Re: Raw minutes from 11 July UAWG teleconference

[Sorry for the repeat, David.  This one is even better]

At 05:06 AM 2002-07-12, David Poehlman wrote:
>al, do you have propposed wording?

<asIs
cite="">

    1. Allow global configuration of the scale of visually
       rendered text. Preserve text size differences when the user
       changes the scale.

</asIs>

<option1>

    1. Allow one master configuration setting the general sizing of visually
       rendered text.  Preserve larger/smaller relationships 
       between text fragments of different sizes as their respective 
       sizes change in response to this setting.

</option1>
<option2>

    1. Allow configuration of the general sizing of visually
       rendered text.  Preserve text-size distinctions in 
       the text as the general sizing increases and decreases.

</option2>

Note that either optional language offered above does not _disallow_ meeting
this by proportional scaling, it just does not *require* proportional scaling.

The second option is more compact, but is more likely to be misunderstood.
What we want to be true is that if text A is smaller than text B at one value
of this configuration setting, that text A will still be smaller than text B
at another value of this configuration setting.  I think I favor the shorter
one, but I could live with any of the above.

The as-is version does have the difficulty that some will read it literally
as the numeric difference in the size number should be preserved, and it is
not our intent to require that, specifically, either.  It is also confusing
as to the sense of 'global' configuration.  It is meant to imply global across
the whole corpus of pages processed through the User Agent at that setting, 
as well as an adjustment that applies globally to all the sizes applied to
distinct classes of text content.

[Aside: In my model of the extreme style adjuster, this is not always true, just
a guideline to follow so far as possible.  Eventually we start shifting the
means of conveying distinctions from text size to other property dimensions 
such as bolding and prefix/suffix text inserts such as we use in Braille.  The 
ultimate rule is that a) distinctions and b) relative prominence shall be 
preserved so far as possible.  And that to the extent possible, distinctions 
that the author conveyed in text size will be expressed in text size etc.  But 
the latter will be violated before we give up on the former.  However, we are 
not asking them to implement the extreme capability.  That's for the 
experimental agenda.]

Al

>----- Original Message -----
>From: "Al Gilman" <asgilman@iamdigex.net>
>To: "Ian B. Jacobs" <ij@w3.org>; <w3c-wai-ua@w3.org>
>Sent: Thursday, July 11, 2002 9:44 PM
>Subject: Re: Raw minutes from 11 July UAWG teleconference
>
>
>
>At 03:32 PM 2002-07-11, Ian B. Jacobs wrote:
>>------------------
>>On checkpoint 4.1
>>------------------
>>
>>IJ: What about new wording?
>>
>>   1.  Allow global configuration of the scale of visually rendered
>>   text. Preserve text size differences when the user changes the
>>   scale.
>>
>>IJ: Suggest: "Text rendered at different sizes should scale
>>proportionally."
>>
>>HB: What about rounding errors?
>>
>>Action IJ:
>>
>> 1. Change second sentence in bullet one to
>>    "Text rendered at different sizes should scale proportionally."
>
>AG: I object to this change.  I was quite careful in the earlier
>language
>_not_ to suggest proportional scaling.  It is not a good idea.  The
>order
>of the differences should be preserved; the smallest should wind up
>smallest
>etc.  But the ratios of the sizes should be compressed in the
>enlargement.
>You have to pack the same number of different sizes into a smaller
>dynamic
>range, that is to say the ratio between the largest feasible font size
>to
>the smallest feasible font size.
>
>People with low vision have less available dynamic range with regard to
>font size than people with 20:20 vision.
>
>The nose reader at the limit of who can use the screen at all has a
>short
>distance from them to the screen and the amount of text that they can
>see
>is limited by their field of vision, not the size of the physical
>display.
>
>By the time one gets to the edges of the physical display screen, it is
>beyond where they can focus.
>
>So one is fighting the reduced dynamic range between where the details
>in
>the glyphs are too fine for them to resolve on the small font end and
>where
>the amount of text that fits in their field of vision is too few
>characters to
>support reading.  This is the extreme case just before the screen gets
>useless,
>but this is where to think about.
>
>So one want to compress the range of type sizes, into something like new
>sizes
>of 18, 20, 22, 24, 28 from old sizes
>of  6,  8, 11, 14, 20.
>
>You don't want to force the top size to 60 just to
>make the stuff that was in size 6 (before enlargement) be size 18 to be
>legible.
>
>Please reconsider this change.
>
>Al

Received on Friday, 12 July 2002 09:22:56 UTC