Paul Libbrecht wrote:
Jonas Sicking wrote:
http://www.activemath.org/~paul/tmp/Ideal-Browser-Behaviour.html
I am still slightly worried about pages denying access to copy data,
but I think it would be worth giving it a shot. Browsers could always
require a higher level of security in case it turns out that pages
abuse it too much.
Cool!
However, I would like to see it defined in terms of an event rather
then attributes.
Good idea, although I fear I may be half-skilled with the appropriate
language for it.
In any case, do you have other pointers to the matter outside of ?
http://www.w3.org/TR/DOM-Level-3-Events/events.html
Are there some current convention in the webAPI documents for events ?
That would be it. The formal language isn't that important at this stage
since we'll hammer that out if/when this is put in a spec.
Also, it is not clear to me that onselected or oncut is needed.
I think oncut is absolutely needed if implementing something like a
visual editor... and it matches the standard gestures.
onselected is fragile... at least... it was just invented.
But I know it matches what I see the two use-cases would need to be
complete:
- for the formula case, the selection should, ideally, only be made to
whole terms... so that the handler or event could trigger an enlargement
to a span representing the term at least if crossing the term
- for the travel-schedule case, I would at least see value of such
enlargement in case more than one cell of the schedule is selected.
You don't want to dynamically screw with the selection while user is
selecting text, that will probably cause confusion at the least.
There is already ways to do this. CSS3 (I believe) adds properties that
allows you to control whether an element can be selected in part, only
as a whole, or not at all.
/ Jonas