W3C home > Mailing lists > Public > public-webapps@w3.org > July to September 2010

Re: XBL2

From: Ian Hickson <ian@hixie.ch>
Date: Thu, 9 Sep 2010 19:19:14 +0000 (UTC)
To: Doug Schepers <schepers@w3.org>
cc: Arthur Barstow <art.barstow@nokia.com>, Jonas Sicking <jonas@sicking.cc>, "public-webapps@w3.org" <public-webapps@w3.org>, "hyatt@apple.com" <hyatt@apple.com>
Message-ID: <Pine.LNX.4.64.1009091909320.27869@ps20323.dreamhostps.com>
On Thu, 9 Sep 2010, Doug Schepers wrote:
> Arthur Barstow wrote (on 9/8/10 1:55 PM):
> > 
> > On 9/4/10 6:36 AM, ext Doug Schepers wrote:
> > > To that end, could you provide a link to the requirements document, or
> > > if there isn't one, could you start one?
> > FYI: when the Web Application Formats WG transitioned XBL2 from Last
> > Call WD to CR (March 2007), the transition request included the
> > following re requirements:
> > 
> > [[
> > 5. Evidence that the document satisfies group requirements:
> > 
> > The spec satisfies more requirements than those defined in the
> > group's 10 February 2006 XBL Use Cases and Requirements document:
> > 
> > http://lists.w3.org/Archives/Public/public-appformats/2006Feb/att-0000/XBL-UCs-and-Reqs-2006-02-10-DRAFT.html
> > 
> > ]]
> Thanks for the pointer.  But it seems that the requirements have 
> changed, according to Hixie, so I'd like to understand better the 
> motivation for the changes he made. 

I didn't examine the above list in depth, but apologies if I made it sound 
like the requirements had changed; they haven't. What changed is the 
context in which the spec is viewed -- one in which HTML has seen a 
resurgence as the recognised core platform, in which our understanding of 
latency and synchronicity implications in API designs is far improved, and 
in which we (or at least I) have a far more appreciation of the importance 
of incremental design in specifications, to lower the initial cost of 
implementations without closing doors on the features the language can 
support over the long term relative to the full vision of the spec.

> Some record of the implementer feedback he's received would probably 
> suffice.

People pointed out the above (in particular the possibilities that would 
be afforded by a closer integration with with HTML and XHTML, rather than 
having a separate vocabulary, and the benefits of the incremental 
approach). Hyatt and I then discussed how to apply these lessons to the 
XBL2 spec for an initial proposal, which is what I then edited.

Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'
Received on Thursday, 9 September 2010 19:19:43 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 18:13:11 UTC