Re: [widgets] Zip vs GZip Tar

Dear All,

On Tue, May 25, 2010 at 6:16 AM, Doug Schepers <> wrote:
> Hi, Folks-
> Sorry to jump in on this thread so late; I've been busy and traveling.
> As W3C Team Contact for this group, I strongly agree with Ian here regarding
> the tone of some of the responses.  Technical comments on this list should
> be treated with the respect they are due.  If you feel something has been
> adequately covered in the archives, point to an example email.  Please keep
> this list civil, technical, and productive.
> On a logistical level, I again agree with Ian.  I'm rather disappointed that
> we can't solve this problem more quickly.  I think Gregg raised worthwhile
> use cases and points for consideration [1], and wonder if this might not be
> dealt with in the Widgets Embedding spec... after all, that is intended for
> the latter case he mentions.  I can think of many worse things than having 2
> alternate compression schemes, if the use cases are different.  (Yes, I
> realize I'm speaking loosely and there might be serious technical problems
> with this approach... I'm just brainstorming here.)
> Aaron Boodman suggested something [2] on the WHATWG list that sounds
> suspiciously like Widgets, and it would be a real shame to miss out on this
> opportunity for increasing the applicability of the Widgets specs in
> multiple scenarios and platforms.
> [1]
> [2]

Apologies, I'm not subscribed to the WHATWG list so I will respond
here. Please feel free to forward my respond there in necessary.

W3C's widget specs are mature (i.e., most at CR or LC) and the working
group believes them to be technically sound and, with a few
extensions, able to meet the use cases of [2] (particularly in light
of Google using the crx format to package applications - which is more
or less identical on a conceptual level to the W3C Widget work).

Opera and many other browser vendors support W3C Widgets [3] - FWIW,
the spec also has received industry support from OMTP (BONDI) and WAC
[5] (hence, implementations can now be found on Limo, Android, and a
range of other devices). The specifications have also been shown to be
a suitable complimentary technology for Google Gadgets [5].

In the past, the working group has conceded that some of the technical
solutions we have adopted are not ideal (e.g., using XML for config
instead of JSON, XML Dig Sig vs JAR signing, zip as a non-streamable
format Vs tar.gz); perhaps a consequence of consensus-driven
specification design, but one that focused on a consistant set of
requirements [8]. Nevertheless, we have worked extremely hard for 4
years to simplify these technologies and provide robust error handling
throughout our specs. As a result, we have modular specifications that
are highly testable [6] and very extensible (see processing model
[7]). Implementers (such as Opera) have not reported any technical
issues with implementing the W3C Widgets specifications and many have
praised the specifications for their implementability, clarity, and

On the issue of closer system integration and policies for access
control, this would certainly be in the scope of the DAP work.

Opera encourages Chromium/Google to help the Working Group address any
outstanding use cases and get industry-wide harmonization for
packaging Web applications. As Doug has already indicated, there is
scope for this in the upcoming charter under the banner of "widget

I sincerely hope that Google/Chromium take up this opportunity to work with us.

Kind regards,

Marcos Caceres
Opera Software ASA,

Received on Tuesday, 25 May 2010 08:20:34 UTC