Re: Improvements to Use Case System (Was: Request to re-open issue 131)

Charles Pritchard sent the following reply directly to me, but has said
I'm welcome to follow it up on the list:

> On Dec 19, 2011, at 2:04 AM, Smylers <Smylers@stripey.com> wrote:
> 
> > > I know I've had a difficult time with the "use case" system.
> > 
> > In what way? It seems to make sense to me that features are only
> > designed once it's clear what problems they are solving.
> 
> In that the problems we are solving are from the WCAG checklists and
> system accessibility inspectors. When we state that an item on WCAG is
> not met, or a field in the system a11y API is left empty, those
> specific technical issues are not accepted as use cases.

Well an item on a list created by somebody else is a level of
indirection away from an end-user use case (in that ticking off items in
a list isn't _inherently_ useful; it's only useful because the items on
the list are themselves useful).

But if those checklists were created base on actual use cases then
clearly it could save this working group time to simply adopt them
wholesale rather than re-evaluate each one from first principles.

The chairs have been encouraging working group members to raise bugs on
the decision-making process itself. If you'd like items on WCAG
checklists to be automatically considered as sufficient use cases (by
proxy) then how about raising a bug on the process suggesting exactly
that? That would seem a better approach than tying this to Issue 131).

> When we bring up real-world products, ones on the market or in
> development, we are told that authors should have programmed them in a
> different language (SVG).

That would seem to be an issue of granularity. Currently use cases are
considered for HTML as a whole (or indeed the open web platform as a
whole), so they are met if there is some way of achieving them in HTML.
After all, if something can be done in HTML then anybody can choose to
do it that way.

You've elsewhere referred to 'canvas developers', a group of people I
for one hadn't previously realized existed. If it's important for use
cases to be met with specific sub-systems of HTML rather than HTML as a
whole then again you could raise a bug suggesting this and explaining
why it is important and which sub-systems we should support in this way.

> That is why the use case system is not working.

Thank you for the explanations. Hopefully any problems can be fixed.

Cheers

Smylers

Received on Wednesday, 21 December 2011 17:18:34 UTC