MAGIC Magic Mailing List
 
 

From: R. Timothy Edwards (tim AT stravinsky DOT jhuapl.edu)
Date: Fri Jul 19 2002 - 09:35:35 EDT

  • Next message: Atul Gadgil: "lef/def reading in magic"

    Dear Rob,
    
    Anyone have any insights into the problem?  Any known bugs or
    idiosyncrasies in the interaction between Magic and Xlib in the newer
    releases of XFree86?
    
    As far as I know (and I've looked at the code pretty thoroughly), magic's
    16-bit color methods are pretty straightforward.  It's using a "TrueColor"
    visual, which means every color is specified by R, G, B.  To get white
    where another color was intended has to be something very subtle.  I
    would guess that it has something to do with the stipple pattern
    generation.  Anyway, the bottom line is that I find it highly doubtful
    that the magic code is at fault, and that the fault may lie with the
    X server but could be something that only shows up in magic due to the
    combination of colors/stipples/paint functions it uses.  Your thorough
    experiment exonerates the video card, which is the usual source of such
    problems (unless both machines have the same video card?).
    
    Questions:  Have you tried this in 24-bit color, or is the graphics
    card too old/too short on memory to support 24-bit color in any
    reasonable resolution?  Have you tried updating XFree86 to the very
    latest version, which is now 4.2 (see http://www.xfree86.org)?  Also,
    did you install a pre-compiled binary for XFree86, or compile it from
    source?  Are you certain that the X server you're using is the best
    one for the video card you have (e.g., standard SVGA server vs. an
    accelerated server) and/or have you tested more than one X server
    appropriate for your video card?  Any of these things could make a
    difference, as well as obscure XF86Config file settings for the
    server, the screen, the monitor, the video card. . .
    
    I've used an assortment of XFree86 versions, including 3.3, 4.0, and
    4.1, on various machines, and I've never come across the peculiar
    behavior you're seeing.  Not that I haven't come across various other
    peculiar behaviors, some of them fatal to the X server.
    
    						Regards,
    						Tim
    


  •  
     
    Questions? Contact Rajit Manohar
    cornell logo