You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Matt Garrish: when I was making the URL spec change this came up
… when property values expand, they are supposed to expand into valid URL string
… assuming that if it is invalid, then RS should ignore property
Brady Duga: there are also situations where RS is suppose to ignore in failure case
Dave Cramer: right, this causes as little disruption to reading experience as possible
… e.g. not doing something to alert the reader
Proposed resolution: If a property is invalid, ignore it, close issue 1673(Wendy Reid)
Ben Schroeter: +1
Shinya Takami (高見真也): +1
Brady Duga: +1
Matt Garrish: +1
Dave Cramer: +1
Matthew Chan: +1
Fred Chasen: +1
Toshiaki Koike: +1
Wendy Reid: +1
Resolution #3: If a property is invalid, ignore it, close issue 1673
Also will close automatically with #1650. I've added the statement to the updated section on expanding property data types (as linked in the above comment).
Activity
iherman commentedon May 12, 2021
My vote would be to ignore the value (and, possibly, issue a warning)
bduga commentedon May 13, 2021
Where is the current conformance language? I am somehow not seeing it.
mattgarrish commentedon May 13, 2021
Current language is here: https://w3c.github.io/epub-specs/epub33/rs/#sec-property-datatype
Updated language in the open PR is here: https://cdn.statically.io/gh/w3c/epub-specs/fix/issue-808/epub33/rs/index.html#sec-property-datatype
iherman commentedon May 14, 2021
The issue was discussed in a meeting on 2021-05-13
List of resolutions:
View the transcript
7. Handling of invalid properties
See github issue #1673.
Matt Garrish: when I was making the URL spec change this came up
… when property values expand, they are supposed to expand into valid URL string
… assuming that if it is invalid, then RS should ignore property
Brady Duga: there are also situations where RS is suppose to ignore in failure case
Dave Cramer: right, this causes as little disruption to reading experience as possible
… e.g. not doing something to alert the reader
iherman commentedon May 14, 2021
@mattgarrish I did not know whether you want to change something on the spec related to this, so I did not close this issue yet.
mattgarrish commentedon May 14, 2021
Also will close automatically with #1650. I've added the statement to the updated section on expanding property data types (as linked in the above comment).