https://bugs.documentfoundation.org/show_bug.cgi?id=132891
--- Comment #12 from Mike Kaganski <mikekagan...@hotmail.com> ---
(In reply to Telesto from comment #11)
> .. Starting of paragraph/text in table cell.
That is consistent, too.
Putting cursor anywhere puts it into some existing text span. The decision is
that we always put it into the same span as the character to the left. The
result is that previous formatting is continued.
Starting of a paragraph is tbe place where no characters to the left exist. So
this is *not* inconsistent: it doesn't break the rule. Simply when characters
to the left don't exist, we still end up in existing span - the very first one
of the paragraph.
If you revert the rule, and always get to the same span as character to the
right, you also get consistent behavior - just the opposite - you would prepend
text to yhe formatting to the right of cursor. It's less intuitive - people
tend to think more about continuing than about prepending; but still it would
be consistent. And it would have the same additional point - at the end of
paragraph, where no span to the right exist. Then you have the same thing to
"fix": "I can prepend red text to red part, and not append - but in the end, I
append. Lets expand it from both ends!"
But if you try to do that, you suddenly can't create anything consistent. You
need to decide in which span your cursor is situated, to inherit the
properties. And now it's not something you can easily explain: "if your text
has no character-level formatting on one side of cursor, but some formatting on
the other, the cursor will stick to the formatted side. Oh, but what to choose
when both sides are formatted, but differently? Choose left side or right side?
Let's choose left side. Oh, but now it looks inconsistent for Benjamin, when
left side is formatted English, and right side is formatted Russian+red! User
doesn't see rhe formatting to the left, and expects cursor to stick to the
right part. Now user needs to consider more variables when expecting which
effect their typing will have. And you special-cased not a logical thing (when
just one side exists), but something illogical - tge case where number of
character-level properties in the property bag is 0, not 1, or 2, or... . It's
not consistent; it's not visible to user (consider formatting coming from
paragraph, or invisible formatting). It can't be explained easily. It's just
bad.
--
You are receiving this mail because:
You are the assignee for the bug.
_______________________________________________
Libreoffice-bugs mailing list
Libreoffice-bugs@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/libreoffice-bugs
