NeOS with disk

ZXNet echo conference «real.speccy»

From Roman Petrov To All 3 December 1998

Hello, All!

I was thinking about Denis Dmitriev's proposal about separating drivers, about
how to do this more efficiently. I came to the following conclusions (I won’t do all this
for now, I’ll wait for approval/criticism):
1. Forget about the "Amiga" way of naming disks and return to the native
one-letter one.
2. Enter 2 system variables, one of which will point to the table
of disks, and the second - the number of these same disks.
3. The device type driver contains 3 functions:
1) disk initialization
2) reading a block of sectors, the initial position is specified by the block number, and not
by track/sector (made for hard drives)
3) writing a block of sectors.
4. The disk table looks like this:
For each disk there are 4 bytes: 2 bytes - the address of the device driver,
2 bytes - the address of the current disk system driver for the disk.

Are you satisfied? 2 bytes are allocated for speed ;-) so that due to clicking
pages no extra cycles will appear, and my functions are not suitable for
reading pages while sitting on the page ;-)

PS: Finishing all this will take no more than half an hour. In the meantime, I continue to write
TR-DOS driver.

Bye, All! With the best wishes, Megus of Brainwave.

[Team ZX-Spectrum] [Team NeOS] [Team Rock'n'Roll forever]


From Denis Dmitriev To All 4 December 1998

Great, Roman!

One day 03 Dec 98 at 19:48:53 comrade Roman Petrov noticed comrade
All about , where I intervened:

RP> how to do this more efficiently. I came to the following conclusions (I won’t do all this
RP> for now, I’ll wait for approval/criticism):
RP> 1. Forget about the “Amiga” way of naming disks and return to the native
RP> single-letter one.
Well, tell me in more detail, or throw soap at the doc. Let's see what's what ;)

RP> 2. Enter 2 system variables, one of which will point to the
table
RP> of disks, and the second - the number of these same disks.

Logical disks, as I understand it? Or physical?

RP> 4. The disk table has the following form:
RP For each disk there are 4 bytes: 2 bytes - the address of the device driver,
RP> 2 bytes - the address of the driver of the current disk system for the disk.
^^^^^^^^
file system! don't be confused :)

The current disk system is installed after calling the setup function on
disk?
(only for FDD, of course).

RP> Are you satisfied? 2 bytes are allocated for speed ;-) so that because of the clicking
Completely. Simplicity is the key to reliability :)

RP> there were no extra cycles in the pages, and my functions are not
adapted
RP> for reading on the pages while sitting in the page ;-)
And this is already bad. It turns out that the program cannot loadfile into pages?
Or is this the function of transferring the downloaded content to designated
pages?

With best wishes, Denis Dmitriev.

... I’ll buy a moonshine still. intermediary phone number 02.

From Pavel Fedin To All 5 December 1998

RP> are adapted for reading on pages while sitting on the page ;-)

Where to store the driver and file system? In lower memory? IMHO it’s not good.
We need to think about the user’s applications. For them, the lower memory is the most valuable. In addition, IMHO programs from TR-DOS should be fairly easy
to re-adapt. There are valuable things in it (STS, ZX-Word, ZX-Tasm and others
that would be nice to move).

Bye, with a pack of viciousness, Pavel Fedin.

... UI125:/0EPJFAGAECCVNGLFHGiGDGFFFPAFfBCKBRRDQCBKHIEBCCKC


From Valerij Kozhevnikoff To All 8 December 1998

Hello, Roman!

03 Dec 98 at 19:49:33, Roman Petrov => All:

RP> 3. The device type driver contains 3 functions:
RP> 1) disk initialization
RP> 2) reading a block of sectors, the initial position is specified by the block number, and
not
RP> track/sector (made for hard drives)
RP> 3) recording a block of sectors.

Add here “switching to device” (it would be more correct to say “selecting a device”), you will get an analogue of the isdos block driver.
Maybe it makes sense to make them compatible?
Imagine if you could use ready-made *.blk from I wonder how much unnecessary
work would be possible without doing?

WBR, Jason.