Re: [css-style-attr] SVG WG comments on CSS Styling Attributes Level 1

Also sprach Chris Lilley:

 > >> Looking back in time by reading the source code of Unix
 > >> edition 7 which dates back _a long time_ (the '70s) I see that the
 > >> 'C' function 'scanf' can handle parsing scientific notation.
 > 
 > Indeed, lex allows scientific notation (as the original comment
 > from SVG notes). CSS however uses a modified lex grammar which does
 > not allow scientific notation.

A better term for the notation in question may be "e notation".

  http://en.wikipedia.org/wiki/Scientific_notation#E_notation

This notation is an artefact from limited i/o capabilities of
calculators and computers. It looks quite differnet from classic
scientific notation, though.

 > AvK> (The group is somewhat split on whether we should add scientific notation,
 > AvK> but most implementors are in favor (some already support it), so it will
 > AvK> very likely happen in due course.)

 > As Hakon is sometimes fond of saying, 'the implementations are
 > going to do it anyway' although I don't agree with his conclusion
 > in that case that 'maybe it doesn't need to be in the spec but
 > people will do it'. This also seems to be Anne's point.

The context of such comments is legacy markup: browsers will continue
to read and decipher legacy HTML even if there is no W3C specification
for it. (HTML5 will hopefully change that.)

Also, UAs will continue to experiment and will add functionality
before it has been standardized. As long as this is done with vendor
prefixes, this is mostly harmless and, yes, vendors will and should do
it even if it's not in a spec.

I hope no UA starts accepting the e-notation in CSS unless it -- over
my body -- ends up in a W3C Recommentation. For example, I don't want
to see this in style sheet:

  h1 { font-size: 2.6e-4em }

It's a feature we don't need, one that will cause harm in terms of
human legibility.

If we need very small numbers in some properties, I'm sure there are
better solutions.

 > Tab Atkins argued in favour, and Daniel Glazman pointed out that
 > everyone who has a need of scientific notation understands it - its
 > on calculators and computers, schoolchildren use it - but Bert Bos
 > and Håkon Lie blocked any change, Bert appealing to 'no-one wants
 > it' and the inviolacy of the core grammar and Håkon appealing to
 > 'no one understands it', suggesting units with multipliers instead
 > (font-size: 12 angstroms, as if that would help or was what we
 > wanted) and, while opposing adding it, also insisting that if
 > added, it use a different syntax with Unicode values for the
 > exponents (he hates the standard scientific notation used on
 > computers for the last 40 years, it seems). Interoperability of the
 > twenty or so SVG implementations over the years (which allow
 > scientific notation in attribute values but are barred by CSS rules
 > from allowing the same thing in stylesheets) was simply ignored or
 > treated as irrrelevant. Because HTML doesn't use it, therefore
 > no-one wants it.
 >
 > The end result of the lengthy, unfriendly and divisive discussion
 > was not only that the proposal failed to get scientific notation
 > allowed (as I predicted, in SVG WG discussions, it would) but that
 > also a number of people took the opportunity to deride SVG in
 > general, to posit strawman outlandish examples that they could
 > demolish, to declaim "I don't understand!" and so forth. A big
 > change from the generally positive, lets-make-this-work attitude
 > that has characterized CSS-SVG discourse, for example on the FX
 > task force, in recent years.

This is a peculiar summary, one I must take exception to. I didn't
find the discussion to be unfriendly. Ideas were tossed around, like
they should be in a WG setting. People argued for and against the
e-notation, like one should in a meeting. Angstroms were mentiond in
cheerful mood on the way to lunch, and not by me.

The most unfriendly part of this debate seems to be your summary. And
that's saddens me; I thought we had a very good meeting.

-h&kon
              Håkon Wium Lie                          CTO °þe®ª
howcome@opera.com                  http://people.opera.com/howcome

Received on Friday, 27 August 2010 09:47:11 UTC