From: Groepaz (groepaz_at_gmx.net)
Date: 2003-04-28 14:02:51
On Monday 28 April 2003 13:51, Ullrich von Bassewitz wrote: > On Mon, Apr 28, 2003 at 01:42:39PM +0200, Groepaz wrote: > > i would prefer to have it changed (the current implementation can not > > handle more than just one connected drive either, right?) > > It can. Just change _curunit before opening a file. uhmz well... not very portable :=P (i'm not only looking at cc65 targets) i wont call this more than "workaround" until we have something better. > I think it is. read(), write() and friends will have to adjust their work > depending on the output device, which in turn means you need "real" > drivers, with function vectors, the handle must encode the driver, and so > on. Currently, the POSIX functions are machine dependent, which means that > the changes have to go into all the existing sources. Doesn't sound like an > easy task for me... mmmh...well still easy, maybe lot of work (correcting myself here) :O) > Hmmm. How about implementing a RAM disk via a driver, and add another > (platform dependent) module that links the first one into the CBM kernal > calls? ? explain :=) however, i would still like the more generic approach... in the future it would allow to have special drivers for each drive, use "irq loaders", have dedicated 1541 code with seeking and all...etc pp. maybe needs more effort to create at first, but IMHO its a lot more useful in the end. gpz ---------------------------------------------------------------------- To unsubscribe from the list send mail to majordomo_at_musoftware.de with the string "unsubscribe cc65" in the body(!) of the mail.
This archive was generated by hypermail 2.1.3 : 2003-04-28 14:06:34 CEST