[Accessibility] [Accessibility-ia2] Text attributes issues

Li Yuan Li.Yuan at Sun.COM
Tue Dec 9 19:07:01 PST 2008


Hi Will and Marco,

The situation in ATK is these strings are auto-generated by 
glib-mkenums. So if we want to change these strings, we need to change 
the definition of AtkTextAttribute which is API change I think. Another 
possibility the solve the problem is to use 
atk_text_attribute_get_name/atk_text_attribute_for_name which are 
provided by atktext. Applications will not care what strings are used in 
this case. Are there similar APIs in IA2?

Li

Willie Walker wrote:
> Hi Marco:
>
> Thanks for the quick reply.  It was a while ago, so my memory of the 
> whole thing is fading.  Now that you mention the -moz- attributes, I 
> believe I recall one of the issues was how the GTK+ font attributes 
> were being exposed through ATK.
>
> I'm CC'ing Li Yuan on this to see if he can provide more detail about 
> how we might (or might not) be able to use CSS attributes in the 
> AT-SPI infrastructure.
>
> Will
>
> Marco Zehe wrote:
>> HiWill et al,
>>
>> I wasn't involved in that original discussion for ATK text 
>> attributes, but from our recent conversation in e-mail, from what 
>> Aaron said, I thought it was intended for these text attributes to be 
>> CSS-named. You provided a list of discrepancies, many of which seem 
>> pretty obvious candidates to me. Some of them we have to fix, esp the 
>> -moz- attributes we expose, but the others would seem to be obvious.
>>
>> Am I missing something?
>>
>> Marco
>>
>> On 09.12.2008 08:14, Willie Walker wrote:
>>> Hi Pete:
>>>
>>> Thanks for doing this analysis and writeup.  Very useful.  As we saw 
>>> earlier this summer, normalization of text attributes is important 
>>> across toolkits (e.g., GTK+, Gecko, Java, OOo) as well as across 
>>> platforms.
>>>
>>>  
>>>>> Will, I don't know what to say about the Linux spec other than I 
>>>>> know there is some pressure from Marco and Aaron to change it.       
>>>
>>> It would be great if Aaron and Marco could come up with a list of 
>>> proposed changes.  Are they working on this?
>>>
>>> Thanks again!
>>>
>>> Will
>>>
>>> Pete Brunet wrote:
>>>  
>>>> I should have cc'd the open a11y list too...
>>>>
>>>> ----- Forwarded by Pete Brunet/Austin/IBM on 12/09/2008 09:25 AM -----
>>>> From:     Pete Brunet/Austin/IBM at IBMUS
>>>> To:     Accessibility-ia2 at lists.freestandards.org
>>>> Cc:     Willie Walker <William.Walker at sun.com>
>>>> Date:     12/09/2008 01:29 AM
>>>> Subject:     [Accessibility-ia2] text attributes issues
>>>> Sent by:     accessibility-ia2-bounces at lists.linux-foundation.org
>>>>
>>>>
>>>> ------------------------------------------------------------------------ 
>>>>
>>>>
>>>>
>>>>
>>>>
>>>> Aaron, This is the list of text attributes issues:
>>>>
>>>> Existing entries in Mozilla's bugzilla
>>>>
>>>>     * text-indent and text-align should really be object attributes -
>>>>       _https://bugzilla.mozilla.org/show_bug.cgi?id=460932_
>>>>     * font-size text attribute should be exposed in pt units -
>>>>       _https://bugzilla.mozilla.org/show_bug.cgi?id=467146_
>>>>     * IAccessibleText::caretOffset should return -1 if the system 
>>>> caret
>>>>       is not currently with in that particular object -
>>>>       _https://bugzilla.mozilla.org/show_bug.cgi?id=448744_
>>>>
>>>>
>>>> Other Mozilla issues
>>>>
>>>>     * The Mozilla spec for the language attribute refers to IETF RFC
>>>>       1766 not IETF RFC 3066 (which obsoletes 1766). The Mozilla spec
>>>>       probably needs to be changed.
>>>>     * In the Mozilla spec, font-weight has values of bolder, lighter,
>>>>       and inherit while IA2 doesn't. These probably need to be removed
>>>>       from the Mozilla spec.
>>>>     * auto-generated and writing-mode aren't documented.  You 
>>>> mentioned
>>>>       there were bugs opened for this but I couldn't find the bug 
>>>> numbers.
>>>>
>>>>
>>>> IA2 issues
>>>>
>>>>     * text-position with no offset uses the term "baseline" vs 0. It
>>>>       looks like IA2 should reference the CSS2 vertical-align spec, 
>>>> not
>>>>       the ODF spec.
>>>>           o Do you want to change the name of this attribute to
>>>>             vertical-align?
>>>>           o I propose the value list contain only baseline, sub, and
>>>>             super and not the rest (top, text-top, middle, bottom,
>>>>             text-bottom, <percentage>, <length>, inherit).  I don't
>>>>             think an end user will care about the nuances of 
>>>> exactly how
>>>>             far above or below the baseline the text is.
>>>>           o I propose we drop the existing second parameter in the IA2
>>>>             spec which specifies the size of the font of the
>>>>             sub/subscript.  That can be covered via the font-size
>>>>             attribute.
>>>>           o The default can be baseline.  (It was 0%.)
>>>>
>>>>
>>>> Other comments
>>>>
>>>>     * In the Mozilla spec, many of the attributes (like font-family,
>>>>       font-style, and font-size) refer to CSS 1, 2.1, and 3 while IA2
>>>>       only refers to CSS 2.  I'd prefer to keep the IA2 spec 
>>>> referencing
>>>>       CSS2 because 2.1 and 3 are not at Recommendation state yet 
>>>> (as far
>>>>       as I can tell).
>>>>
>>>>
>>>> I'd appreciate it if Marco could take a look at the two specs and 
>>>> see if he can spot any issues.
>>>>
>>>> Will, I don't know what to say about the Linux spec other than I 
>>>> know there is some pressure from Marco and Aaron to change it.  Are 
>>>> there any issues with using the IA2 spec?  Note that the IA2 spec 
>>>> is mostly based on CSS2, plus one case each of an attribute based 
>>>> on WAI-ARIA and XSL 1.1.  However, there are several attributes 
>>>> based on ODF which are needed to provide access to ODF docs:
>>>>
>>>> text-line-through-mode, text-line-through-style, 
>>>> text-line-through-text, text-line-through-type, 
>>>> text-line-though-width, text-outline, text-shadow, 
>>>> text-underline-mode, text-underline-style, text-underline-type, 
>>>> text-underline-width
>>>>
>>>> I didn't mention text-position because, as mentioned above, I'm 
>>>> proposing that we refer to the CSS vertical-align spec instead of 
>>>> the ODF spec.  (I've also asked the ODF office list why they are 
>>>> using a non-standard attribute when text-position is so close.
>>>> *
>>>> Pete Brunet*
>>>>                                                                        
>>>> IBM Accessibility Architecture and Development
>>>> 11501 Burnet Road, MS 9022E004, Austin, TX 78758
>>>> Voice: (512) 286-5485, Cell: (512) 689-4155
>>>> Ionosphere: WS4G_______________________________________________
>>>> Accessibility-ia2 mailing list
>>>> Accessibility-ia2 at lists.linux-foundation.org
>>>> https://lists.linux-foundation.org/mailman/listinfo/accessibility-ia2
>>>>
>>>>
>>>> ------------------------------------------------------------------------ 
>>>>
>>>>
>>>> _______________________________________________
>>>> Accessibility mailing list
>>>> Accessibility at lists.linux-foundation.org
>>>> https://lists.linux-foundation.org/mailman/listinfo/accessibility
>>>>     
>>>
>>> _______________________________________________
>>> Accessibility-ia2 mailing list
>>> Accessibility-ia2 at lists.linux-foundation.org
>>> https://lists.linux-foundation.org/mailman/listinfo/accessibility-ia2
>>>   
>>
>



More information about the Accessibility mailing list