See also: IRC log
<shadi> http://lists.w3.org/Archives/Public/public-wai-evaltf/2014Jun/0010.html
Finalising draft: Please go down disposition of comments and ensure everything is cloded one by one
Shadi: we should finalise the draft for the WG to
revie
... addressing questions regarding reorganisation of the text
<shadi> http://lists.w3.org/Archives/Public/public-wai-evaltf/2014Jun/0011.html
<shadi> http://www.w3.org/WAI/ER/conformance/ED-methodology-20140610
<shadi> http://www.w3.org/WAI/ER/conformance/ED-methodology-20140610#step1d
Shadi: Addressing several suggestions - first reg. step 1e
<shadi> Step 1.d (define evaluation procedures) seems more related to the
<shadi> use of techniques (or other methods) during evaluation (i.e. Step 4)
<shadi> than to Step 1
Shadi: section 1d was flagged in test run: people
were not sure what exactly to d
... suggest keeping the suggestions from step 1d and move to step 4
<Kathy> +1
Shadi: info on techniques role for testing seems a better fit in step 4
+1
<martijnhoutepen> +1
<Mike_Elledge> +1
<Vivienne> +1
Resolution: move section 1d to step 4
<shadi> http://www.w3.org/WAI/ER/conformance/ED-methodology-20140610#step3d
Shadi: Next on is step 3d (complete processes)
... (describes content of section 3d)
Shadi: suggestion to move parts of 3d (complete proccesses) move selection of pages that are part of the process to step 4 (explained by Martjin)
Kathy: We still need to identify complete processes in step 3
Vivienne: How does moving the choosing of the sample (incl. complete processes) to step 4 ensure that we have actuall ycovered the whole process?
Shadi: (explaining) 1. Mark start of process in step 3 / 2. specify the exact sample in step 4
Mike: Important to retain the idea to include complete processes in sample - step 3 still needed
Detlev: processes with decision points are often too complex to be covered exhaustively anyway - step 4 is picking those pages and states on pages that are present and need to be evaluated, as type
Shadi: Step 3 still valid for making the complete processes included
<Vivienne> I'm fine with that as you identified just now Shadi. As long as the step stays in 3 that we identify any complete processes
<martijnhoutepen> +1
Shadi: OK to move evaluation of complete processes to step 4 - first two steps in algorithm remain in step 3 - items 3 and 4 in algorithm (looking at branches) go to step 4
1+
<Kathy> +1
<Mike_Elledge> +1
Resolution: Move items 3 and 4 in algorithm in step 3d to step 4
<shadi> http://www.w3.org/WAI/ER/conformance/ED-methodology-20140610#step4a
<shadi> http://www.w3.org/WAI/ER/conformance/ED-methodology-20140610#step4c
<shadi> http://www.w3.org/WAI/ER/conformance/ED-methodology-20140610#step1d
Shadi: Fourh point in Email was mergin 4a and 4c (referring to mailing list discussion a while back)
<martijnhoutepen> +1
Mike: Makes sense to merge 4a and 4c - important not to refer to alternative version pages as compliant before they have been checked (and passed WCAG)
Detlev: repeating some content of the discussion of checking of pages with alternative versions
Shadi: Tow issues: what we want in the nethodology regarding the process of checking alternative versions, second merge the text (then it has to be updated anyway)
Vivienne: Not sure why these to sections 4a nad 4c should be merged
Vivenne: important to have them as separate steps - important to identify it as second steps where we discover references to alternative versions
Shadi: Point is that the alternative is usually
identified during sampling/testing so it belongs together - now the two
connected steps are interrupted by step 4b complete processes
... none of the information should be removed - the issue is moving it to
those sections in the methodology where it is needed
Alistair: OK with merging 4a and 4c
... cursory check on default pages will be needed to check whether alternative
versions are properly linked to default pages
... if the link is fine, evaluation should move to alternative version, homing
in on default version would be unfair since the alternative stands in for it
... Methodology needs to rest on definition of "conforming alternative
version"
Shadi: Alistair's case refers to a situation where the alternative version intends to be a full replacement of the default page - bu tthere are many other cases where only some content is renderd on alternative version pages
<Vivienne> that's my understanding also, Shadi
Alistair: once a page is seen as conforming alternative version needs to be thrown out of the sample
Shadi: Understanding is that page and conforming version need to be seen as one entity
Alistair: Must not include issues for which conforming alternatives exist
Shadi: commissioner may want to know what is needed to avoid alternative version altogether
<Vivienne> +1
<martijnhoutepen> +1
<alistair> Merge ok by me
I have an alternative suggestion
<Mike_Elledge> +1
Suggestion to move section 4c to place 4b so it is adjacent to 4a - keeping checing SCs and checking alternative versions close together
Shadi: having them as separate checkpoints is a bit unnatural
I agree (/pesonally)
<Vivienne> +1
Resolution: Merge step 4a and 4c, make sure that content is preserved
Martjin: moving 4b back could mean that you lose a check whether a complete process as an alternative accessible version
Marttjin: swapping may not work well therefore
Martjin: 4d is needed on every page as well
Shadi: difficult to design it so that it flows well
Alistair: Wonders whether 4 c and 4d should be
merged to ensure that the five conformance requirements are check in sync
... $c and $d are linked - this should be checked first to then split off into
checking accessible alternatives, for those without alternaitves you would
check all SCs
<martijnhoutepen> step 4a+4c+4d == ¨Check each page for the conformance requirements¨
Shadi: There were no requirements to remove info
- changed order has a different flavour
... will ty to take a stab at reordering step 4 sections
Shadi; Last issue is to merge 3a 3b and 3c because they are small and some evaluators were confused
Shadi; Last issue is to merge 3a 3b and 3c because they are small and some evaluators were confused
Shadi: Wants to merge without losing any info - would be more straighrfoerward guidance
+1
Martlin: Likes hoe 3 a - b- c- maps onto 2 a - b -c
Martjin: Seemd natural in the test run
Shadi: Some comlpained about frequent cross-referencing between sections, so this is what should be addressed
<Vivienne> sure
Shadi: will finish final editor draft together
with Eric
... Final draft brought to WG review
... Either TF wants to see editor draft first or Shadi and Eric are given
carte blanche to finsalise it and move it to WG review
<alistair> Go straight to review by other groups
Mike: Would like to take a final look at the draft
Shadi: fair dos - we discussed a lot of changes today, also not all comments addressed
<Vivienne> I'd like another look
Alistair: straight to review
<martijnhoutepen> agree with alistair
Detlev: don't mind if you finalise and it goes straight to review
Either way!
<Vivienne> yes please just because there are lots of changes
Alistair: no worth to have a version to look at without option to change it over again
Shadi: Sure changes could still be made
Alistair: why not going via the list so people can review it and object if they find issues
Detlev: What we discussed today aren't substantial changes
Shadi: if preview is wanted that should respected
Vivienne: will go with wht group as a whole decides - may still be useful to look at the changes, also on the list
Shadi: to be clear: sending it to the WG means there wil be a survey where anybody can provide feedback
Vivienne: easy with that
Shadi: another few weeks needed to address coments from WG review proccess
Mike: will go with group as a whole, yields to majority
Shadi: suggests to wrap up comments (earliest on
Monday for ppl to look at) - WCAG meting already on Tuesday so it's tight
... will send update to list ppl can then agree / object to sending to to WG
review
<Mike_Elledge> Thanks all! Especially Shadi and Eric!!