On Dec 27, 2007 7:56 AM, Bruce Markham <[EMAIL PROTECTED]> wrote: > That, my friend, is a really slick idea.
Well i can't take full credit for the idea since singularity does something simular (on the language level, they have a modified version of C#) The idea to use properties like this is mine though (and i'm not a 100% sure if it's a good idea, because it looks like a field now, even though it's really an IO port.. it looks 'clean' and readable though) > But I'm wondering... > > I've been working with PCI. And PCI you use an out() to one port, and then > an in() from another, to read configuration data. > > The parameters written to the first port determine from which PCI > configuration "registers" (and on which devices), that the read from the > second port returns. Can you define better how that would match your > abstraction? I'm not exactly sure what you mean here... but the only reason i implemented only one get and one set per property is because those ports in the example only read or write.. there's no reason why a port wouldn't be able to read and write (get/set) In fact, i see no reason why we couldn't have some sort of symantic that would allow you to have a different port for the get and the set of the property.. There would be attributes for 'fixed' IO ports & memory, for older hardware, and other attributes for dynamically allocatable IO ports and memory.. Other than that, i haven't dived into pci specifics yet, so if i misunderstood you'll have to elaborate.. ------------------------------------------------------------------------- This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2005. http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ _______________________________________________ SharpOS-Developers mailing list SharpOS-Developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/sharpos-developers
