Most of the MSX1 dont have the floppy disk included ( few have like the SVI , but it is an' MSX1.5' ).And the hard disk interfaces work only on MSX2.
I have seen that novatec had done an interface compatible with diskrom1 ( well If I well unserstood the asm code of it ) based on an i8255 PPI to interface to IDE ports .

So there is not fundamental reason to limit access of ide interface to msx2.

Of course with dos1 you dont have sub-dirs and it can be helpfull to organize the HD with it.( I may still exit a msx-dos1 version with sub dirs, version1.8?? )

I have tried to put on an MSX1 a 512K mapper and a sunrise interface and still it was not working. Does it mean Dos2 needs more than 128k RAM but also an MSX2 ?

There so many MSX1 useless (in my cupboard for instance , ) since they cannot store any data, that is one reason also why there is so few MSX1 disk games.( R-Type , .. ).

Have you tried switching the slots where the mapper and IDE is inserted?

DOS2 needs a 128K (or more) memory mapper available. Maybe also a functional Real-Time Clock but I'm not sure about that.

The only sure-fire way of using a harddisk on MSX1/DOS1 is using the HSH interface which allows one to use a HD without subdirectories. But it's a big loss of potential.

RTC wouldn't be a problem in MSX-DOS2 I think. Mapper configuration most likely is, and there might be other problems.

About DOS1, the Sunrise IDE's ROM is flashable... I think it should be possible to alter the IDE's DiskROM so that it loads the diskrom, but not automatically loads the DOS2 ROM afterwards.

~Grauw

The floppy disk access are through adress mapped IOs ( except brazilian's msx), I guess the SLOTSELECT is also used , so the address decoding of the floppy disk controller is a combination of the right slot select and address.

We could imagine that a rom, from slot , could take control over the FDC located in an other defined SLOT.

In that way the sunrise-ide could be flashed with modified diskrom , that access the HD in its own slot and the Floppy to an other slot.

I have just recompiled the source of Novatec interface with low level access to Hd changed so to make it compatible with the ATA-IDE interface from sunrise.

I hope i can then get dos1 + hardisk + msx1 running after flashing this bios !
( but if it works at first try I would be surprised )
I check if i have an old HD to make the test ...

Wow this sounds really promising, keep us posted, Leo!

Well Loaded the flash rom of the sunrise with the code and then the computer was stuck!
It semed there was no way also to load back the ide bios !!!
...
Finally I tried pressing CTRL at the boot and then I could boot normally in dos1 and reload the ide bios ... I was afraid to screw my ide interface !
...
I though that it was looking for a connected HD, so I attached mine ( no other left ) , and it did not work.
...
So, part of the rom might be working since it is sensitive to ctrl, but I should trace the rom execution with an emulator to debug the reason of this stuck.
But I have seen that there were a logo that should be printed but even that does not work.
...

Quote:

But I have seen that there were a logo that should be printed but even that does not work. ...

Maybe the logo is drawn in a MSX2 screenmode? Otherwise I wouldn't know ...

Yup, the logo users width 80 iirc.

In that case the MSX might not have locked up, but just stuck in a illegal screenmode. Easiest way to test this is pressing the CAPS LOCK. If the light goes on/off, the MSX has not crashed.

New results:

I have forgotten to translate some piece of routines from previous type of access ( NOVATEC: through a 8255 addressed in base address 30h ) into sunrise type ( address mapped io on base 7E00h ).

Now it works better, I have the logo "BEER - 202" plus an info message on the type of hard disk attached.

But this info is wrong !
- When I dont attach any Hd it reports anyway a HD with 27 head and 33sec/tracks.
then it goes into basic , with a nice 'disk basic version 1.3' version !
- When I attach my HD , it seems to init fine but then the info are wrong and the msx
stucks ( maybe it tries to boot from HD but with the wrong geometry info ).

I will try to modify the bios so now it boots on floppy by default instead of HD.
And also review the geometry info retreiving routine.

Note: I do the trials on my old VG8235 - MSX2, then if something works I'll try it on my sony HB75F - msx1. And the logo is not graphical , it is 40 characters text width. So it should worl also on an MSX1.

When flashing a custom-made BIOS, make sure your first 3 lines of initialization code are somewhat like this:

ld a,(#FBE5+8)
bit 0,a
ret z

This for example checks whether the space bar is pressed, and RETs if that is the case. If you do not utilize such code to make it possible to skip the execution of the contents of the FlashROM (by keeping space pressed), in case of an incorrect flash it could prove a very hard task to reflash your BIOS, because you won't be able to get 'past' the booting process with the cartridge inserted. In such a case you would need a means to disable the cartridge during the boot process and enable it later on, which can be done with a slotexpander for example, but in order to do that you first need to have one.

~Grauw

Well , I did something like this once , hiding the SLOTSECLECT signal at the boot then connect it back.