RE: [Issue 512] Seeking lower bound on text size requirement

>    (or width) than others. As the Latin script is rather at the lower
>    end (for upper-case only, a height of 5 pixels is enough), 

This is perhaps 'English' rather than 'Latin'.  eg. Eastern European
languages regularly use accents above and below capital letters, and would
therefore require additional height.  

> >  - What internationalization issues enter into this
> >    discussion?

In case it is of use, here are requirements we came up with a couple of
years ago with a customer re. minimum readable font sizes in pixels for a
number of Asian languages.  The minimum here is to allow reasonable
representation of the complicated glyphs involved.

Thai	7-8 pixels in width, but 22 pixels minimum height (because vowel
signs below and above base characters, and tone marks above that.

Korean	16 pixels square (includes a 1 pixel gutter vertically and
horizontally)

Japanese	14 pixels square (includes a 1 pixel gutter vertically and
horizontally) (some strokes have to be sacrificed but context aids
recognition)

Traditional Chinese (Taiwan)	24 pixels square (includes a 1 pixel gutter
vertically and horizontally) (this is obviously a lot, and is not needed for
all Chinese characters, but there are a large number of very complex ones
that do require 24 pixels for all strokes to fit. eg. ?[轟] ?[9F98]
?[9EA4] ?[9A6B] ?[7065] ?[91C1] ?[99A8] 


Note also that inter-line spacing may also be a factor - particularly in Far
Eastern scripts.  Two lines of 16x16 pixel ideographic characters separated
by a single pixel gutter can be very difficult to read - plus the inter-line
spacing generally helps indicate writing direction.  Note that additional
space may be required in the presence of underlined text.


RI

X______________________
Richard Ishida
Globalisation Consultant,
International Document & User Interface Design
Xerox
tel: +44 1707 353395 (Voicemail always available)
http://www.xerox-emea.com/globaldesign/



> -----Original Message-----
> From: Martin Duerst [mailto:duerst@w3.org]
> Sent: 21 May 2001 07:50
> To: Ian Jacobs; clilley@w3.org
> Cc: w3c-wai-ua@w3.org; w3c-i18n-ig@w3.org
> Subject: Re: [Issue 512] Seeking lower bound on text size requirement
> 
> 
> Hello Ian,
> 
> Just a few thoughts:
> 
> - There is two issues of resolution, namely a) the screen resolution
>    (in pixels) and b) the visual resolution (in terms of 
> optical frequency
>    per degree angle measured from they eye.
> 
>    For a), certain scripts may need more pixels for character height
>    (or width) than others. As the Latin script is rather at the lower
>    end (for upper-case only, a height of 5 pixels is enough), 
> you won't
>    get i18n problems, I guess (e.g. if the details for your checkpoint
>    say: "...control...size...of text down to at least 5 pixels...",
>    then that won't make it impossible for other scripts to 
> get to their
>    smallest feasible size). In other words, a tight (i.e. as high as
>    possible) lower bound requirement can differ for different scripts,
>    but you can just take the minimum, and the minimum for Latin is a
>    good overall minimum.
> 
>    b) is more difficult, because the UA software doesn't 
> really control it.
> 
> - Being able to increase the reference size of rendered text 
> can also be
>    an I18N issue, as for some scripts, you really want to 
> make the reference
>    size larger.
> 
> - The hight should be the overall hight, not the x-height. The aspect
>    ratio doesn't have anything to do with it as far as I understand.
> 
> - Please note that there are programs (mostly layout software such
>    as pagemaker,...) that show a line of text below a certain size as
>    a grey strip. This is called 'greek text' (as always, typographers
>    use strange terms :-). This can be quite helpful to get an overview
>    of a large page, less disturbing that actual text that is too small
>    to be read, but I'm not sure it's an accessibility issue.
> 
> 
> Regards,  Martin.
> 
> 
> At 11:21 01/05/18 -0400, Ian Jacobs wrote:
> >Chris, Martin,
> >
> >The UAWG would like your input on a question of visual text
> >rendering and internationalization. Checkpoint 4.1 of the 9 April
> >2001 draft [1] starts:
> >
> >   "4.1 Allow global configuration and control over the reference
> >        size of rendered text ..."
> >
> >This is a Priority 1 checkpoint. One reviewer pointed out that it
> >is not really a P1 requirement to allow the user to choose very
> >small text sizes. Indeed, the intention of this checkpoint is
> >primarily to allow users with low vision to increase text
> >size. [I would note here that small text is useful to some users
> >(e.g., so that users with screen readers can scroll less), but
> >that's not a P1 requirement.]
> >
> >At our teleconference yesterday we asked ourselves whether we
> >could come up with some lower bound on the requirement. Thus,
> >user agents would not be required to provide access to very
> >small text size as part of meeting this checkpoint (or,
> >for example, they might allow configuration, but not
> >actually be required to render very small text).
> >
> >Our questions are thus:
> >
> >  - How might we express a lower bound in text size?
> >    What units would we use? What parameter to measure
> >    size (x-height? aspect ratio?)?
> >
> >  - What internationalization issues enter into this
> >    discussion? Does a lower bound requirement
> >    make sense across different scripts?
> >
> >Thank you for your help on this topic,
> >
> >  - Ian
> >
> >Note to the Working Group: For checkpoints 4.1 and 4.2, we should
> >change "rendered text" to "visually rendered text" to be more
> >precise.
> >
> >[1] http://www.w3.org/TR/2001/WD-UAAG10-20010409/
> >[2] http://server.rehab.uiuc.edu/ua-issues/issues-linear-lc3.html#512
> >
> >--
> >Ian Jacobs (ij@w3.org)   http://www.w3.org/People/Jacobs
> >Tel:                     +1 831 457-2842
> >Cell:                    +1 917 450-8783
> 

Received on Tuesday, 22 May 2001 13:41:00 UTC