See also: IRC log
<shadi> http://www.w3.org/WAI/ER/EARL10/issues#testmode
<shadi> http://lists.w3.org/Archives/Public/public-wai-ert/2007Jul/0000
SAZ: Issue #7 Test Mode
<shadi> http://www.w3.org/2007/07/04-er-minutes#item05
SAZ: Discussion before but without conclusions so far
JR: could you summarize?
<shadi> Where a software tool WAS ABLE TO IDENTIFY POTENTIAL PROBLEM INSTANCES,
<shadi> BUT WHERE HUMAN judgment WAS STILL REQUIRED TO DECIDE OR HELP TO DECIDE
<shadi> WHETHER AN ACTUAL PROBLEM EXISTS. This should be additionally noted
<shadi> through the earl:compoundAssertor class of the Assertor.
SAZ: semiautomatic is more about the tool taking the decision but with a little human help
JR: manual in ATAG means that tool provide info
and so but the person makes the decision
... in automatic mode the tool makes decision without the need of any human
interaction
... a tool ask to the person look at the image about description
... the person looks at the image and says it's not
JK: other scenario
... the tool ask the user: is this a layout table?
... if the user says yes and there are structural table elements in the
layout table, then the tool takes the decission that this is wrong
SAZ: the difference seems to be that in same cases the tool just go through the document and list things and in other cases it also makes something more "elaborated", a sort of preprocessing
CV: try to distinguish who has the main role could add a lot of confussion
CI: The key question is who has the final word,
if it's the machine it's automatic, if it's the person is manual
... dropping semi-automatic would be the most clear
SAZ: focus on who is doing the hard work instead of the decission
CI: if you need the human intervention it will be manual
JK: so, if there's no human intervention at all it's going to be automatic
SAZ: focus on semi-automatic which is the tricky case
JK: A test tool presents a document to the user
with a tree view
... The user finds a layout table in the tree and comunicates this to the
tool, using a checkbox for example
JR: you need at least have automatic instance location
SAZ: you can look also at "good things", not just problems
JR: will take the rewording to the ATAG WG
... are there repair modes in EARL?
SAZ: no, just evaluation
<scribe> ACTION: SAZ to take the wording back to ATAG and have a final decision next week [recorded in http://www.w3.org/2007/07/11-er-minutes.html#action01]
<shadi> http://lists.w3.org/Archives/Public/public-wai-ert/2007Jul/0008
<scribe> ACTION: everybody to response to the thread on the mailing list [recorded in http://www.w3.org/2007/07/11-er-minutes.html#action02]
<shadi> http://www.w3.org/2007/07/04-er-minutes#item01
SAZ: CV, we expected something by yesterday
CV: will be for the tonight!
<shadi> http://www.w3.org/2007/07/04-er-minutes#item02
<shadi> http://lists.w3.org/Archives/Public/public-wai-ert/2007Jul/0016.html
<shadi> http://lists.w3.org/Archives/Public/public-wai-ert/2007Jul/att-0016/PointersChangesDraft.htm
CI: it's a summarize of the current state of
the art taking into consideration what we have discussed until now, specially
at the last F2F
... there're some open issues to discuss about
... we need to decide what are we going to address before next draft
SAZ: would be good to have issues highlighted to help distinguish them
<scribe> ACTION: CI to highlight issues in the document [recorded in http://www.w3.org/2007/07/11-er-minutes.html#action03]
SAZ: is there anything missing or can we freeze this requirements for the next draft?
<scribe> ACTION: everybody to check the document and look for missing things [recorded in http://www.w3.org/2007/07/11-er-minutes.html#action04]
<shadi> http://lists.w3.org/Archives/Public/public-wai-ert/2007Jul/0009
SAZ: updated with discuss so far