dwb wrote:
Writing a set of routines to put shapes on the canvas and move them is relatively trivial. Grouping those shapes is also trivial (but don't get me started on how non-trivial grouping is if we add transforms to the toolset). You'd be surprised how a simple 90º transform on a square becomes a nightmare when we have a group of squares.) Connecting shapes with elastic lines is another totally new nightmare.
But is this not a possible problem for Pages? There is plenty of functionality I would not have have added to the product, and which I guess was added just because it was easy: shapes, bezier curves, alpha-channels, picture adjustments, and so on.
On the other hand there are things like the tables, that people still do not come to grips with. One problem is when people feel the need to let a cell span pages, but I have a feeling that there are much more problems than that with tables. I just find it too much of a hassle to isolate exactly what goes wrong.
I realise that fixing tables may be difficult, especially as I cannot even pinpoint exactly what is wrong. But surely tables is much more of word processor functionality than bezier curves?
From a programmer perspective, I realise it is cheaper to add bezier curves. From a user perspective, I think it is strange that tables do not work "properly".
From a programmer perspective, I realise it is easier to add just shapes than adding connector points. However, from a user perspective, I think it is strange to implement what I perceive as half baked functionality: shapes without connectors.