- From: Frédéric WANG <fred.wang@free.fr>
- Date: Tue, 29 May 2012 15:26:36 +0200
- To: www-math@w3.org
> So in summary what I think the spec says is > > * a missing actiontype is an an error (not detected by the schema) but > any value should be accepted > Yes. > * If the actiontype is known by the processor then the it may ignore > any selection attribute and may have other requirements on the number > of children required for that action I don't think the spec is so explicit about ignoring the selection attribute or having other requirements on the number of children. However, it essentially describes the expected behavior of these known action types so I assumes one obviously understands that the specific description supersedes the general one (I don't know if it was intentional or not but note however that the original developer of the Mozilla code considered the selection attribute even with the statusline actiontype). > > * If the actiontype is not known then the child identified by the > selection attribute (or 1 by default) should be rendered. Yes. > > Does that sound like reasonable behaviour or should we be considering > an errata to change something? Yes, I think that's fine. At least that should allow Andrii to write a patch to conform to this behavior.
Received on Tuesday, 29 May 2012 13:25:34 UTC