Re: gtkhtml2 vs. gtkhtml1

[Bill Haneman <bill haneman sun com>]

On Fri, 2002-09-20 at 10:57, Michael Meeks wrote:
> Hi Bill,
> 
> On Thu, 2002-09-19 at 14:15, Bill Haneman wrote:
> > > 	Since we effectively provide a sort of 'DOM' via the Atk system, it
> > > seems particularly pointless to waste a huge amount of time creating
> > > another parallel-but-different way to do so, since it cannot (shouldn't)
> > > add anything useful that Atk+ doesn't do.
> > 
> > But a "sort of DOM" is not w3c DOM, which is what the w3c UAG requires.
> 
> 	Yes, my question is why bother with the w3c spec, when it mandates that
> we waste our time[1], providing duplicate functionality with a different
> API. It seems that only Australia has 'reference to w3c' standards. [cf.
> http://developer.gnome.org/projects/gap/laws.html]

Well:

* the gap "laws" pages are not 100% up-to-date

* the w3c spec is just now reaching "Recommendation" status (I argued
vehemently against mandatory DOM in discussions with the Working Group,
and did manage to get some recognition of the issues, but the
requirement remains).  While it was in 'draft' status it's very unlikely
that anyone would adopt the recommendations.

I think we can expect these UAGs to become much more widely cited,
adopted, and legislated in the future.  I don't think they are a great
match for the work we are doing, but if we wish to comply with or
respect standards body guidelines we sometimes have to adopt solutions
that diverge from our preferences.

I hope that we can comply with the UAGs in a reasonably timeframe, and
more importantly that the UAGs in practice actually help ensure
meaningful improvements in end-user accessibility; time will tell on
both counts.

best regards,

- Bill

> 	So - does that mean we have to spend months duplicating effort,
> producing an interface that it's unlikely anyone will ever use, and
> doesn't give any more power ?
> 
> 	Hmm,
> 
> 		Michael.
> 
> [1] - in absolutely huge truck loads.
> -- 
>  mmeeks gnu org  <><, Pseudo Engineer, itinerant idiot
>