When editing a document, it is often desireable to choose what format context to
be in when inserting text at the boundary between two formats. For instance, if
I have inserted a Hyperlink into my document and I click at the "point" (don't
know the OOo term for the point between one character and the next where the
cursor is positioned... so I'll use "point" in this description...) immediately
after the last character of the Hyperlink and before the next character in my
document (just regular text for the sake of argument), I might want to type in
the Hyperlink text context (to add something to the text for the Hyperlink) or
the regular text context (perhaps to add a space or punctuation). As it is now,
typing at that "point" will always be within the context of the Hyperlink, no
matter how my cursor got to that "point". A good example is a Hyperlink followed
by a comma. If I want to add a few words after the Hyperlink but before the
comma, I have to position the cursor to the right of the comma, type the space
and words I want and a new comma, then go back to the "point" after the original
comma and press backspace to delete the original comma.

The same is true of editing at the boundaries of text with italics or other text
formats, and some other text format.

The solution that I am used to is that the text format context at "point" is the
context that the cursor was last in. So, for instance, if I clicked somewhere in
the text for a Hyperlink (although any text format/context will do) and used the
cursor keys to move to the "point" immediately after the last character in that
context, if I typed I would still be in that context. However, if I clicked on
text in the region after the Hyperlink context (where "after" means further
along in the direction of left/right or right/left text), and used the cursor
keys to move back to the "point" immediately after the last character in the
Hyperlink -- but without "entering" that context -- when I typed text, it would
be in the context of the text after the Hyperlink. If I clicked at a "point"
with the mouse, then the behaviour would be identical to its current
implementation, and the context would be that of the text immediately before the
"point".

I find that when I'm working on documents that I am faced with this issue
several times an hour, and I presume that others are facing similar issues, so I
set the priority a bit higher.