Is there is no scanline function in the bitmap object of Lazarus. Is there no way we can get access to the raw data?
Met vriendelijke groet, Pieter Valentijn Delphidreams http://www.delphidreams.nl -----Oorspronkelijk bericht----- Van: Mattias Gaertner [mailto:[EMAIL PROTECTED] Verzonden: zondag 4 februari 2007 14:56 Aan: [email protected] Onderwerp: Re: [lazarus] Carbon Canvas implementation dilemma On Sun, 4 Feb 2007 14:26:31 +0100 "Ingolf Schäfer" <[EMAIL PROTECTED]> wrote: > On 2/3/07, Tom Gregorovic <[EMAIL PROTECTED]> wrote: > > Hi, > > I think over how to implement canvas for Carbon interface. There are > > two possibilities to use: QuickDraw or Quartz 2D. I want to consult > > it with you. Here is the list of disadvantages, I have found: > > > > QuickDraw > > - DC of Carbon controls must use Window Port > > - deprecated in Mac OS X v10.4 > > > > Quartz 2D > > - unable to implement GetPixel, FloodFill, SetPixel must be > > simulated via line drawing > > - different coordination system (change Current Transformation > > Matrix) > > - bledning instead of ROP > > - bitmap canvas implementation (needs Mac OS X v10.4 because of > > CGBitmapContextCreateImage, but we can get round it) > > > > Feel free to point out anything I have omitted. > > From my point of view the Quartz 2D appears to be more suitable, > > although it doesn't provide per-pixel operations. > Hi, > > although I am not a lazarus developer myself, I dare to to add my two > cents here... > > I fully agree with Tom. Quartz 2D is the ONLY future compatible mac > way in the moment. For the non-mac people, the problem is follows: You > don't have direct access to pixels in Quartz anymore, just floating > point coordinates. As I understand some of the canvas methods require > pixel access. Yes, and they are slow on all platforms, so no time critical application use them. > I have not really thought of this all the way through, but maybe it > would be simple to have an offscreen bitmap graphic which acts as > canvas and use Quartz 2D to map it to the screen. In this way one > would really have pixels to manipulate. Yes, possible. IMO the carbon intf should not use the deprecated API. But the decision is up to the implementer - Tom. Mattias _________________________________________________________________ To unsubscribe: mail [EMAIL PROTECTED] with "unsubscribe" as the Subject archives at http://www.lazarus.freepascal.org/mailarchives _________________________________________________________________ To unsubscribe: mail [EMAIL PROTECTED] with "unsubscribe" as the Subject archives at http://www.lazarus.freepascal.org/mailarchives
