[Coco] Sugestions for new NitrOS-9 versions

[L. Curtis Boyle]

That might work. I do know that the Eliminator driver had a flag to indicate if the drive was just formatted (set in the Write track function, track 0, from what I remember), so that it would force a refresh of the LSN0 cache, in case the drive parameters had changed with the new format.

L. Curtis Boyle
curtisboyle at sasktel.net



On Apr 2, 2014, at 10:38 AM, Aaron Wolfe <aawolfe at gmail.com> wrote:

> On Apr 2, 2014 12:26 PM, "L. Curtis Boyle" <curtisboyle at sasktel.net> wrote:
>> 
>> Some of the hard drive drivers do sector 0 caching already (since they
> know they are hard drives, and don't have to worry about swapping disks),
> like the B&B and the Eliminator.
>> 
> 
> I remember that being mentioned when I asked about it a while back.  Some
> tests with DriveWire show that it certainty does improve overall system
> speed a lot for such a simple thing, so not surprised this was an
> optimization added in specific drivers.
> 
> Is concern over disk changes the only reason RBF reads it so often in the
> first place?  I'm wondering if there is a universal way to address this or
> if individual drivers is the best approach.  Most modern drive types
> (emulator FDD and hard drives, DriveWire, ide/cf/sd card hdd systems, etc)
> can easily know when a disk has changed (or that it will never change) and
> indicate this to RBF much more efficiently than polling sector 0.
> Wondering if there might be an even simpler way to achieve the same
> improvement by setting a flag in the DD that RBF knows means "this disk
> isn't going to change" or something like that.
> 
> --
> Coco mailing list
> Coco at maltedmedia.com
> http://five.pairlist.net/mailman/listinfo/coco
>