. If something is not interoperable in the 
spec, then it should be fixed. Now we are back at Domenic's email. No spec 
editors should be, or will be, referencing v1. It's simple pointless to think 
otherwise. Look, all browser vendors already implement promise-using 
WebIDL-based APIs, which means that they've already had to implement v2 
features. 

I think a large segment of the WG has made it pretty clear that's it harmful to 
pretend that WebIDL 1 has any value to anyone but patent lawyers. Technically, 
it's just going to be bit-rotting trash sitting on TR (as you even acknowledge 
below). 

> 
> The -1 spec explicitly states that people wanting to implement WebIDL are 
> invited to read
> the main WebIDL specification (that, ideally, should be automatically 
> published as /TR/WebIDL ) because yes
> WebIDL-1 is _not_ the WebIDL specification, just a frozen snapshot of what 
> was implemented as the 
> time of publication, not more than that, and bound to be replaced by a 
> subsequent level later on.

Yes, but it's grossly obsolete and no one but patent lawyers should be, or will 
be, looking at it. So why bother putting it on TR? 

You can't seriously say that anyone writing specs would be using it to 
implement against - not even as joke. It has zero value from a technical 
perspective - yet huge value from an IPR perspective. 

I'm all for getting the IPR protection, but let's stop with putting useless 
things on TR. 

Re: Quick update on WebIDL "Level 1"

[Yves Lafon]

> On 10 Jul 2016, at 16:35, Marcos Caceres  wrote:
> 
> On July 9, 2016 at 6:24:56 AM, Domenic Denicola (d...@domenic.me) wrote:
>> From: Travis Leithead [mailto:travis.leith...@microsoft.com]
>> 
>>> The purpose of the “Level 1” document is to serve as a stable reference for 
>>> W3C specs that
>> link to WebIDL. It contains a subset of the WebIDL syntax that is considered 
>> stable (as
>> verified by interoperable tests). Implementations should not use the Level 1 
>> document
>> as a guide, but instead track changes to the editors draft. We expect to 
>> follow-up Level
>> 1 with a Level 2 as additional editor’s draft syntax and behavior 
>> stabilizes, are implemented
>> as part of other specs, and shown to be interoperable.
>> 
>> Why is it acceptable for specs to reference a version of Web IDL that nobody 
>> should implement?
> 
> This is a totally valid question, but we've had this debate 1001
> times. Perhaps a better question is: how can we get patent protection
> (making this subset of WebIDL royalty free for society), but without
> harming the ecosystem by confusing implementers and developers by
> publishing on the "/TRash" space (as most of us now unfortunately
> referring to it).
> 
> We need a way to clearly indicate that, for a subset of documents,
> RECs on TR represent a royalty free set of ideas (as kindly and
> honorably granted by the W3C Membership) - and should only be referred
> to by patent lawyers and government officials. That it's for those
> groups should be stated and promoted proudly, not disparagingly. And,
> that implementers should be looking at the living document instead.
> The value of TR need not be diminished - in fact: it should be
> correctly used to published the documents that enshrine the royalty
> free status of particular specifications.

The goal of publishing this as a REC is not to have a final document nor to 
please only
the lawyers. The goal is to provide a document that contains the parts of the 
WebIDL
syntax that are implemented, and the associated implemented ES-binding, as a 
guide
for spec authors that are not following the main WebIDL spec evolutions (as not 
everybody 
has your knowledge of what is or is not usable in WebIDL).

The -1 spec explicitly states that people wanting to implement WebIDL are 
invited to read
the main WebIDL specification (that, ideally, should be automatically published 
as /TR/WebIDL ) because yes
WebIDL-1 is _not_ the WebIDL specification, just a frozen snapshot of what was 
implemented as the 
time of publication, not more than that, and bound to be replaced by a 
subsequent level later on.

> Perhaps we need a new space just for documents that represent and
> agree to set of royalty free ideas? (i..e, if it's a REC, it does into
> this new space - and gets clearly marked for the appropriate target
> audience, which is not implementers or developers - but patent lawyers
> and government officials)...
> 
> I think we've also had this debate 10001 times too... but we need to
> do something folks, as the division between the forks and the reality
> of how web specs are developed is hurting everyone :(
> 
> Kind regards,
> Marcos
> 

-- 
Baroula que barouleras, au tiéu toujou t'entourneras.

~~Yves

RE: Quick update on WebIDL "Level 1"

[Marcos Caceres]

On July 9, 2016 at 6:24:56 AM, Domenic Denicola (d...@domenic.me) wrote:
> From: Travis Leithead [mailto:travis.leith...@microsoft.com]
>
> > The purpose of the “Level 1” document is to serve as a stable reference for 
> > W3C specs that
> link to WebIDL. It contains a subset of the WebIDL syntax that is considered 
> stable (as
> verified by interoperable tests). Implementations should not use the Level 1 
> document
> as a guide, but instead track changes to the editors draft. We expect to 
> follow-up Level
> 1 with a Level 2 as additional editor’s draft syntax and behavior stabilizes, 
> are implemented
> as part of other specs, and shown to be interoperable.
>
> Why is it acceptable for specs to reference a version of Web IDL that nobody 
> should implement?

This is a totally valid question, but we've had this debate 1001
times. Perhaps a better question is: how can we get patent protection
(making this subset of WebIDL royalty free for society), but without
harming the ecosystem by confusing implementers and developers by
publishing on the "/TRash" space (as most of us now unfortunately
referring to it).

We need a way to clearly indicate that, for a subset of documents,
RECs on TR represent a royalty free set of ideas (as kindly and
honorably granted by the W3C Membership) - and should only be referred
to by patent lawyers and government officials. That it's for those
groups should be stated and promoted proudly, not disparagingly. And,
that implementers should be looking at the living document instead.
The value of TR need not be diminished - in fact: it should be
correctly used to published the documents that enshrine the royalty
free status of particular specifications.

Perhaps we need a new space just for documents that represent and
agree to set of royalty free ideas? (i..e, if it's a REC, it does into
this new space - and gets clearly marked for the appropriate target
audience, which is not implementers or developers - but patent lawyers
and government officials)...

I think we've also had this debate 10001 times too... but we need to
do something folks, as the division between the forks and the reality
of how web specs are developed is hurting everyone :(

Kind regards,
Marcos

Re: Quick update on WebIDL "Level 1"

[Chaals McCathie Nevile]

On Fri, 08 Jul 2016 22:21:10 +0200, Domenic Denicola  wrote:


From: Travis Leithead [mailto:travis.leith...@microsoft.com]

The purpose of the “Level 1” document is to serve as a stable reference  
for W3C specs that link to WebIDL. It contains a subset of the WebIDL  
syntax that is considered stable (as verified by interoperable tests).  
Implementations should not use the Level 1 document as a guide, but  
instead track changes to the editors draft. We expect to follow-up  
Level 1 with a Level 2 as additional editor’s draft syntax and behavior  
stabilizes, are implemented as part of other specs, and shown to be  
interoperable.


Why is it acceptable for specs to reference a version of Web IDL that  
nobody should implement?


That's not what Travis describes. To restate his message above, the Level  
1 spec is "what people already implement". The Level 2 editor's draft is  
"what you should look at if you want to make a new implementation with all  
the new stuff - but be aware that some if it is up for debate and might  
change".


cheers

--
Charles McCathie Nevile - web standards - CTO Office, Yandex
 cha...@yandex-team.ru - - - Find more at http://yandex.com