Re: [widgets] Editorial Comments on 18-Aug-2009 LCWD of A&E spec

On Fri, Sep 11, 2009 at 9:30 PM, Arthur Barstow <> wrote:
> On Sep 11, 2009, at 2:47 PM, ext Marcos Caceres wrote:
>> Can I be a PITA and ask you to set up the wiki page and just reuse the
>> old text. Then I will link to it.
> I took a first pass at this:

Looks great. Linked both P&C TSE and A&E.

>>> 9. Section 2: Getting and Setting: seems like there should a [Reference]
> My thought here was that if Getting and Setting are defined e.g. in a DOM
> spec, a reference to that spec should be added; if not, ignore my comment.

No, I nicked those from WebStorage... I don't think there is much
point in referencing web storage, as no further information can be
found there.

>>> 10. Section 2: Initialization: I don't understand this sentence, which
>>> when
>>> shortened is effectively "The first run through X, prior to runtime.".
>> How about:
>> "When a user agent first runs a widget package through the Steps for
>> Processing a Widget Package, as specified in the [Widgets-Packaging]
>> specification, prior to runtime."
>> That boils down to:
>> When a user agent first runs a widget package through X, prior to runtime.
> I still can't parse/grok it but it's probably just me and I'll bug you about
> it in IRC (someday) :-).
> Perhaps part of the confusion is that "user agent" in the context of this
> defintion is presumably a P&C user agent yet that qualification isn't made
> and the only reference to that definition is in Section 6.1 and in that
> context the UA is the A&E UA.

Yeah, you nailed it (well, section 4.1 Support of Other Specifications
[1] kinda defined it). Reworked the definition of a UA to:

A user agent is a software implementation that supports:
 * The widget interface.
 * The [Widgets-Packaging] specification.
 * The [Widgets-URI] specification.
 * Storage areas.

It is optional for a user agent to support the widgets
[Widgets-DigSig] specification.

Where supports means "a user agent implements a mentioned
specification or conformance clause."

I removed "4.1 Support of Other Specifications", as it was incorrect anyway.

However, the question about if "initialization" is clear enough
remains. We should definitely work through that in IRC.


Marcos Caceres

Received on Sunday, 13 September 2009 17:07:17 UTC