WCAG WG telecon

13 Jan 2005

Agenda, IRC log


Alan, Wendy, Mike_Barta, John_Slatin, Matt, Bengt, David, JasonWhite, rellero, Gregg, Ben, Loretta_Guarino_Reid, Chris, Takayuki, Andi, Avi, Becky, Kerstin, Alex_Li
Yvette, Roberto_Castaldo, Sailesh



text equiv summary

<wendy> http://lists.w3.org/Archives/Public/w3c-wai-gl/2005JanMar/att-0067/text-equiv-summary.html

<MattSEA> for avi: http://www.bestkungfu.com/archive/date/2004/05/understanding-semantics/

grouped issues into 5 groups, 1) close easy, 2) some work) 3) need discussion and some elephants

170# bugzila advertising, - recommend we stay silent on advertisers leave for polcy makers, think it is handled in 1.1

gv agregation we may make a comment -wendy had it removed but will put it back in

<scribe> closed 170


requested about read aloud using assistive technology rather than screen readed---will close


<wendy> http://trace.wisc.edu/bugzilla_wcag/show_bug.cgi?id=664

respond to it buy saying to reviewer, it is beyond us to require that

gv id this about alt text

this is about the delete button

for an icon like garbage can, the alt t4ext shut just say the function rather than a garbage can description...Ben says these discussions should go in General technique. john will note that

Andi...we should close it doesn't belong in our guidelines...gv will respond to say we will recommend the function of the button

Chris .... suggest alt= description but title should say where it was going to go....gv...the purpose of alt text is to be able to remove the images, and have the alt perform the function

Jason - agrees we should close it....GV will say close it and say ALT should be function based...unless it is complex...then a longdesc

<wendy> http://trace.wisc.edu/bugzilla_wcag/show_bug.cgi?id=665

GV...closes issue

<wendy> http://trace.wisc.edu/bugzilla_wcag/show_bug.cgi?id=791

665# a screenreader should.....we've changed the example that they complained about..so its ok


benefit..."1.1 does not benefit people who cannot read test...Wendy thinks we shold decline make no changes and close issue...

andi thinks we should not say text equivalents will help people who have trouble reading text...

<wendy> ACTION: wendy propose a new benefit based on this response

wendy will propose new benefit based on tis response

GV says we should refer bugs related to informative over to ediors and save our group time

<wendy> http://trace.wisc.edu/bugzilla_wcag/show_bug.cgi?id=890

Wendy will skip ones about examples (since they are informative)


guideline leaves no room for null alt

wendy thinks no change and close issue

<wendy> http://trace.wisc.edu/bugzilla_wcag/show_bug.cgi?id=951


text equivalents should be easy to use...its a user issue, close issue

gv: text alternative should be in a standard format

wendy: should be provided according to spec

<wendy> http://trace.wisc.edu/bugzilla_wcag/show_bug.cgi?id=1004

clarify sc in guideline 1.1....its old we have fixed ll the proposed issue, overome by events


covered close..re null alt text



1082 andi...need a sc to cover when to use alt verses LONGDESC.. when to use each....150 characters....

wendy thought problem...150 is english, language specific and html specific...close...but must have istances where 1075 get to later

<wendy> http://trace.wisc.edu/bugzilla_wcag/show_bug.cgi?id=1082

gv lets cover in techniques and checklists not in sc

<wendy> http://www.w3.org/WAI/GL/WCAG20/tests/test3.html

techniques are wreslting with this test #3 no resolution yet

1232 done

<wendy> http://trace.wisc.edu/bugzilla_wcag/show_bug.cgi?id=1321

1321 alt text for graphics should be lower priority...response if alt text is moved it will be done less..let's decline...

1322 related to 1321

<wendy> http://trace.wisc.edu/bugzilla_wcag/show_bug.cgi?id=404

#404 proposals for examples...infomative...adopt it...created a table for SC...an example for every SC...#5 don't do an example...not leanding itself...Wendy would like feedback for examples

people wonder what example are associated with what examples....GV suggests a code, andi says lets figure out how to do itlater but we should do it and associate examples

437# informaive

588, 633 informative

587 def of text equivalent...wendy will close 683 and....nothing to discuss yet...

#666 waiting ofr [roosal about metadata

937 more about examples, suggest we go to general techniques with these examples

<wendy> http://trace.wisc.edu/bugzilla_wcag/show_bug.cgi?id=1079

1079# modify "text alternatives...., " propose ...andi thinks irrelevant till we fix 1.1 GV says let's look at it

<wendy> 1079 - can close

GV says close...if non text is decorative an alt text not required and therefore second sentence is not correct...close it because, it says we need alt for decorative images

# clse

1207 will be closed because related GV says we can close

<wendy> 1207 - can close b/c OBE

GV its clear that it is marked so it can be closed cause current wording covers it

1080 close....

1104# html specific

1075 andi will summarize

<wendy> http://trace.wisc.edu/bugzilla_wcag/show_bug.cgi?id=1075

it should be possible to provide a Text alternative that is not associated with the technology, IE, chart and graph, had alt text...but alt is not enough...longdesc should be used or put on the pages or a link to the page

JW: disagree with proposal, about aainst explict content

person reading who has dsability will not know that the external text is an alternative.. so there will be an explicit assoc copy and an external, very confusing,

if it is done on a server the issue will go away...

GV: agrees

an automated tool may take care of it...erhaps the explcit assoc should be in metadata

Allan: can be in the same page. long desc pointing to a paragraph below the graph. not a separate page

John agrees with Jason

HPR can find a long desc on same page

Mike agrees with andi, he understands it includes linking to same page is in spec

wendyl. propose to both camps: have alt that says has a description. current wording..."maybe say ..."at least one...." play with wording to satisfy both groups

GV: don't have bits and peices of the description in different places...

<Kesh> Kerstin: I have to go to another meeting, sorry -- just wanted to put my hand up and agree with Andi and Mike. I agree that we should not create rules for testing facilitation as Gregg is saying, as well.

GV let's not create rules just for testing...and besides you can't test if a alt text is necessary

GV I had a chart, and chart was described below so no need to have a LONGDESC to it cause it will be read twice, we should just make sure that it is easy to find the text alternative

Andi, OK with that, but if there has to be a programatic association, then that's a problem

GV: so every image should have a caption and it could be an alt text?

wendy what about SVG,

Wendy.. no defn of explicitly associated

Jason disagress cause it should be a machine readable so machine can decide what to do it. disagrees that new technologies will not support it. because our specs can be followed by mfg

he disagrees with andi's position

oops andi's postion

prefers, saying that we put explicit assoc in metadata, HTML does satisfy explicit association ALT text satisfys even if more content alternative is not machine associative

GV. can'[t just say there is a chart here

If you have a technology that doesn't associate then does not satify 1.1

If it wasn't necessary to put in the ALT text.

Wendy....if someone invents new format that brings in benefits, but not acessible, but we provide acessible html version it is the text alternative...to have expicit association necessary then this would not be conforming


<wendy> isn't that 4.2? "Any programmatic user interface components of the content conform to at least the default set of conformance requirements of the UAAG 1.0 at Level A plus the sets of requirements (a) through (i) (below) that apply. If the custom user interfaces cannot be made accessible, an alternative solution is provided that meets WCAG 2.0 (including this provision) to the level claimed."

sub andi for wendy...

Jason...not what's happening in 1.1...just make conformance claim on the html version, rather than saying its a 1.1 issue...an alternative version that meets guideline stands on it s own.

andi: so if I have an image on a page, diecide to leave alt null and put description below..is that conforming

GV says no put "Chart...descibed below" in the alt text...because its not a decorative issue

john what would happen if "use title in paragraph" using title element as an alt but

GV we have closure,,, we must have ALT text exicitly associated but if there is more. text it should be easy to find..

<wendy> ACTION: andi propose definition of 'explicitly associated'

<wendy> dmd: it "is must be easy to find" is that testable?

<wendy> gv: immediately adjacent, explicitly associated, or longer description explicitly associated.

<wendy> we'll need defn of "immediately adjacent" especially if the content can reflow...how code it in case it gets repurposed?

<wendy> i.e., to ensure adjacency

GV: if its not immediately agascent then it would not confom like text books images are not immediatly ajascent

spelling adjacent

JW: i think we killed the elephant

Ben will send John the minutes from the bug...ANdi will work on the defintion

of expicitly association...andi doesn't think the elephant is dead....gredd says we'fe not behind it anymore we are in front of it

issue 895

baseline discussion,

questions about scripts, noscripts

related to non-text content , and related to 1.1 that wendy has an action item

gv; uaag if you have a script y0u must do following, but doesn't say must support scripts,

<wendy> related to defn of non-text content (issue 587)

UAAG compliance as baseline would say script need not be supported...

Mat May..but that is the users own fault for not using a browser that supports scripts

<wendy> gv: UAAG says, if you support scripts, you need to do x, y, z. It does not say, you must support scripts.

gv: if uaag is baseline we must assume scripts are not accommodated...we would hve to make a special provision in our baseline to accomodate this

jw; uaag not designed to accept any technology including html???

the developer makes a decsion on what they will support,

one way is to recognize uuag does not require speicif technolgy

content dev. will make those associaitons

gv: that leaves our guidlines ambiguous

wendy: in doublin we should expicitlysay our assumption...

<wendy> http://www.w3.org/TR/UAAG10/conformance.html#conformance-claims

this is still in the air....jason have problem with expliit technology..stuff..

jason guidleines should not be technology specif

wendy...we are not oracles but we must say what aour assumptions are,,esp for HTML techniques...ie do we assume scriting is supprted or not,

Jason: the content author should e allow to assume what they want

GV this is a hydra and we will come back to it

wendy is working on 1.2 and we will come back to it

Summary of Action Items

[NEW] ACTION: andi propose definition of 'explicitly associated'
[NEW] ACTION: wendy propose a new benefit based on this response
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.106 (CVS log)
$Date: 2005/01/13 23:17:30 $