>> Nope! In MegaSCSI you can define the size of each partition (when formatting
>> the device) with any amount of sectors. When you change the disk you don't
>> need to do nothing if the new disk has the same partitionning layout as the
>> previous; else, you must only to run the partition changer.
>
>What if you have one Zip-disk with 3 partitions and other with 4
>partitions ? Everything is still working Ok ?
Yes, the partition changer works no matter of the partitionning state of the
disks. Howewer this program is not perfect yet, and sometimes goes wrong
(and even hangs the computer) when you use non-ASPI disk.
>> However the most normal (and logical) case is to act as you say: making all
>> the partitions of the same size (32MB) except the last one.
>
>No exepts are needed if I'm correct. 3 * 32MB = 96MB that is actual size
>of formatted Zip-disk. Atleast my fdisk proggram says that all of the
>capasity is used, when I use 3 * 32MB format.
The maximum amount of clusters handled by MSX-DOS is 4085, not 4096.
Therefore the actual maximum size of the partitions is somewhat smaller than
32MB, and a little of disk space remains for a fourth partition. I define my
partitions with 64000 sectors, so this extra partiton grows to about 2MB,
enough for some little utilities.
>> This is not necessary if you boot from the SRAMdisk (that is, if you assign
>> it to drive A:). This is very useful, because you have the same booting
>> state regardless of the inserted disk.
>
>Is it allways A-drive ? With Novaxis I have notised, that boot HD can be
>A-drive or C-drive depending of your slot configuration. I think, that
>C-drive is better because then I can boot from disk, and HD-boot is
>enabled only if disk was not present. (Just like in stupid PC)
>Anyway if I want to use this feature for example in MSX tR I need to add
>another disk controller before SCSI controller. (This is all tested
>information)
Boot drive is always A:, but you can assign this drive to ESE-RAMdisk or to
any device, or even leave it unconnected (but this is a little stupid!) 8-).
Booting from diskette is never allowed in MegaSCSI, except if you boot
pressing the appropriate key for disabling it.
>> That's right. Maybe patching any hook on the disk work area this can be
>> solved, because for check the disk change it is enough to read the serial
>> number on the boot sector, and this is independent of the controller.
>
>It is not that easy... You have to update also all partition info from
>RAM and then starts battle against DOS. If someone has any glue about DOS2
>internal variabbles, then please... DOS2 seems to keep same information
>usually in many places.
ESE guys have this information and much more, else it is not possible to
implement all the features of MegaSCSI (it has even a internal call for
creating a new drive letter and assign it to a device!!)
>> It's only that MegaSCSI designers took some time using the brain before the
>> construction step. It is not so difficult.
>
>We can learn ( = COPY ) much from them. :-)
Unluckily the only way is to DISSASSEMBLE their stuff, because now they are
not easy to find via e-mail... because of this, my FAT16 project is frozen.
-------------------------------------------------------
Konami Man - AKA Nestor Soriano (^ ^)Y
Itsumo MSX user
[EMAIL PROTECTED]
http://www.geocities.com/SiliconValley/Bay/9797/msx.htm
This Pentium will become obsolete in 5 seconds...
4... 3... 2...
-------------------------------------------------------
****
MSX Mailinglist. To unsubscribe, send an email to [EMAIL PROTECTED] and put
in the body (not subject) "unsubscribe msx [EMAIL PROTECTED]" (without the
quotes :-) Problems? contact [EMAIL PROTECTED] (www.stack.nl/~wiebe/mailinglist/)
****
