Re: Virtual serial drive

[BLuRry <brendan.robert@gmail.com>]

On Friday, September 7, 2012 12:09:44 AM UTC-5, Michael J. Mahon wrote:
> <aiiadict@gmail.com> wrote:
> 
> > On Thursday, September 6, 2012 6:35:39 PM UTC-7, BLuRry wrote:
> 
> >> Here is a crazy but feasible idea: a replacement rom for the super
> 
> >> serial card that has a prodos block driver signature and all the
> 
> >> transfer code implented as mli calls? 
> 
> > 
> 
> > Even better, piggyback a ROM on the motherboard, so no specific card is
> 
> > required.  Perhaps a scheme like the NSC to map the ROM into address space?
> 
> 
> 
> The NSC scheme is fine for an IC, but is much more complex than just adding
> 
> a ROM. 
> 
> 
> 
> And since you need an SSC anyway, where's the saving?
> 
> 
> 
> -michael - NadaNet 3.1 and AppleCrate II: http://home.comcast.net/~mjmahon

The rom would help turn the SSC into a card that is bootstrapped to work as a block device without the need of loading custom drivers or a slow (and sometimes error-prone) bootstrap step.  The key different is really one of memory management.  If the I/O and prodos block drivers reside in the card rom space, then there's no need to worry about that much less space available to programs or obliterating something else of value like the ram drive or disk drive routines in Prodos to make space.  I'm pretty confident this is a viable option because I just finished implementing two different prodos MLI devices and a generalized MLI handler in Jace.  I know that you can pretty much get away with just implementing the prodos MLI and not the smartport MLI -- provided you don't care about supporting 2MG images of games that use the smartport API directly for disk access.  Even then, the smartport disk access for read/write is extremely similar in function to the prodos MLI so even supporting both is feasible.  (Prince of Persia over a serial port?  Yes please!)

-B