Re: [agenda] Tuesday Sep 9 teleconf

http://flickr.com/photos/hendry/2824592876/ - Chrome seems to fail the
cookie test amongst others. Haven't had time to look into this.
http://flickr.com/photos/hendry/2840053692/ - ie8beta2 is also misbehaving

On Mon, Sep 8, 2008 at 4:03 PM, Dominique Hazael-Massieux <dom@w3.org> wrote:
>  * Testing Widgets?
> ACTION: Kai to contact editor of Widgets packaging spec to see if testing is appropriate [recorded in http://www.w3.org/2008/09/02-mwts-minutes.html#action03]

I'm in good contact with Marcos (the widget spec editor) and according
to him there are some parts of the specification that are stable and
testable. Though a more realistic date of November to begin a suite
was suggested.

I'm also in touch with Andrei Popescu (Geolocation spec editor). He
also suggested to wait a little before undertaking any testing work.
I'm hoping to meet with him and we should have a better idea next
week.

>  * Testing plugin support?
> ACTION: Kai to look into testing for plugin support [recorded in http://www.w3.org/2008/09/02-mwts-minutes.html#action05]

I fear external mimetypes
http://static.webvm.net/supportedMimeTypes.html and (NPAPI) plugins
has been ignored by the W3C. I can understand why, but at the same
time I recognise the importance of being able to extend the
capabilities of a Web browser in order to innovate (e.g. Gears, Flash
video). Disclosure: I work at the Aplix Corporation which does just
that with WebVM http://wiki.webvm.net/

The grounds for innovation are less fertile on mobile devices, which
are generally locked down from extension. So aiming a plugin test at
help could prevent stagnation of the Web like we've seen in the past
on Desktops.

At work, we have developed a (freely licensed) minimal NPAPI plugin
http://git.webvm.net/?p=npsimple

NPAPI is as close as we have to a plugin standard. It is supported by
Opera, Webkit & Gecko. IE/Trident uses ActiveX and at Aplix we are
putting together a minimalistic test for that too.

The NPAPI standard is worked upon on a private mailing list:
https://mail.mozilla.org/listinfo/plugin-futures

So the "plugin support test" could be just something like
http://foo.webvm.net/simple.html However how does the platform know
where to find and install the plugin?

A plugin finder service has been employed by Mozilla in the past, but
the service seems to be poorly maintained:
https://bugzilla.mozilla.org/show_bug.cgi?id=453632

So besides actual plugin support, the packaging of plugins needs some
attention. So to conclude I propose using 'npsimple' for testing basic
plugin support in browsers.


Kind regards,

Received on Tuesday, 9 September 2008 10:53:29 UTC