19:51:14 RRSAgent has joined #wai-wcag 19:51:19 RRSAgent, make log world 19:57:22 zakim, this will be wai_wcag 19:57:22 ok, sh1m, I see WAI_WCAG()4:00PM already started 19:57:32 zakim, who's on the phone? 19:57:32 On the phone I see ??P0 19:57:40 Zakim, ??P0 is Tom 19:57:40 +Tom; got it 19:57:43 Zakim, I am Tom 19:57:43 ok, sh1m, I now associate you with Tom 19:59:11 +Yvette_Hoitink 19:59:58 +James_Craig 20:00:06 +Katie_Haritos-Shea 20:00:10 +Michael_Cooper 20:00:11 zakim, I am Michael_Cooper 20:00:11 ok, Michael, I now associate you with Michael_Cooper 20:00:48 +Bengt_Farre 20:01:29 +Matt 20:01:36 zakim, mute Bengt_Farre 20:01:36 Bengt_Farre should now be muted 20:01:45 zakim, I am Bengt_Farre 20:01:45 ok, bengt, I now associate you with Bengt_Farre 20:01:46 MattSEA has joined #wai-wcag 20:01:56 sh1m has left #wai-wcag 20:02:02 Zakim, I am James_Craig 20:02:02 ok, james, I now associate you with James_Craig 20:02:06 +??P7 20:02:22 zakim, ??P7 is Ben_and_Gregg 20:02:22 +Ben_and_Gregg; got it 20:02:33 zakim, who's on the phone? 20:02:33 On the phone I see Tom (muted), Yvette_Hoitink, James_Craig, Katie_Haritos-Shea, Michael_Cooper, Bengt_Farre (muted), Matt, Ben_and_Gregg 20:02:46 Rob has joined #wai-wcag 20:03:01 bcaldwell has joined #wai-wcag 20:03:33 + +1.540.998.aaaa 20:03:47 zakim, who is here/ 20:03:47 I don't understand 'who is here/', bcaldwell 20:03:56 zakim, who is on the phone? 20:03:56 On the phone I see Tom (muted), Yvette_Hoitink, James_Craig, Katie_Haritos-Shea, Michael_Cooper, Bengt_Farre (muted), Matt, Ben_and_Gregg, +1.540.998.aaaa 20:04:01 sh1m has joined #wai-wcag 20:04:13 +Wendy 20:04:50 zakim, +1.540.998.aaaa is Rob 20:04:50 +Rob; got it 20:05:01 zakim, Rob is Robert_Fentress 20:05:01 +Robert_Fentress; got it 20:05:35 http://www.w3.org/2001/12/zakim-irc-bot.html 20:05:36 +Becky_Gibson 20:05:40 Zakim Guide 20:05:57 Becky has joined #wai-wcag 20:06:09 example: q+ to say "note to self" 20:07:51 q+ Content mus be accessible with JS disabled? 20:08:04 + +1.512.918.aabb 20:08:14 zakim, +1.512.918.aabb is Andi 20:08:14 +Andi; got it 20:08:41 q+ JS alternatives 20:08:44 Scribe: Andi 20:08:57 rob: Javacript is not on the agenda today. 20:08:58 Andi has joined #wai-wcag 20:09:12 wendy, i thought it was 20:09:27 Agenda: http://lists.w3.org/Archives/Public/w3c-wai-gl/2004JulSep/0368.html 20:09:39 Other discussions that have occurred on the list recently: 20:09:39 1. Javascript and alternatives 20:09:39 20:09:50 it's in the list 20:09:52 yes, you're right. 20:09:53 sorry. 20:09:55 its very hot here too :) 20:10:04 It's last on the list 20:10:07 Wendy, can I assign myself as scribe? 20:10:24 Andi, I've already assigned you. 20:10:29 gregg has joined #wai-wcag 20:10:33 +Alex_Lee 20:10:43 zakim, Alex_Lee is Alex_Li 20:10:43 +Alex_Li; got it 20:10:47 Authored unit 20:11:13 Discussion with device independence working group 20:11:26 zakim, who is on the speaker queue 20:11:26 I don't understand 'who is on the speaker queue', james 20:11:33 their definition is different from ours - orthogonal - not same dimension we are using it in 20:11:40 q? 20:11:47 q- JS 20:11:54 q- alt 20:11:59 arose at the last f-2-f 20:12:08 haven't figured out if it works but have been exploring it 20:12:23 way to talk about conformance 20:13:12 authored chunk to us is something that has a uri, not a "#" uri within another one 20:13:38 can contain other authored chunks 20:14:26 conformance of an authored chunk is dependent on the conformance of the other chunks it contains 20:15:22 +[Microsoft] 20:15:33 Wendy speaking about differences between our use and DI use 20:15:50 q? 20:15:57 we're trying to describe how to collect components in order to make a conformance claim 20:16:33 person who creates the image might be different from the person who writes the alt text 20:16:48 image and alt text is a collection that might be included in a larger collection 20:17:07 zakim, [Microsoft] is Mike_Barta 20:17:07 +Mike_Barta; got it 20:17:13 Wendy proposes "collection of authored units" 20:17:22 q+ 20:17:29 Gregg speaking 20:18:02 Wendy are you suggesting we use "collection of authored units" as the term or that we come up with a new term 20:18:14 ack [Microsoft] 20:18:17 Wendy - think it's okay to use "collection of authored units" 20:18:35 zakim, [Microsoft] is Mike_Barta 20:18:35 sorry, bcaldwell, I do not recognize a party named '[Microsoft]' 20:18:43 ack tom 20:19:03 Tom - agree with Wendy - talking about compositions of authored units 20:19:49 DI did not intend for authored unit to refer to a collection of authored units 20:21:21 Gregg - if authored unit contains other authored units, then an authored unit is a collection of authored units with additional information 20:21:41 q+ to say thats why I like the term composition 20:21:48 q- 20:23:00 Wendy believes that authored unit does work for us - doesn't understand the DI issue 20:23:03 ack Alex 20:24:33 -Robert_Fentress 20:24:37 Wendy - DI covers both aggregation of content and converting content 20:25:30 Gregg - DI says "authored unit" has multiple modes of perception depending on delivery 20:25:42 might look different if fetched by computer, PDA, or phone 20:25:45 also might change over time 20:25:55 q? 20:25:59 changing over time fits our definition 20:26:29 our use of it also allowed for content negotiation 20:26:34 q+ to say that content negociation doesn't work that way 20:26:36 ack tom 20:26:37 Tom, you wanted to say that content negociation doesn't work that way 20:26:54 Tom - does not agree that our use covers content negotiation 20:27:04 zakim, mute me 20:27:04 Yvette_Hoitink should now be muted 20:27:13 issues with definition of URI and concept of content negotiation 20:28:16 DI definition includes content that is not human authored; that is, delivered from a database perhaps 20:28:21 q+ to "clarify the diff pieces for DI" 20:29:10 Wendy - DI has separated between the chunks of material an author creates (authored units) 20:29:25 chunks delivered to the device (delivered units) 20:29:34 and the chunks perceived by the user (perceived units) 20:29:52 q+ 20:30:00 we may be talking about combining authored and delivered units, and maybe the perceived unit too 20:30:01 ack wendy 20:30:01 wendy, you wanted to "clarify the diff pieces for DI" 20:31:12 Tom - can have a case where the template can enforce accessibility on the delivered unit 20:31:47 Delivery Unit 20:31:47 A set of material transfered between two cooperating web programs as the response to a single HTTP request. The transfer might, for example, be between an origin server and a user agent. 20:31:47 Users are not normally aware of individual delivery units. 20:31:56 aggregation - hard to make claims on the delivered unit because can't guarantee that the authored units that are aggregated are conforming 20:32:06 +Loretta_Guarino_Reid 20:32:11 Gregg - thinks we want to talk about delivered unit 20:32:15 http://www.w3.org/TR/di-gloss/ 20:32:24 Delivery Context 20:32:24 A set of attributes that characterizes the capabilities of the access mechanism and the preferences of the user 20:32:32 if delivered unit is accessible, doesn't matter if the authored unit was accessible or not 20:33:01 content may be in content management system but are run through a process that makes them accessible 20:33:41 Perceivable Unit 20:33:41 A set of material which, when rendered by a user agent, may be perceived by a user and with which interaction may be possible. 20:33:41 User agents may choose to render some or all of the material they receive in a delivery unit unit as a single perceivable unit or as multiple perceivable units. 20:33:41 Most perceivable units provide both presentation and the means for interaction. However, on some types of device, such as printers, perceivable units might contain only presentation. 20:33:51 q+ to ask about conformance 20:33:55 how do you make conformance claims on a delivered unit 20:34:17 maybe we use "served unit" 20:35:28 Wendy reads definition of "Delivery Unit" 20:35:39 also consider "delivery context" 20:36:42 Tom - user profile allows users to request a piece of content that is particular to their needs but might be inaccessible to someone else 20:37:03 Gregg - should be able to negotiate for one that meets the WCAG guidelines 20:37:34 person serving content is not responsible for what happens to it in the pipeline 20:37:39 q+ to say "page?' 20:37:50 Wendy clarifies that it is "delivery unit", not "delivered unit" 20:38:11 Gregg - what is delivered by default must be accessible. 20:38:38 even if it is possible to negotiate for a version that is not accessible 20:39:38 Tom - should still be able to make a claim for an authored unit if you can guarantee it will conform 20:40:00 Gregg - what would be the purpose of making a claim on the authored unit 20:40:10 Wendy - important for aggregation 20:40:30 AOL provides the delivery unit for various authored units 20:40:38 syndicators provide claims on the authored units 20:40:47 AOL can then make a claim on the delivery unit 20:41:07 q+ 20:41:50 q- 20:41:52 Gregg - if syndicators deliver accessible authored units, aggregator may strip out accessibility information 20:42:07 zakim, mute me 20:42:07 Michael_Cooper should now be muted 20:42:07 aggregator could also add accessibility information to syndicated authored units 20:42:15 so what really matters is the delivery unit 20:42:15 zakim, unmute me 20:42:15 Yvette_Hoitink should no longer be muted 20:42:29 q+ to say "but it may be a business rule that AOL would only accept accessible authored units" 20:42:32 easier for aggregators if authored units come in with conformance claims 20:43:16 Gregg - delivery unit is the only thing that really matters but it is useful to those creating the delivery units if the authored units have conformance claims 20:43:17 ack yvette 20:43:17 Yvette_Hoitink, you wanted to say "page?' 20:43:24 q+ to talk about other use cases 20:44:04 Yvette - difficult to understand this concept 20:44:13 Gregg - a delivery unit is that what you get form a URI 20:44:53 Yvette - delivery unit is non-HTML specific version of "page" 20:45:32 ack bcaldwell 20:45:42 Ben - don't see the distinction 20:45:55 syndicators deliver a unit to AOL which they pass on 20:46:11 image by itself can't be accessible 20:46:23 must be combined with alt text 20:46:55 Gregg - if you don't actually deliver content but you sell content to those who do deliver it 20:47:24 wouldn't it be good to say that your authored unit (aggregate delivered unit) will be accessible when delivered 20:48:03 Tom - many ways that something could be delivered 20:48:07 ack Mike 20:48:36 Mike - if you receive something and you strip out something, it is a new authored unit and you are responsible for the claim 20:48:56 still useful to have the claim on the original authored unit 20:49:00 zakim, mute me 20:49:00 Yvette_Hoitink should now be muted 20:49:14 Gregg - claims are all on delivery units 20:49:19 q- 20:49:27 q+ 20:49:28 either content delivered from someone else that I re-deliver or something that I authored 20:49:36 q+ to ask, "are RSS feeds http requests?" 20:50:13 ack tom 20:50:13 Tom, you wanted to talk about other use cases 20:50:29 Tom - need to be able to make claims about content management systems 20:50:52 possible to define CM system so that it is impossible to deliver inaccessible content from it 20:51:18 helpful if people can make conformance claims against something that is more abstract 20:51:32 ack gregg 20:51:37 shouldn't prevent people from making claims against authored units 20:52:26 Gregg - want to guard against people making claims against authored units that might not be delivered in an accessible way 20:53:32 user only cares about the deliverey unit 20:54:16 Wendy - looking at definition of delivery unit, still have sense that there is a difference for aggregators 20:54:52 still thinks that what content providers are providing is authored unit 20:55:11 ack wendy 20:55:11 wendy, you wanted to ask, "are RSS feeds http requests?" 20:55:15 zakim, mute me 20:55:15 Yvette_Hoitink was already muted, Yvette_Hoitink 20:55:28 Alex - example - if intellectual property belongs to someone else, aggregator would have limited ability to alter it 20:55:44 similar to Reuters where IP belongs to someone else and you host it 20:55:57 does that put aggregator in a position where they can't make a claim 20:56:08 q+ to say wendy, the rss feeds are still "delivery units" even though the "receiving user" is a program aggregating those feeds into another "delivery unit" 20:56:19 Gregg - if you buy inaccessible content, may make it impossible for you to make a claim 20:57:05 james, thanks. but are rss feeds "responses to a single http request?" if so, then they don't meet the current defn of delivery unit: 20:57:05 Delivery Unit 20:57:05 A set of material transfered between two cooperating web programs as the response to a single HTTP request. The transfer might, for example, be between an origin server and a user agent. 20:57:05 Users are not normally aware of individual delivery units. 20:57:13 aggregators would like to be able to ask sources to send accessible content 20:57:25 s/if so,/if not, 20:57:28 in order to make claim based on contractual agreements 20:57:43 accessibility is not about "effort", it's about "actual" 20:57:52 -Mike_Barta 20:58:22 legislators can decide if doesn't have to be accessible because of undue burden 20:58:38 Alex - can't alter someone's IP material 20:59:07 copyright law allows you to change someone's content in order to make it accessible 21:00:20 q? 21:00:23 ack James 21:00:23 James_Craig, you wanted to say wendy, the rss feeds are still "delivery units" even though the "receiving user" is a program aggregating those feeds into another "delivery unit" 21:00:45 James - RSS feed is still a delivery unit even though not delivered to a human user 21:01:40 Wendy - definition of delivery unit talks about between two computers in response to a single http request 21:01:46 does this fit RSS feeds 21:01:52 James - yes, in my experience 21:02:09 Wendy - if RSS feeds meet the definition of delivery unit then agree that this is a good term to use 21:02:30 q+ 21:02:42 Gregg - we said it's something you get from a URI. DI says it's a single http request 21:02:58 isnt it URL 21:02:58 can you get something from a single http request that is not from a URI 21:03:14 URI includes http but is not limited to http 21:03:33 http is the protocol you use to access the authored (or delivery) unit 21:03:55 Wendy - go back to DI group and validate use of delivery unit in our context 21:04:27 bengt, URL (locator) and URI (identifier) have a slightly semantic difference... 21:04:49 Gregg - anything that is not http is out of our scope 21:05:07 james, yes http gets from URL ? URI is more than both ? 21:05:10 James - what about video rt 21:05:16 ack matt 21:05:18 RTSP 21:05:22 Gregg - doesn't meet the definition of delivery content - single http request 21:05:36 real time streaming protocol... SMIL, RM, etc 21:05:54 from James - rt is rtsp 21:06:27 RSS is not a final-form language. 21:06:41 Like SOAP, it carries a payload which can be accessible content or not. 21:06:42 it is in my rss reader 21:07:02 It is not intended to be referenceable by a URI, for example. 21:07:08 Rather it points _to_ a URI containing full content. 21:07:08 ga 21:07:14 Wendy - do I need to re-type this for the minutes? 21:08:37 no 21:08:50 yes 21:08:58 One HTTP call brings down the feed for an entire site. 21:09:03 Gregg - how did rtsp movie get initiated? 21:09:15 For RTSP, you usually send something over HTTP to get a pointer. 21:09:59 +Matt_May 21:10:25 Gregg - trying to get closure on term 21:10:48 delivery unit seems to work but it's definition includes "delivered by a single http request" 21:10:51 q+ to say, "you have a template (an authored unit) that generates content. you could make a claim on the authored unit for WCAG and the process that fills in the template for ATAG and then those together is the delivery unit" 21:11:12 want to make sure it covers other kinds of streaming media transferred via different kinds of requests 21:11:54 Matt - protocol is irrelavant to accessiblity 21:12:15 Gregg - know it is transferred in other formats and accessibility is not affected by protocol 21:12:19 issue: delivery unit brings up issues related to web services and overlap with atag 21:13:17 but delivery unit defined as http so we are trying to determine if can use delivery unit term for these types of content 21:13:46 Matt - really talking about content that can be returned as result of URI request 21:14:13 zakim, unmute me 21:14:13 Yvette_Hoitink should no longer be muted 21:14:34 Gregg - if DI group says we can use delivery unit for something that is delivered as the result of a single URI request, then are we okay using that term? 21:15:39 action: ask DI group if delivery unit can be used to for content that is delivered as the result of a single URI request 21:15:45 any idea when? 21:16:51 Gregg, Wendy, Tom, James, Matt, Mike will pursue with DI group - get on their agenda next Wednesday 1500 UTC (10 Eastern) 21:16:57 same time as techniques task force 21:17:25 action: Gregg e-mail to DI chair to propose getting time on their agenda next week 21:17:31 action: Gregg send email to Rhys to request 1/2 hour of agenda to discuss authored unit at their call next week (10 am Eastern) 21:17:50 Gregg is hoping for a short discussion :) 21:18:07 ack wendy 21:18:07 wendy, you wanted to say, "you have a template (an authored unit) that generates content. you could make a claim on the authored unit for WCAG and the process that fills in the 21:18:10 ... template for ATAG and then those together is the delivery unit" 21:21:43 Wendy - just breaking it down into which parts have to conform to WCAG and which parts have to conform to ATAG 21:22:52 trying to define how to combine ATAG piece in the conformance claim for content that comes from a content management system 21:23:29 q+ 21:24:01 Tom - with flat html file can make claim against delivery unit 21:24:25 but for dynamic content, not reasonable to make delivery unit claim when there is potential for pages to become inaccessible 21:24:37 need mechanism that shows why it is accessible 21:24:55 pieces of content that are plugged into a template that meets WCAG 21:25:10 q+ 21:25:12 process for plugging the content into the template conforms to ATAG 21:25:18 ack tom 21:25:56 Gregg - nothing in guideline that allows you to claim a template is compliant 21:26:41 using ATAG compliant generator does not guarantee WCAG compliant content 21:27:40 q+ to "wrap up" 21:27:52 Tom - trying to propose something that is more than just a promise of conformance 21:27:53 i think i have an idea for action for me, tom, matt and auwg 21:28:06 feasible to test a template for WCAG conformance 21:28:13 feasible to test something for ATAG conformance 21:28:40 not feasible to test something that is dynamically generated frequently for WCAG conformance 21:28:52 ack gregg 21:28:55 Wendy - this is a big issue 21:28:58 ack wendy 21:28:58 wendy, you wanted to "wrap up" 21:29:06 Tom and Wendy willing to take action item 21:29:26 first issue is how to reference ATAG from WCAG - specific request from ATAG at f-2-f 21:29:39 q+ 21:29:51 second - need to work with ATAG group to understand how to make conformance claims for content management systems 21:30:07 third - go through concrete example - maybe ATAG has already done this 21:30:28 ack matt 21:30:39 Matt - agree with Wendy 21:31:09 in order to conform to ATAG, authoring tool has to check for accessibility at a certain level and prompt for content needed for accessibility 21:31:42 action: matt, wendy, tom: how ref atag from wcag, conformance claims for wcag/atag, concrete example to demonstrate overlap (to help raise awareness of relationship between the two, primarily within WCAG WG since ATAG fairly aware due to dependence on WCAG) also joint action with AUWG to agree on solution. 21:31:43 can't measure 100% on the delivery context because that changes from session to session 21:32:14 ack greggg 21:32:18 ack gregg 21:32:23 Gregg - agree that we want to emphasize ATAG in introduction 21:32:37 don't think ATAG should be incorporated into any of the guidelines themselves like we did for UAAG 21:32:45 possible to embed a user agent in content 21:32:53 rather than re-write, we refer to sections of it 21:32:59 it being UAAG 21:33:07 q+ to say, "yes, can embed ua into web content, however web content could be an authoring tool. ATAG applies to CMS" 21:33:10 ATAG is a tool for creating WCAG conforming material 21:33:37 q+ to say, "Authoring Tools can also be embedded in accessible content" 21:33:43 content generated using ATAG conforming tool does not guarantee accessibility - can prompt for information but information provided may not be accessible 21:33:55 and the tool can only check certain things 21:34:11 ack wendy 21:34:11 wendy, you wanted to say, "yes, can embed ua into web content, however web content could be an authoring tool. ATAG applies to CMS" 21:34:29 Matt - true of all tools - need process that makes it as difficult as possible to create something that is not accessible 21:34:43 Wendy - web content can itself be an authoring tool 21:34:52 that's why the relationship is so critical 21:35:35 web applications that create web content have to also conform to ATAG 21:36:40 but this is a circular reference so we have to resolve that 21:36:50 Gregg - sounds like we need a separate guideline on this 21:37:24 Wendy - not clear if it is a guideline or if it affects conformance 21:39:18 Gregg - doesn't ATAG reference UAAG for accessibility of user interface 21:39:28 Matt - no, it references ISO 16071 21:39:43 q- 21:39:52 sh1m has joined #wai-wcag 21:42:39 discussion about "process" 21:42:49 never measure accessibility by "process" 21:43:25 -Michael_Cooper 21:45:28 -Katie_Haritos-Shea 21:45:29 -Alex_Li 21:45:33 -Becky_Gibson 21:45:34 -Bengt_Farre 21:45:35 -Wendy 21:45:36 -Loretta_Guarino_Reid 21:45:37 -Andi 21:45:38 -Matt_May 21:45:39 -Ben_and_Gregg 21:45:40 -Yvette_Hoitink 21:45:41 -James_Craig 21:45:42 -Matt 21:45:43 -quit 21:45:44 -Tom 21:45:47 bengt has left #wai-wcag 21:48:30 Rob has left #wai-wcag 21:48:34 RRSAgent, bye 21:48:34 I see 4 open action items: 21:48:34 ACTION: ask DI group if delivery unit can be used to for content that is delivered as the result of a single URI request [1] 21:48:34 recorded in http://www.w3.org/2004/08/12-wai-wcag-irc#T21-15-39 21:48:34 ACTION: Gregg e-mail to DI chair to propose getting time on their agenda next week [2] 21:48:34 recorded in http://www.w3.org/2004/08/12-wai-wcag-irc#T21-17-25 21:48:34 ACTION: Gregg send email to Rhys to request 1/2 hour of agenda to discuss authored unit at their call next week (10 am Eastern) [3] 21:48:34 recorded in http://www.w3.org/2004/08/12-wai-wcag-irc#T21-17-31 21:48:34 ACTION: matt, wendy, tom: how ref atag from wcag, conformance claims for wcag/atag, concrete example to demonstrate overlap (to help raise awareness of relationship between the two, primarily within WCAG WG since ATAG fairly aware due to dependence on WCAG) also joint action with AUWG to agree on solution. [4] 21:48:34 recorded in http://www.w3.org/2004/08/12-wai-wcag-irc#T21-31-42 21:48:53 zakim, bye 21:48:53 leaving. As of this point the attendees were Tom, Yvette_Hoitink, James_Craig, Katie_Haritos-Shea, Michael_Cooper, Bengt_Farre, Matt, Ben_and_Gregg, Wendy, Robert_Fentress, 21:48:53 Zakim has left #wai-wcag