Nestor,

> MegaSCSI uses actually two systems at the same time to manage
> partitions:

        Yes, I know this. But using MEGASCSI.TBL is fudeba. :)

> >       So, SCSI users allways have to know the IDs of their SCSI devices?
> On the case of MegaSCSI, there is no need for this.

        Wait, wait... Forget about PS for a moment. I know that using
ESET and PS you can forget about SCSI IDs after the initial setup.
        I'm asking all this things 'cause I'm writing MegaSCSI/Novaxis
drivers for UZIX. And I need to know about SCSI IDs handling.
        In UZIX you don't have the 8 drives limitation. For example, take
the Sunrise IDE interface (that already have a driver in my UZIX): If I
had
two IDEs connected to my MSX, and each IDE had two HDs connected to it,
AND if each HD (total: 4) had 31 partitions, I would have 31*4=124
partitions to manage.
        How do I handle this on UZIX? I have 124 devices in /dev:
        /dev/idea0              (1st IDE, master device, partition 0)
        /dev/idea1              (1st IDE, master device, partition 1)
        ...
        /dev/idea30             (1st IDE, master device, partition 30)
        /dev/ideb0              (1st IDE, slave device, partition 0)
        /dev/ideb1              (1st IDE, slave device, partition 1)
        ...
        /dev/ideb30             (1st IDE, master device, partition 30)
        /dev/idec0              (2nd IDE, master device, partition 0)
        /dev/idec1              (2nd IDE, master device, partition 1)
        ...
        /dev/ideb30             (2nd IDE, master device, partition 30)
        /dev/idec0              (2nd IDE, slave device, partition 0)
        ...
        /dev/idec30             (2nd IDE, slave device, partition 30)

        Ok, now I can access each partition of each HD of each IDE I
have on my MSX. Cool, huh?
        Stupid usage example: I want to see the directory of the DOS
partition, FAT 16 bits, in my 2nd HD (slave device of first IDE),
partition 27. I would type

        dosdir /dev/ideb27

        Take the same example for MegaSCSI: each SCSI interface can
handle seven devices (six (right?), since one ID is used by the
interface
itself), and suppose I have two MegaSCSIs on my MSX:

        /dev/scsia0             (first MegaSCSI, device with ID 0, partition 0)
        ...
        /dev/scsia255           (first MegaSCSI, device with ID 0, partition 255)
        /dev/scsib0             (first MegaSCSI, device with ID 1, partition 0)
        ...
        /dev/scsib255           (first MegaSCSI, device with ID 0, partition 255)
        ...
        [and so on, until /dev/scsio255]

        So, if my SCSI ZIP drive, with ID 5, is connected to my MegaSCSI, the
corresponding entry of it is /dev/scsif### (where ### vary from 0 to
255). So,
I need to know that my ZIP drive has ID 5 or I can't access it because
I don't know which entry to use in /dev.
        I hope I made de problem clear. If not, I can explain it again.
        So, master-of-MegaSCSI-NestorPrograms ;), am I right about all this
things or there is an easier way to handle all this stuff?


Adriano Camargo Rodrigues da Cunha          ([EMAIL PROTECTED])
http://www.adrpage.cjb.net                  MSX: more fun per less MHz

Aibohphobia n.  -  The fear of palindromes


_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com

--
For info, see http://www.stack.nl/~wynke/MSX/listinfo.html

Reply via email to