14:01:45 RRSAgent has joined #wai-wcag 14:01:45 +Tim_Boland 14:02:07 +[Microsoft] 14:02:13 rrsagent, make logs world 14:02:57 +Ben_Caldwell 14:03:58 Becky has joined #wai-wcag 14:04:11 ben has joined #wai-wcag 14:04:12 wendybrb has joined #wai-wcag 14:04:24 zakim, who is here? 14:04:24 On the phone I see ??P3, Jim_Thatcher, +1.202.312.aaaa, Becky_Gibson, James_Craig, Chris_Ridpath, Takayuki (muted), Tim_Boland, [Microsoft], Ben_Caldwell 14:04:28 zakim, [Microsoft] is temporarily Janae_Andershonis 14:04:28 On IRC I see wendybrb, ben, Becky, RRSAgent, jslatin, JimT, ChrisR, nabe, cookiecrook, Zakim, MichaelC 14:04:30 +Janae_Andershonis; got it 14:04:30 i will be there in ~2 minutes. please check your email. *just* sent gateway drafts. 14:04:34 +Michael_Cooper 14:04:44 zakim, I am Michael_Cooper 14:04:44 ok, MichaelC, I now associate you with Michael_Cooper 14:05:15 zakim, ??P3 is Tom_Croucher 14:05:15 +Tom_Croucher; got it 14:05:18 zakim, +1.202 is Alex_Lee 14:05:18 +Alex_Lee; got it 14:05:34 zakim, I am Becky_Gibson 14:05:34 ok, Becky, I now associate you with Becky_Gibson 14:08:33 jslatin, will you be joining by phone? 14:10:35 +Paul_Bohman 14:10:53 +Wendy 14:11:43 zakim, who's on the phone? 14:11:43 On the phone I see Tom_Croucher, Jim_Thatcher, Alex_Lee, Becky_Gibson, James_Craig, Chris_Ridpath, Takayuki (muted), Tim_Boland, Janae_Andershonis, Ben_Caldwell, Michael_Cooper, 14:11:46 ... Paul_Bohman, Wendy 14:12:23 Meeting: WCAG WG Techniques Task Force 14:12:29 Chair: Michael Cooper 14:12:37 agenda+ gateway 14:12:41 agenda+ testing framework 14:12:44 agenda+ end-to-ends 14:13:00 zakim, take up agendum 1 14:13:00 agendum 1. "gateway" taken up [from wendy] 14:13:34 http://lists.w3.org/Archives/Public/w3c-wai-gl/2004JulSep/0239.html 14:18:24 wac goes through the issues in the message 14:18:45 tb is this normative? 14:18:57 tb how relates to test framework? 14:23:02 wac informative. not sure if needs to be objective or not. difficult to make appropriate text alternative objective. 14:23:10 tc views: prefer the 3rd - divided into chunks. 14:24:00 tc would like to see w/more information 14:24:23 bc each criterion has its own page. borrowed navbar from pubrules. 14:24:30 bc menus are generated dynamically 14:24:39 bc if another guideline in there, all the info would be generated. 14:24:53 tc could we generate an internal draft of the original source to see what it would look like? 14:25:06 bc we made some change to structure, so some editing needed, but could be odne 14:25:14 bg liked the 3rd one as well. 14:25:31 bg the numbering scheme is confusing. could like to see it at the top. 14:26:33 appropriate text alternative is human-testable and sometimes machine-testable. 14:27:10 wac curious about how people feel with moving forward. realize just got 20 minutes ago, but curious about how much time people feel they need and about publishing the documents 14:27:24 mc have to delay everything? 14:27:50 mc techniques point back to guidelines 14:28:12 tc they are a collection, have been developed that way 14:28:17 mc do we want to slide by a week? 14:29:21 wac would like to avoid 14:29:50 mc, tc happy with what we have 14:30:30 js ok with what we have 14:32:21 but may need to add item abt multimedia to list in gateway 14:33:47 wac good point. not sure what it would look like - just a reference to guideline 1.2? 14:34:43 action: wac create placeholder technique, ednote "1.2 not covered in this draft" (similar to level 3) 14:37:28 action: wac include placeholders for 1.2 and at least one guideline for each of the other principles (to flesh out the TOC a bit more) 14:37:43 mc group leaning towards 3rd view 14:37:50 (slices view) 14:39:09 bc anything to do w/numbering as interim? 14:39:57 action: wac ednote about numbering, bc remove numbering for techniques 14:40:54 zakim, close this item 14:40:54 agendum 1 closed 14:40:55 I see 2 items remaining on the agenda; the next one is 14:40:56 2. testing framework [from wendy] 14:41:01 zakim, take up item 2 14:41:01 agendum 2. "testing framework" taken up [from wendy] 14:41:17 +John_Slatin 14:41:25 jenae's message: http://lists.w3.org/Archives/Public/w3c-wai-gl/2004JulSep/0196.html 14:41:47 ja less of a proposal, more of a "here's everything we can have in the test framework" 14:41:52 ja we need to figure out what should be in it 14:42:40 http://lists.w3.org/Archives/Public/w3c-wai-gl/2004JulSep/att-0196/H_1_1_001.htm 14:42:52 ja everything that could be associated with one test case 14:43:10 ja starting w/metadata section type: automated or manual 14:43:21 ja automated - write a script to automate it 14:43:31 .. manual - person needs to decide it if passes or fails 14:44:00 .. test level - just look at on a page or in regard to whole site? 14:44:20 .. if have 100 pp and all titles need to be unique, then test throughout "all" of site 14:44:44 .. anything that should be added? 14:45:13 .. next is test case: prereq - if there is a test that has to be run before this test? 14:45:35 .. pass and fail instructions 14:45:58 .. 3. Examples 14:46:20 .. 4. compatibility (test matrix from march comes into play here) 14:46:38 ..5. resources 14:46:49 ..6. status 14:48:02 .. could then sort by any of these fields. i.e., all test cases for img field, or all that work with jaws or ... etc. whatever can do in xslt. :) 14:48:25 tb requirement that the test files are validated? 14:49:03 ja that could be a test case. however, this isn't so much about format but here are the things we can commit to giving to our audience. 14:49:34 tb user interaction, i.e., prompt author for more information. a self-explanatory test purpose. 14:50:06 tb perhaps platform specific issues that the user will want more clarification. i.e., using a particular browser 14:50:17 ja that could be in the compatibility section. "if it fails on xyz, put that there" 14:51:15 tc human or automated - some tests are not automatable, but can show the user what highlighting. 14:51:42 .. may not be able to automate, but user could automate the bits used to show the author. 14:51:55 ja look at next file: http://lists.w3.org/Archives/Public/w3c-wai-gl/2004JulSep/att-0196/H_1_1_002.htm 14:52:09 .. it's passed the 1st test, now you need to look at it and make a decision about which test should be run next 14:52:21 .. should it have a text equiv or be null 14:52:30 .. if null goes to a and if not null goes to b 14:57:27 -Tom_Croucher 14:57:39 discussion about knowing where in the path you are given any file. 14:57:46 -Alex_Lee 14:57:59 e.g., H1.1.002 has prereq of H1.1.001. 14:58:22 the way jenae has created, there is a decision tree. if you pass 001 you go "left" and if you fail you go "right" 14:58:37 mc at each node there is a question that either leads to other nodes or terminates in an error report 14:59:21 question/issue about information where you are in the path or how to walk the user through. people feel that is an ert question. wcag should define the path and let the ert/au define how to implement. 14:59:36 UAAG test suite: http://www.w3.org/WAI/UA/TS/html401/ 14:59:42 cr more info here than in UAAG. need to trim to that format? 15:00:07 cr how do we want to format it? 15:00:32 ja give people access to it, they write their own xslt to format it 15:01:17 mc element and attribute are html-specific. need a generalized way to handle that. e.g., "selector" and "property" 15:01:22 ja gateway may have own as well 15:01:29 mc other than that, could easily put in xml 15:01:40 cr assumed these were going to be html-specific. another set for other technologies. 15:01:44 mc framework should be the same 15:02:14 -Tim_Boland 15:03:58 other w3c test suites: http://www.w3.org/QA/TheMatrix 15:04:12 cr what do we need to do to satisfy the w3c QA process 15:04:19 mc execute the tests 15:08:50 discussion about purpose of the test suites: testing the guidelines, verifying the accessibility created by following the guidelines 15:09:16 js need user testing to get complete answer 15:09:31 bc everything that has a test file is required...however we're going to have many optional techniques 15:09:47 mc making sure that guidelines consistently interpretable 15:10:10 ja mapping the examples to html, it wasn't clear 15:10:30 ja we say short equivalent and short label - are those the same things? 15:11:11 js the short label for an audio file will be alt-text for an icon or screen text for what plays the audio recording 15:13:00 js said would like to have examples for each success criterion. (examples that are specifically associated with each criterion) 15:13:12 js e.g., heading, "this is an example of 1.1.a" etc. 15:13:32 js still in the example section, but calling them out as explicitly associated with 15:13:36 ja make it so! 15:13:46 bc the examples for 1.1 are written for the old criterion 15:14:00 bc there are several 1.1 bugzilla bugs related to that 15:14:40 mc when talking about testing files and interpreting guidelines consistently, going back to tool implementations, if chris implements in bobby and michael implements in bobby...will have different results. 15:14:52 mc the only way have consistent results is to develop to the test files 15:15:04 mc if want test files to validate interpretability, tey will define an interpretation 15:15:41 mc there are many differences: one uses a threshold of 60 chars another uses 150 on alt-text length 15:15:52 mc one prompts for "null alt-text or not"... 15:16:12 mc subtle differences...this is why cr wanted such richly defined test files 15:16:40 bc sounds like a scary road. if teach to a test, don't get useful results. 15:16:45 mc would push in that direction 15:16:49 js does consistent mean identical? 15:17:01 js at the F2F talked about "how long is short?" 15:18:41 mc another example: one tool is provide good alt-text another says provide alt-text and it needs to be xyz 15:19:05 mc could be argued that they are consistent, but the 2nd calls out in more detail 15:19:37 cr providing a specific interpretation is scary. however, people will have a clear idea of what we mean. 15:19:53 q+ to say "test files for diff languages? e.g., alt-text length" 15:20:31 js unavoidable that there will be different interpretations, that's not a bad thing. 15:20:39 js there are humans involved. :) 15:20:46 js we're trying to define a range of interpretation. 15:21:16 js i used to teach poetry. exercise: here's a poem, come up with the worst interpretation you can come up with. people come up withwacky stuff. 15:21:29 js so, you do believe that some interpretations are wrong? right. 15:21:34 js we're trying to define the floor. 15:21:47 mc could need more test files to cover the range. 15:22:27 ack wendy 15:22:27 wendy, you wanted to say "test files for diff languages? e.g., alt-text length" 15:22:35 stunned silence 15:22:58 add language to metadata for test cases? 15:23:05 cr for each language that wcag is translated into 15:23:30 wac - re: lang - would need sets of test files for each language, e.g., 150 for english, 1/2 that for german? 15:23:43 js a diff set of requirements might have diff set of numbers 15:25:14 wac perhaps part of metadata: this test can be modified to suit your situation vs this test can not be modified 15:25:26 js configurable by the end user 15:25:35 bc and note the rationale for those 15:25:51 bc 150 may be good measure for current technology quirks, but we may update in future as tech changes 15:26:09 js contextual as well. i don't want to hear 150 alt text for an img on every page 15:26:21 js that is in that line between usability and accessibility 15:26:28 cr the guidelines are open to interpretation. 15:26:39 cr someone could say "images don't have to have alt-text, you can describe them someplace else" 15:26:58 cr we're saying "every img must have an alt-text" if don't like, then suggest a change. 15:27:03 cr it is required of us to provide an interpretation 15:27:31 js agree. question is, "how to we avoid creating canonical interpretations that are more rigid that what we are comfortable with" 15:27:44 mc techniques and test files are not normative, therefore, despite all the work, this is our interpretation 15:28:04 mc if you have a good reason, you could create your own, but we feel we've done a good job 15:29:05 cr number of characters before it is suspicious 15:31:39 [returning to discussion about ja's proposals] 15:31:54 action: wac make sure ja's issues are entered into bugzilla (http://lists.w3.org/Archives/Public/w3c-wai-gl/2004JulSep/att-0196/End_to_End_Guildline_1.1.htm) 15:32:04 -Michael_Cooper 15:32:10 -Chris_Ridpath 15:33:09 action 4 = wac help ja entered issues into bugzilla 15:34:06 RRSAgent, make log world 15:37:12 zakim, who's on the phone? 15:37:12 On the phone I see Jim_Thatcher, Becky_Gibson, James_Craig, Takayuki (muted), Janae_Andershonis, Ben_Caldwell, Paul_Bohman, Wendy, John_Slatin 15:39:01 zakim, close this item 15:39:01 agendum 2 closed 15:39:02 I see 1 item remaining on the agenda: 15:39:03 3. end-to-ends [from wendy] 15:39:33 -Paul_Bohman 15:41:01 bc do we want to do somethin gin css and html to tie back to gateway? 15:41:06 bc want to show the relationships better 15:41:29 -James_Craig 15:44:01 action: ben work on xslt to link from html and css back to gateway 15:44:44 -Becky_Gibson 15:44:45 -Jim_Thatcher 15:44:46 -Ben_Caldwell 15:44:47 -Wendy 15:44:48 -Janae_Andershonis 15:44:49 -John_Slatin 15:45:19 -Takayuki 15:45:20 WAI_WCAG(techniques)10:00AM has ended 15:45:21 Attendees were Jim_Thatcher, +1.202.312.aaaa, Becky_Gibson, James_Craig, Chris_Ridpath, Takayuki, Tim_Boland, Ben_Caldwell, Janae_Andershonis, Michael_Cooper, Tom_Croucher, 15:45:23 ... Alex_Lee, Paul_Bohman, Wendy, John_Slatin 15:45:50 Wendy, thank you for mentioning international problem on "short" alt-text 15:46:01 :) your welceom 15:46:10 welcome 15:46:36 good bye 15:46:47 nabe has left #wai-wcag 16:17:48 cookiecrook has left #wai-wcag 16:40:18 ben has left #wai-wcag 16:55:04 rrsagent, bye 16:55:04 I see 5 open action items: 16:55:04 ACTION: wac create placeholder technique, ednote "1.2 not covered in this draft" (similar to level 3) [1] 16:55:04 recorded in http://www.w3.org/2004/07/28-wai-wcag-irc#T14-34-43 16:55:04 ACTION: wac include placeholders for 1.2 and at least one guideline for each of the other principles (to flesh out the TOC a bit more) [2] 16:55:04 recorded in http://www.w3.org/2004/07/28-wai-wcag-irc#T14-37-28 16:55:04 ACTION: wac ednote about numbering, bc remove numbering for techniques [3] 16:55:04 recorded in http://www.w3.org/2004/07/28-wai-wcag-irc#T14-39-57 16:55:04 ACTION: wac help ja entered issues into bugzilla [4] 16:55:04 recorded in http://www.w3.org/2004/07/28-wai-wcag-irc#T15-31-54 16:55:04 ACTION: ben work on xslt to link from html and css back to gateway [5] 16:55:04 recorded in http://www.w3.org/2004/07/28-wai-wcag-irc#T15-44-01