Repository navigation
Normatizing the fixed layout property statements #1936
Description
The Fixed Layout properties section of the RS spec makes a number of informative default statements for the various rendition properties, but only one of them is a normative statement.
The default value reflowable MUST be assumed by EPUB Reading Systems as the global value if no meta element carrying this property occurs in the metadata section section [EPUB-33].
The default value statements for rendition:orientation and rendition:spread are both informative (and the same).
The default global value is auto if no meta element carrying this property occurs in the metadata section.
I would propose making these both normative MUST statements to match the statement made for rendition:layout and to clarify this section.
Activity
mattgarrish commentedon Dec 6, 2021
Earlier in the revision we changed from Reading Systems must support the default value to the current prose without a normative keyword because it was said the statements weren't useful for auto values: #1313
The assertion cannot be failed, but that was always by design to allow the default behaviour of reading systems to pass.
Are we okay with testing auto again?
iherman commentedon Dec 10, 2021
The issue was discussed in a meeting on 2021-12-09
List of resolutions:
rendition:orientationandrendition:spreadto MUSTs.View the transcript
3. Normatizing the fixed layout property statements (issue epub-specs#1936)
See github issue epub-specs#1936.
Wendy Reid: i noticed that in RS spec we say in
rendition:layoutthat the default value is a MUST.… but for the other 2 FXL properties even though we also define default behaviour, we don't make those statements normative.
Dave Cramer: there are a couple instances where we say that the default value MUST be assumed, but i didn't know how that would be testable.
Wendy Reid: the default behaviour is auto, and that means to respect what the RS is doing.
… so the test would pass if the epub appears in way that would respect the app settings or the user settings.
… which is essentially what auto is meant to do.
Brady Duga: you could not specify in one case and specify in another case, and if the result is the same then the RS passes?.
Wendy Reid: my proposal was to turn the 2x non-normative default behaviour statements into normative MUSTs.
… i've already made the tests for both of them.
Dave Cramer: so we just ask mgarrish to made the change, then.
adding changes from #1936