W3C home > Mailing lists > Public > public-html@w3.org > August 2008

Re: Flickr and alt

From: David Poehlman <david.poehlman@handsontechnologeyes.com>
Date: Tue, 19 Aug 2008 10:29:37 -0400
Message-ID: <3AB56F39A2494F3A886159842BF54C8E@HANDS>
To: "Boris Zbarsky" <bzbarsky@MIT.EDU>
Cc: "Patrick H. Lauke" <splintered@gmail.com>, "Philip TAYLOR \(Ret'd\)" <P.Taylor@rhul.ac.uk>, "Maciej Stachowiak" <mjs@apple.com>, "Anne van Kesteren" <annevk@opera.com>, "James Graham" <jg307@cam.ac.uk>, "Steven Faulkner" <faulkner.steve@gmail.com>, "Ian Hickson" <ian@hixie.ch>, "W3C WAI-XTECH" <wai-xtech@w3.org>, <public-html@w3.org>

easy is a cop out.  compliant is compliant.  the spec should be written to 
achieve an end which is to provide clear information of syntactical 
correctness.  Other documents can then lay out the course for deriving 
results up from that.

If we fork the spec, the fork of least resistance and most likely, less 
accessible in this case will be the one most travelled.

----- Original Message ----- 
From: "Boris Zbarsky" <bzbarsky@MIT.EDU>
To: "David Poehlman" <david.poehlman@handsontechnologeyes.com>
Cc: "Patrick H. Lauke" <splintered@gmail.com>; "Philip TAYLOR (Ret'd)" 
<P.Taylor@rhul.ac.uk>; "Maciej Stachowiak" <mjs@apple.com>; "Anne van 
Kesteren" <annevk@opera.com>; "James Graham" <jg307@cam.ac.uk>; "Steven 
Faulkner" <faulkner.steve@gmail.com>; "Ian Hickson" <ian@hixie.ch>; "W3C 
WAI-XTECH" <wai-xtech@w3.org>; <public-html@w3.org>
Sent: Tuesday, August 19, 2008 10:19 AM
Subject: Re: Flickr and alt


David Poehlman wrote:
> Right, either you are compliant with a spec or you are not.  It is like
> programming.  you can write a lousy app and it may work but... 
> interestingly
> in some programming environments, they won't run at all if not done right.

Not that great an example, since you can write an app "per spec" and it
won't run in some programming environments.  In fact this is a quite
common problem with a number of existing languages: following the
language spec is no guarantee of things working, because of bugs in the
compilers, interpreters, virtual machines, standard libraries, etc.

> Ths is not to say you can't choose to be non compliant, but we cannot
> provide for exceptions in the spec because those who don't want to be
> compliant but want to claim compliance will work hard to fit themselves 
> into
> the exceptions.

This seems to be using a circular definition of compliance....

Ideally the spec would have the following properties:

1)  Being compliant with the spec is the easy thing to do (leads to
     the "right" behavior from the point of view of authors).
2)  Being compliant with the spec leads to the "right" behavior
     from the point of view of those viewing (hearing, smelling,
     whatever) the content.

Sadly, this is pretty difficult because at heart authors are lazy (just
like all of us, and this is not a bad thing!) and the "right" behavior
from the viewer's point of view varies so widely by viewer....

-Boris
Received on Tuesday, 19 August 2008 14:30:32 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 9 May 2012 00:16:22 GMT