Re: [tlug] ACID Tests (was Re: Browser share in Japan?)

["Stephen J. Turnbull" <stephen@example.com>]

Raymond Wan writes:

 > I think there are sometimes tests where 100% is difficult to 
 > achieve or losing a few percent doesn't necessarily mean 
 > we're failures.  Driver's road tests, for example.  :-)

Yeah, sure, but my point is that I don't think that kind of analogy is
appropriate here.  I failed my first road test here.  I assure you
that although I made mistakes the second time around, I didn't make
the same ones.  But Firefox will lose points on the ACID test the same
way every time.  That's why it matters whether the standard is a good
one or not.

For example, the first comment said "I made an SVG font in Inkscape,
and it didn't work!  ... By the way, it didn't work in <some other
browser>, too."  (Well, d'oh.)  That suggests that the ACID test sucks
for sure, since it doesn't test features that people actually use, and
maybe that the standard sucks too, since nobody (well, 0 of 2 browsers
tested ;-) implements the standard.

On the other hand, several people mentioned the "semantic graphics"
use case.  So maybe the standard sucks by being too featureful, but
there is a subset of SVG Font that browsers "should" implement to make
users happy.  That would *really* suck, because there would be no way
to accurately describe or test "good" implementations based on the
standard.

 > As I understand it, SVG is often associated with web 
 > browsers but they can also be used for other purposes.  So 
 > it also wouldn't make sense for the SVG working group to 
 > define a standard with only the web in mind.

Well, in most modern standards there is a "conformance" section,
describing the features that must be correctly implemented to conform
to the standard.  SVG 1.1 is no exception, and in fact it describes
several conformance profiles for exactly the reasons you have in mind.
One of those is "static viewer" which browsers clearly fit well.
Fonts are a required feature for conformance to the SVG static viewer
profile.  :-(

So you see, the standard authors *do* have some common sense.  It's
just that *in hindsight*, sometimes we realize that their common sense
told them that something should be standard when it's a broken idea.