Re: Last Call review of XML Pointer WD Version 1.0

"Martin J. Duerst" wrote:
> At 16:41 1999/12/16 +0100, Chris Lilley wrote:
> > "Martin J. Duerst" wrote:
> > > Please note that for RTL scripts (Arabic, Hebrew), graphical user
> > > selection can lead to the selection of more than one logical parts
> > > (ranges) of a document.
> >
> > And that logical selection, which is more likely to be useful (ie, give
> > you text you can paste somewhere else and have it still make sense) will
> > give a contiguous selection on the document (a single start point and a
> > single end point) although it may display as multiple discontiguous
> > portions on the screen.
> >
> > This is the more common case, since selection is typically initiated by
> > some sort of pointer click and the other end is determined by tracking
> > the current pointer position (and finalised by a second click); this
> > produces a single start and a single end, and the selected portion is
> > all characters in between. This is exactly the same for LTR, RTL and
> > mixed scripts; only the visual result differs.
> As far as user preference is concerned, it seems that using
> graphical selection (setting two points and selecting all the
> text that is visually in between these two points) is more
> common. 

To distinguish these cases, we need to consider bidi otherwise they are

Please demonstrate that visual selection is "more common" than logical
selection in a reasonable sampling of bidi-enabled editors,
wordprocessors or browsers of your choice. My impression is quite the


Received on Wednesday, 22 December 1999 06:44:05 UTC