W3C home > Mailing lists > Public > whatwg@whatwg.org > June 2004

Transition from Legacy to Native rendering - (was Re: [whatwg] repetition model)

From: Jim Ley <jim.ley@gmail.com>
Date: Mon, 14 Jun 2004 00:58:24 +0100
Message-ID: <851c8d31040613165877a9a7f9@mail.gmail.com>
On Sun, 13 Jun 2004 23:36:46 +0000, Peter-Paul Koch
<gassinaumasis at hotmail.com> wrote:
> >Seen as (by how I read the process) the spec cannot reach a mature
> >stable level until there are native implementations, why should we
> >bother with this restriction?  
>
> Because I want to start using it right away instead of in five years (at
> which date it won't be supported by you-know-which-browser, which will
> probably still have a large market share, and at which date there will be
> some incompatibilities between browsers that do support the specification).

Interesting, as I think you agreed earlier there's little in the spec
we can't already do interopably with scripting on many platforms, with
the same graceful fallback, or indeed better fallback.  So what's the
motivation of using a spec and implementation which is almost
certainly going to be incompatible with future user agents? You're
well known for disliking libraries, not as much as I dislike them, but
this seems to be a challenge to directly implement a library system
for web-forms?

I've also yet to see anything that shows the transition from script
implemented to natively implemented. How will my legacy content work
in these future native renderers - my script will still be there, how
will the hypothetical native renderers know not to execute it?  Will
they have to have a whole load of heuristics to guess if I'm using one
of the script libraries, or will I have to somehow guess that an impl
may be providing support and then not execute my script bits.

Equally as you hint future UAs will likely have incompatibilities -
how do we deal with these incompatibilities in the client? if an
important user agent fails to implement the pattern attribute
correctly, how do we resolve this?  (This is a general problem in all
declaritive "script" such as the pattern element, incompatible UA's
can render all use impossible.)

> Maybe we can use the repeat element after all.

Oh, I never had a doubt about that.  Shame we can't easily have shadow
trees though, might be worthwhile considering as they are possible
though.

Jim.
Received on Sunday, 13 June 2004 16:58:24 UTC

This archive was generated by hypermail 2.3.1 : Monday, 13 April 2015 23:08:17 UTC