Nextor 2.1 Alpha 2 + new IDE driver released

(^______^) Great work!

Thank you Konamiman!

Real nice, this will be very very useful!

Won’t this be a problem when software uses GRAPH as an ALT key, e.g. GRAPH+S to save a file? Maybe consider using the SELECT key in stead.

Also re. fragmentation, hmm, I’m thinking it would be nice if someone made a FAT16 version of Improve…

Thanks Konamiman! Just tried it with my Sunrise CF, but it keeps hanging after showing the "Nextor.SYS version 2.1" + copyright message My TR-ST and Sony HB-F9P show the same behaviour so I guess it's something with the Sunrise cart/driver because I have this problem since the first release of nextor. Booting from a floppy works, just like nestor basic (which i used to partition my CF card). And in nestor basic I can see the files on the cf card.

Hurray! Nice to see Nextor 2.1 and "my" driver there

Only a fix in this announcement: I am not from Tecnobytes. Only the IDETB128.COM is. Probably my bad english skills gives this wrong idea. Write in a foreign language is always a challenge

And now is Nextor 2.1 flash time

Grauw,

The driver is a bit more stable than the previous one and can give some new debug information, it detects OK your CF?

Try with an empty CF card, the traditional "MSX-IDE" partition schema isn't so good for interoperability, and one of advantages of Nextor is to use a MSDOS-compatible disk partition schema.

You can try to use the old IDE driver to see if the problem is new or if it always happened.

**This post was heavily edited, I need to think before write**

Great! Nice work!
Question: which key do I have to use to release the DSK image?

Piter Punk,

I used 2.0.4 and alpha 2, both have the same problem. It boots but the prompt never appears. I also used the older versions of the Sunrise driver and they all had the same problem. I first thought it had something to do with my TR because of the extended internal memory. But today i also tested it on my Sony HB-F9P which has the same problem.
But i'm glad to see that Grauw has the same problem and it's not my cartridge

Thanks Guillian! I searched for it, but I couldn't find it that fast
Now I'm really happy with this feature! I missed it when I sold my Padial SD interface...

No difference, unfortunately.

Multi mente has a similar option but it doesn’t work on FAT16 either.

I use a pc tool for that at the moment, but iirc that's something an MSX could do perfectly as well...

If you want the old driver, you can find it here:

http://www.tecnobytes.com.br/interna.php?cod=118182

The NXTR23TB.DAT is the old Nextor + Sunrise IDE driver with some strings changed.

But your problem looks the same one that Sylvester reports. The driver can detect the CF card, can read and write there but some problem prevents "DOS" prompt to load. And, if this is the same that happens with Sylvester, changing to an old Nextor/IDE driver will not help. Did you try to create a FAT12 partition and, instead Nextor, put there only the plain old MSXDOS2.SYS to see if it boots? Or MSXDOS.SYS + COMMAND.COM?

To sort FAT files/directories I use and recommend fatsort. I use it for both MSX (FAT12,FAT16) and SMS (FAT32) and it works perfectly.

@KonamiMan
Congratulations for another awesome achievement!

@All community on MRC:
I'm doing my part and have set up a monthly contribution to KonamiMan as a thanks for his time and huge effort. What are you all waiting for? Go to PayPal and setup your contribution ASAP! :)

@Grauw
Maybe it's better if you generate an image of your partition and test if the problem is confirmed on openMSX. Then you can send this HDD image for Piter Punk to test.

At first I was using CODE/KANA, but for some reason the computer was hanging... the problem disappeared when I changed to using GRAPH instead! (O_o)

I guess I could add an option to EMUFILE to use a diferent key for disk change.

I was searching for a defrag tool able to work on SD cards and didn't figure out that I could do just as you say. Please kill me!

Sorry, it's my fault. From the first time you contacted me I was assuming that you were part of Tecnobytes staff! I have fixed my web site.

That's a great idea, and I would like to have this image as well!

If you use Windows a very nice tool to do this is ImDisk. If you have this tool installed a "Save disk contents as disk image" option is added to the menu that appears when you right click the drive assigned to the card in Windows explorer.

Congratulations! You found a bug!

It happens not only with Turbo-R, but with any MSX with an empty disk file. I usually use images from my physical CFs to do the tests and this never happened before. Now is time to see what is happening in OpenMSX and to test the bug in real hardware (filling a CF with "nothing" and trying to format it again). The last OpenMSX message:

warning: DI; HALT detected, which means a hang. You can just as well reset the MSX now...

The SD to CF adapters are supported by this new Kernel for Sunrise IDE/CF interface ?

Just to clarify: are you really emulating a Nextor kernel in slot 0-1? Or it is just a typo and you meant 1-0?

Also, this happens only in OpenMSX, right? Not in other emulators or real computers?

Not yet, but I really doubt that I do such thing on purpose. Anyway remember that the source code of FDISK is published in my web site, in case that someone wants to help me spot the nastyness.

@Grauw and @Sylvester: I really would like to have a disk image of the cards that cause your system to hang while loading NEXTOR.SYS. Could you please generate one for me?

@gdx: That's very strange. Looks like your kernel was not flashed properly. Anyone else has similar problems?

By the way, to disable a Nextor kernel press Q while booting if you have it in slot 1, or A for slot 2 (more information in Nextor's user manual).

It's the only method that allows to disable kernels selectively, so you can do things like flashing your IDE controller from a MegaFlashROM SD by disabling only the former. This also allows to use the SHIFT key for disabling only the MSX-DOS kernels.

I really have no idea. We would need to somehow create a dump of the complete ROM from the cartridge so that I can examine it.

I don't have any problems with CF cards. To be fair, in my tests (and with other beta testers) the new driver detects more CF cards than before. But the only tester with SD-CF had the same issues that you have. If you want to try a bit more, send a private message and I can send to you some development drivers.

Press ESC. We try very hard to detect the devices, one of the first tests is to ask if there is some device there (master or slave). Some adapters says that there is a device even when the slot is empty and some doesn't answer nothing when the slot is empty. In both cases we timeout probing for it.

You have to press a key for the specific slot where the Nextor kernel is:

Q,W,E,R -> 1.0, 1.1, 1.2, 1.3
A,S,D,F -> 2.0, 2.1, 2.2, 2.3
Z,X,C,V -> 3.0, 3.1, 3.2, 3.3

If you are in primary slot X use the key for X.0. If you have an IDE-Mapper, use the key for slot X.3

Easy: your NEXTOR.SYS is corrupted. Just copy a good one to your card overwriting the old one, and it should work. It does at least on my blueMSX.

I see a lot of strange hidden files in your card, like _BACKU~1.DOC or ~1.TRA. Are you using this card in another computer as well?

Could you please send me an image file of the whole card that is freezing FDISK?

As for the other problems they seem to be related to the driver, so let's see if Piter Punk sama can do someghint about it.

I tried it again, flashed the sunrise cf, booted to basic, removed all partitions, created 2 new partitions of 32MB.
Then I mounted the CF-CARD in my Mac to copy nestor.sys and command2.com on it. Then booted the MSX with the card, and it hangs again after the nextor.sys message. So maybe it's my mac, downloaded nextor.sys on my mac, packed it with lha, copied that file to my cf card with pmext and msxdos2.sys. Flased the Sunrise CF again with sunrise ROM, then use pmext to unpack the nextor.sys, flased the Sunrise CF again, reboot and it hangs again
When i have some more time, will try to create an hd image of 64MB with openMSX that works with nextor and then copy the dsk image to the CF-Card to see if that works better.

@Sylvester

Use files from Nextor-2.1-alpha2.zip.
Konamiman's site needs to be simplified.

Please try this method :

; Routine without call to Bios (without EI)

LF	equ	0ah
CR	equ	0dh
;CHPUT	equ	000A2H		; Print a character
CSRX	equ	0F3DDH		; X cursor value
CSRY	equ	0F3DCH		; Y cursor value
LINLEN	equ	0F3B0H		; Screen width value
NEWKEY	equ	0FBE5H		; First line of matrice keyboard
SCRMOD	equ	0FCAFH		; Atual screen mode

	org	04000h

	db	041h,042h		
	dw	progStart
	dw	0,0,0,0,0,0

progStart:
	push	af
	push	bc
	push	hl

	ld	a,(NEWKEY+8)
	bit	2,a		; INS Key
	jp	nz,Exit

	ld	a,(0c000h)
	cp	c
	jp	z,Exit		; Jump if second call to progStart
		
	ld	a,c
	ld	(0c000h),a

	ld	hl,Question
	call	Print

; Print Slot number

	ld	a,(0c000h)
	and	3
	add	030h
	call	CHPUT		; Print primary slot number
	ld	a,(0c000h)
	rla
	jr	nc,primslt	; Jump if  slot

	ld	a,'-'
	call	CHPUT		; Print"-"

	ld	a,(0c000h)
	and	00ch
	rra
	rra
	add	030h
	call	CHPUT		; Print secondary slot number

primslt:	
	ld	a,'?'
	call	CHPUT

Again:
	ld	c,5		; C = Keyboard row 5 value
	call	ScrutRow
	bit	6,a		; Test Y Key
	jr	nz,Pass_Y_Key
Y_Pressed:
	call	ScrutRow
	bit	6,a		; Test Y Key
	ld	hl,Disabled
	jr	nz,OK
	jr	Y_Pressed

Pass_Y_Key:
	ld	c,4		; C = Keyboard row 4 value
	call	ScrutRow
	bit	3,a		; Test N Key
	jr	nz,Pass_N_Key
N_Pressed:
	call	ScrutRow
	bit	3,a		; Test N Key
	ld	hl,Enabled
	jr	nz,OK
	jr	N_Pressed
Pass_N_Key:
	jr	Again

OK:
	call	Print

Exit:
	pop	hl
	pop	bc
	pop	af
	ret
	
Print:
	ld	a,(hl)
	cp	'$'
	ret	z
	call	CHPUT
	inc	hl		; Next character address
	jr	Print

ScrutRow:
	in	a,(0aah)
	and	0f0h
	or	c
	out	(0aah),a
	nop
	in	a,(0a9h)	; A = Row 5 value
	ret

; Print a charater routine
; Input: A = Character ASCII code
; Note: Take partially into account the parameters WIDTH instruction

CHPUT:
	push	af
	cp	LF
	jr	nz,NoLF
	ld	a,(CSRY)
	inc	a
	ld	(CSRY),a	; Move cursor to next line
	pop	af
	ret
NoLF:
	ld	a,(hl)
	cp	CR
	jr	nz,NoCR
	ld	a,1
	ld	(CSRX),a	; Move cursor to beginning of the line
	pop	af
	ret
NoCR:
	push	de
	push	hl
	ld	hl,0
	ld	a,(SCRMOD)	; A = Screen mode value
	or	a
	ld	de,028h		; Width max = 40
	jr	z,Scr0		; Jump if SCREEN0
	ld	h,018h
	ld	de,020h		; Width max = 32
Scr0:
	ld	a,(CSRY)
AddLines:
	dec	a
	or	a
	jr	z,NoAddLines
	add	hl,de		; Address for next line
	jr	AddLines
NoAddLines:
	ld	a,(LINLEN)	; A = width of a line
	ld	b,a
	ld	a,e
	sub	b		; A = Width max - width of a line
	rra			; A = A/2

	ld	b,a
	ld	a,(CSRX)
	add	b		; A = A + (CSRX)
	ld	e,a
	add	hl,de		; HL = actual address of character

	ld	a,h
	rlca
	rlca
	rlca
	and	3		; Keep 3 MSB of HL
	out	(099h),a
	ld	a,80h+14
	out	(099h),a

	ld	a,l
	out	(099h),a

	ld	a,h
	and	01fh		; Keep other bits of H
	or	040h
	out	(099h),a
;	ex	(sp),hl
;	ex	(sp),hl
	pop	hl
	pop	de

	ld	a,(CSRX)
	inc	a
	ld	(CSRX),a	; X cursor moves to left

	pop	af
	out	(098h),a	; Shows character
	ret
	
Question:
	db	CR,LF,"Ignore interface in slot $"
Disabled:
	db	CR,LF,"Interface disabled.$"
Enabled:
	db	CR,LF,"Interface enabled.$"
	
progEnd:
	ds	08000h-progEnd

This program makes a Rom of 16kB for you test my method. Insert the Rom in several Slots.
Using:
- Press INS key during startup.
- Always keep INS pressed when answering questions until Beep.

I chose the INS key as Sunrise but you can define another. I hope you like it. It's very easy to use.

@gdx, the "disable key" handling is done by Nextor kernel itself.

@Sylvester, I already knew that DEV_INFO had some bugs (is listed in "Known Bugs") but until now, all bugs are cosmetic, mostly "empty strings" when calling DEVINFO command. With the machine hang problems I'll need to look deeper at this code. I don't know if Sunrise code would help with this specific bug, but for sure it would help in a lot of other problems.

@mesiasmsx: Holding "2" while booting Nextor means "I want to boot on MSX-DOS1 with R800 enabled" as described in the documentation. That's why the FM sounds weird (due to R800)
Hold GRAPH while booting and when the CAPS light switchs on, press "2" to boot from disk 2.

Hi konamiman & Piter Punk,

I made a disk image of my CF card, I’ve sent you a DM with a download link.

Hi everyone, I've waited long enough and wanted to try experimenting with the new Disk Emulation features of Nextor.(Thanks Konamiman! And thanks Manuel for making a simple recovery rom for the MFRSCC+SD!) I've managed to get everything installed correctly and I've figured out how to load a disk into this new virtual memory loader which creates a Next_Dsk.dat file and I see how I can get back to the Multimente launcher by holding down '0', but I've run into some questions of how to best use this new feature and I apologize if these questions have been asked earlier.

My first question would be if this new Disk Emulation option supports loading multiple disks, and if so, what is the best way of going about this? I've seen a few posts above which state to use the GRAPH key and when I see the CAPS LOCK LED light turn on, I can hit another number key to switch to that disk. I tried this with the Melancholia SD-Snatcher this evening (which I finally got around to patching and I can't wait to start playing it!) but it doesn't seem to call any other disks such as when I used GRAPH + a Number Key, but I can at least hit GRAPH + 1 to go back to the first disk and resume the game.

I guess my problem is that I don't know how to add multiple disks, or if perhaps if I'm only supposed to be using Disk Emulation for say the User Disk only (for saving games) or for Disk Games that save data directly to the first Disk. Is there a way to add multiple disk files via Multimente? I tried selecting all using [SELECT] and/or using the TAB key to mark each disk file (out of four, including the userdisk). Or if that is not possible, is there a way to add this all to the Disk Emulation option through MSX-Dos 2 with a .bat file perhaps?

Or maybe I need to load one disk file that has been concatenated (SD Snatcher 1 + 2 +3 + 4(userdisk)) to create just one disk?

Last question would be that if I wanted to save this virtual disk setup (and saved data) would I just need to make a backup of the Next_Dsk.dat file and perhaps rename it to something like "SDSnatch.dat" until I'm ready to throw it back onto the SD card and renaming it back to Next_Dsk.dat when I want to use it again?

Thanks very much for your help!

-Thomas

If by "empty" you mean "full of zeros", then it is not hanging in blueMSX. It should not, as there is no parition table in place and therefore Nextor assumes that it is an unpartitioned device.

You got a nice response from @Guillian for this one already. Anyway, I would like to remind that there is a dedicated forum post for informing about issues when emulating games, please use it instead of posting such problems here.

This image boots perfectly in blueMSX. But there's something strange: the image file size is 15MB, but the contained partition is just 720K! How have you partitioned it?

If you know beforehand which disk you want to boot with, create the .DAT file as follows: EMUFILE -b (disk index) *.dsk. This way you will boot directly in the desired disk.

Also note that you don't press GRAPH+number to change the disk. You press only GRAPH. Then when CAPS lits, you release it and press the desired number.

And finally, GRAPH is not really needed to switch disks when you are in the game. Just press the disk number when the game is going to access the disk and that's it. GRAPH is only needed at boot time to disambiguate between a disk change and a special Nextor boot key (e.g. 2 = switch to R800)

In _FDISK I specified a partition size of 720K, to ensure it creates a normal FAT12 partition. Anyway, looks like the problem booting on the real Sunrise CF-IDE can only be reproduced on a real machine...

In OpenMSX it gives errors. As I said in another post, looks like OpenMSX uses data from FAT first sector to fill up the IDENTIFY request. After I fill some specific data on FAT first sector it works OK, even with this partition table broked (without the size) and FAT first sector totally invalid. But without this data it returns some kind of garbage. I will fix the DEV_INFO to comply with specification and return the hex values when it finds garbage in the IDENTIFY answer. Probably this can help some real hardware, too.

Don't worry, I have one of these as well. A couple of persons have reported me that their computers hang when saving a file from the AKID text editor, but in both cases the problem disappeared spontaneously. No matter what I try I can't reproduce it.

Ok, if I press ESC just after the CF in master slot is recognized, it avoids the scanning of the slave slot. We must be very quick to press the key at the right time.
If I insert the CF in the slave slot the scan does not take much time! Strange the scan is fast as the kernel from Sunrise for only the slave slot.

Another problem: Sometimes Nextor does not boot and once it have booted on the drive B:. (I created just one partition)

The length of identifier name can it be a problem or not?

It's strange this is only for the slave slot!
The kernel from Sunrise doesn't take longtime for any slot.

Default.

Thank you. The identifier name of unrecognized* SD to CF adapter is 40 characters. Are there invalid characters? (see picture)

(*) unrecognized by Nextor's FDISK but recognized by IDEFDISK from Sunrise.

When I try latest stable 2.04 and alpha 2.1 in bleMSX as Sunrise IDE for Daewoo CPC-300/CPC-400 - it doesn't work. System never see a CF, boots in Nextor Basic always.
Same happens with the real hardware in CPC-300.

You can install with the original Konamito driver, or try to make your own, or improve the one that exists now. Without any driver the CF/IDE couldn't control any CF/IDE device. One of the improvements made by Konamito in Nextor is to have the driver clearly separated than the OS itself, so it is easily replaced.

If you want to try Nextor but without some features, you can flash it in another cartridge (there is a standalone Nextor in Konamito's page) and you will have Nextor support, but using the CF/IDE devices in "Legacy" mode (no FAT16 support, no standard partition table, etc), like we use the FDD devices now.

If you connect any cartridge with Nextor (MFRSCC+SD or SD/Mapper) with interfaces that isn't with Nextor, you'll get the same effect, the Nextor will have access to all disks connected, but some in "native" mode (those connected in interfaces with Nextor support) and some in "legacy" mode.

Hi Piter Punk, you did a great job with your driver for Sunrise ATA-IDE cartridge with Nextor. Could you also do this for the Padial SD interface please? That would be great!

Hi guys,

Now most of my MSX stuff are packed. Sadly, to really test the IDE driver, I need to do most of the tests using real hardware, as many timing and corruption issues only happens when testing in the real stuff. Example: for some reason, the Sunrise hardware is much less tolerant to those timing issues than the clones from CIEL and Tecnobytes (didn't try yet with other clones).

I don't know (yet) how the storage device can be tested, formatted and can't be read after that. I really want to continue this development, and hope to do that soon.

An "workaround" that helped some people is to power on first the IDE devices and adapter, wait a few seconds, and later the MSX.

You can try it.

Are there a progress on the original CF Sunrise?

Maybe because Sunrise hardware uses TTL chips (MSXPro's Sunrise IDE uses 74HCT* and 74ALS* chips) and GAL PLDs, while CIEL and Tecnobytes devices uses CPLD chips, which are way faster.

Is the Nextor project still being developed?

After MicroSoft released (let someone else publish) source code base of MS-DOS 1.0 and MS-DOS 2.0, is it possible to have source code base of MSX-DOS 2.30 which is source baseline of beloved NEXTOR ??

Noticed the same problem here. That happens when to edited file is in a subfolder (A:\C\MAIN.C in my case). Only the 500 first bytes of the file are saved when I perform the "E" (save & end) command, and then I'm getting either a "Wrong version of COMMAND2.COM" (on GR8Net) message or an "Invalid file handle" message (on MFRSCC+SD).

Quite weird. And the problem does not disappear hear. I tried to copy my MAIN.C file to another partition, try to copy it to main2.c... same result so far. Will switch to TED for now...