RE: Axis attribute

> Joachim Andersson

> What I was saying is that if a user agent doesn't support an element
> or an attribute, that is in no way a reason not to use it. If it is
useful
> in its declaration in a specification and if it doesn't present a
usability 
> issue like longdesc, it should be used as specified. The user agents
will 
> hopefully get to a point where they can handle these functions
properly.

Unfortunately, although I agree with the principle, it's very tough in a
commercial setting to justify additional quality assurance, training and
development for features that are simply not implemented just yet. It's
a hard sell to managers..."the standard says we should be using X, for
accessibility, although at the moment no user agent actually supports
it, but in future they hopefully will".

And it then becomes a chicken and egg problem: user agents will most
likely only support it once there's a critical mass of pages
speculatively using these features, as otherwise it won't add value to
their product and by extension to their customers; web developers in a
commercial setting will most likely only implement them once user agents
show actual support.


The idealist in me agrees that the cycle needs to be broken, and that we
should use features per spec even when they're not used in any real user
agent, hoping that in future they will...but the pragmatist just knows
that, unless you can sell your boss/management to the idea of "we're
doing it, nobody benefits from it now, but in future they will", it's
unlikely...

P
________________________________
Patrick H. Lauke
Web Editor
Enterprise & Development
University of Salford
Room 113, Faraday House
Salford, Greater Manchester
M5 4WT
UK

T +44 (0) 161 295 4779
webmaster@salford.ac.uk

www.salford.ac.uk

A GREATER MANCHESTER UNIVERSITY 

Received on Friday, 28 November 2008 11:33:26 UTC