assistive-technology-abbr-results: Difference between revisions
AndyMabbett (talk | contribs) (→Valid HTML4: title-trigger) |
JamesCraig (talk | contribs) (→Markup Possibilities: added note on title-trigger addition) |
||
Line 39: | Line 39: | ||
==== Valid HTML4 ==== | ==== Valid HTML4 ==== | ||
* abbr-design-pattern (as currently recommended) | * abbr-design-pattern (as currently recommended) | ||
* SPAN element with title attribute | * SPAN element with title attribute (with or without [[title-trigger]] as it behaves no differently in any known screen reader). | ||
* [[dfn-design-pattern|DFN element with title attribute.]] | * [[dfn-design-pattern|DFN element with title attribute.]] | ||
* Non-linked anchor element with title attribute. (<code><nowiki><a class="dtstart" title="20070713">13th</a></nowiki></code>) | * Non-linked anchor element with title attribute. (<code><nowiki><a class="dtstart" title="20070713">13th</a></nowiki></code>) |
Revision as of 06:08, 6 January 2008
abbr-design-pattern alternatives
This page is for document the precise results of using assistive-technology with real world examples in the wild of the abbr design pattern, as nearly always used by hCalendar (e.g. dtstart, dtend), often by hReview (e.g. dtreviewed, rating) and sometimes by hCard (e.g. bday, type subproperty) and by Geo.
Each assistive-technology listed here should link to a precise description of the tool/service on the assistive-technology page, including version number, when published, etc. Each test result listed for each assistive technology should link to the example that was tested.
Test Cases
The WaSP Accessibility Task Force has developed a set of (extended test cases]) that cover the following data formats and markup formats.
Data Formats
- ISO 8601 date with punctuation (YYYY-MM-DD)
- ISO 8601 datetime with punctuation (YYYY-MM-DDTHH:MM:SS)
- ISO 8601 datetime with punctuation and no seconds (YYYY-MM-DDTHH:MM)
- ISO 8601 UTC datetime with punctuation (YYYY-MM-DDTHH:MM:SSZ)
- ISO 8601 pre-UTC datetime with punctuation (YYYY-MM-DDTHH:MM:SS+HH:MM)
- ISO 8601 pre-UTC datetime with punctuation and no timezone minutes (YYYY-MM-DDTHH:MM:SS+HH)
- ISO 8601 post-UTC datetime with punctuation (YYYY-MM-DDTHH:MM:SS-HH:MM)
- ISO 8601 post-UTC datetime with punctuation and no timezone minutes (YYYY-MM-DDTHH:MM:SS-HH)
- ISO 8601 datetime with punctuation but w/o timezone punctuation. The hypothesis is that something like "eleven o'clock plus oh seven hundred" may be more understandable than "eleven o'clock plus seven o'clock" (similarly for the hyphen, read as "dash").
- pre-UTC (YYYY-MM-DDTHH:MM:SS+HHMM)
- post-UTC (YYYY-MM-DDTHH:MM:SS-HHMM)
- GEO Latitude/Longitude coordinates.
- ISO 8601 durations
- ISO 8601 recurrences
Rejected Data formats
Dates without punctuation as they have proved unusable. Proof is premise for [hAccessibility] article on the Web Standards Project Buzz blog. Window Eyes and JAWS read un-punctuated dates as single number. [JAWS 8 on IE7 citation (MP3 Audio)] of "twenty million seventy-thousand three-hundred twelve" is reason enough for rejection.
- ISO 8601 date without dashes (YYYYMMDD)
- ISO 8601 datetime without dashes and colons (YYYYMMDDTHHMMSS)
- ISO 8601 datetime without dashes and colons and no seconds (YYYYMMDDTHHMM)
- ISO 8601 UTC datetime without dashes and colons (YYYYMMDDTHHMMSSZ)
- ISO 8601 pre-UTC datetime without dashes and colons (YYYYMMDDTHHMMSSZ+HHMM)
- ISO 8601 pre-UTC datetime without dashes and colons and no timezone minutes (YYYYMMDDTHHMMSSZ+HH)
- ISO 8601 post-UTC datetime without dashes and colons (YYYYMMDDTHHMMSSZ-HHMM)
- ISO 8601 post-UTC datetime without dashes and colons and no timezone minutes (YYYYMMDDTHHMMSSZ-HH)
Markup Possibilities
Valid HTML4
- abbr-design-pattern (as currently recommended)
- SPAN element with title attribute (with or without title-trigger as it behaves no differently in any known screen reader).
- DFN element with title attribute.
- Non-linked anchor element with title attribute. (
<a class="dtstart" title="20070713">13th</a>
) - Nested SPAN element.
- Nested ABBR element.
- Nested DFN element (See: mailing list thread, verify at ALA article).
- Empty INS element with DATETIME attr (
<span class="dtstart">13th<ins datetime="20070713"></ins></span>
) - Rarely used event handler. e.g. (
<span class="dtstart" onchange="void(20070713)">13th</span>
) - Hidden INPUT element (undesirable as it requires FORM)
- Empty ABBR element (
<span class="dtstart">13th<abbr title="20070713"></abbr></span>
)
Invalid HTML4
Markup possibilities that are invalid in HTML4/XHTML1 are noted here to be clear that they were considered and rejected. Since microformats build on posh, any technique that is invalid semantic (X)HTML is unacceptable for a microformats solution. There is no need to bother testing these.
- SPAN with RDF dc:date attribute.
- SPAN with RDF geo:position attribute.
- Anchor element with name attr (
<a class="dtstart" name="iso8601:20070713">13th</a>
) *[name] does not need to be unique, but a[name] does. HTML 4.01 Strict, 12.2.1
Markup Rejections
- OBJECT with param value. (requires significant extra markup and CSS in order to *behave* correctly)
- OBJECT with in-page data attribute. (See include-pattern-feedback#Objects_and_Browser_Behavior)
Screenreader Language Pronunciations
Localization is not the job of Microformats, so multiple human languages (English, Spanish, etc.) is irrelevant.
Verbosity settings
- Punctuation, speak none.
- Punctuation, speak some.
- Punctuation, speak all.
- Title attributes, speak none.
- Title attributes, speak some (abbr and acronym).
- Title attributes, speak all.
Screenreaders
- Freedom Scientific JAWS (multiple versions TBD)
- GW Micro Window Eyes (multiple versions TBD)
- Apple Voiceover (multiple versions TBD)
- Dolphin HAL (multiple versions TBD)
Results
Please provide hyperlinks to examples in the wild for each case that is to be tested. Test cases with example in the wild URLs merit more testing than those without. The purpose of these test cases is to find the best solution, whether or not it is currently being used.
Results to be determined.