On Tue, 2 May 2006 16:31:22 -0400 DJ Delorie <[EMAIL PROTECTED]> wrote:
> If the core knows it's drawing on the solder side of the board, it > needs to swap the signs of the *board* coordinates it's using, so that > on the physical board it reads correctly. > > If the GUI happens to be showing the board from that side, it would > re-flip it so it reads correctly on the screen. If not, it would > remain flipped and read backwards on the screen. > > I think what Tamas is complaining about is that the core swaps the Y > coordinate, but Gtk swaps the X coordinate, so even when you're > looking at the correct side of the board, the text is not mirrored > but is upside-down. > > I'm not sure there's a good way to manage this, other than rotating > the parts 180 degrees after flipping them. Eventually the GUIs will > support both X and Y axis board flipping (and perhaps board rotating > too), so no single "default" element flip will be sufficient. > > And if all this confuses you, that only means you're paying attention. > It's confusing. The rule is: nothing in the "core" space should know > about SCREEN COORDINATES (i.e. zoom, pan, flip view). That doesn't > preclude confusing calculations on BOARD COORDINATES, such as flipping > text or elements to put them on the other side of the board, or > rotating them. Then the core is currently defining the flip direction. Shouldn't the macros SWAP_SIGN then be replaced with HID functions so the gui can be in control?