About VDP command operation code = 0

صفحة 2/6
1 | | 3 | 4 | 5 | 6

بواسطة flyguille

Prophet (3031)

صورة flyguille

06-06-2011, 21:26


If i cannot find a way to get another 720KB media, i will not be able to use msx disk. Formatting existing 1.44MB with a black stripe on in does not work for me....

The problem is format 720kb on PC? On the net I found "format a: /f:720" ,
which is said to work under vista, but not XP. I slighly remember that under XP the file explorer format hat a selection box where one could go 720k.

If the MSX drive head does wipe too much with format, then I would suspect that it also does this with every "save" operation. It may work for some time but I would prepare for disk wreck.

I know, for the early MS-DOS days the format ( i'm a programmer since 1989, being born in 1970 ) command.
the problem is that i get (on pc invalid format or track 0 damaged) for a 1.44MB disk formatted as 720KB cannot format
if, by contrast i try to format on msx, it's unreadable on pc at all.

the same DOES not apply if i use a true 720 disk, instead of a 1.44 with a black stripe to 'fool' the pc disk drive.

actually i can successully work only with those 720 thingy.....

I suspect the windows OS try to do some sort of format verification even with the /U switch of the format command.

No, that happens because before to format a real 1.44MB disk as 720K, you needs to demagnetize all data on the disk.

Because if you format wi the stripe that "fools" the drive, you are formating wider track that not necessary aligns with old tracks, so between wide tracks and wide tracks in between can be remnants of the thin old track, and interfere.

so, use a magnent to erase all the surface (spinning the magnet).

then try to format.

بواسطة PingPong

Prophet (4093)

صورة PingPong

06-06-2011, 22:04


No, that happens because before to format a real 1.44MB disk as 720K, you needs to demagnetize all data on the disk.

Because if you format wi the stripe that "fools" the drive, you are formating wider track that not necessary aligns with old tracks, so between wide tracks and wide tracks in between can be remnants of the thin old track, and interfere.

so, use a magnent to erase all the surface (spinning the magnet).

then try to format.

but some of my old 1.44 disk are unformatted, (for sure).
anyway thx for suggestion. now the hard part. need to find magnet....
Question

بواسطة hit9918

Prophet (2927)

صورة hit9918

06-06-2011, 22:39


but some of my old 1.44 disk are unformatted, (for sure).
anyway thx for suggestion. now the hard part. need to find magnet....
Question

And when you format them on PC with 720kb, after the format the PC really can't read it?

I slightly remember formating diskettes with an Emulator, maybe fmsx or RuMSX, which the OS could not read. Maybe this is a solution, maybe the emu just blindly enforces 720k mode.

And with a .dsk mounted in emulators B:, it was a .dsk-file to diskette copy station.

بواسطة PingPong

Prophet (4093)

صورة PingPong

06-06-2011, 22:43


but some of my old 1.44 disk are unformatted, (for sure).
anyway thx for suggestion. now the hard part. need to find magnet....
Question

And when you format them on PC with 720kb, after the format the PC really can't read it?

formatting from pc does try to start, but when processing 0 track i get the error above, format cannot complete.
formatting from msx appear to go fine, but pc see an unformatted media.

بواسطة PingPong

Prophet (4093)

صورة PingPong

06-06-2011, 22:48

Sad with disk to rom I get 128K
i can try to modify my test program of r18 that has this problem because it change r18.
I will add a vdp(47)=0 (in asm) before the write to r18 then the vdp(47)=208 after.
this will take me a bit, because i need to convert the code to a stream of relocatable bytes, then transform to a basic program with a loader, then convert to a wav file, then load with something via cassette interface.

بواسطة Hrothgar

Champion (479)

صورة Hrothgar

07-06-2011, 12:04

Can you elaborate on the use cases of this 'trick'?

You and Artrag mention this could be useful for scrolling games, but I thought the much feared corruption only occured when sprites are off. Can anyone confirm that? As hardware sprites seem a necessity for a smoothscroll game (games without sprites are poorly conceivable; and software sprites hardly possible when doing a full smoothscroll), I don't yet understand which breakthrough this could generate.

بواسطة sd_snatcher

Prophet (3642)

صورة sd_snatcher

07-06-2011, 13:47

I wonder: if the STOP command only pauses the blitter, is there a way to actually abort a command in progress?

بواسطة Hrothgar

Champion (479)

صورة Hrothgar

07-06-2011, 14:37

Not so much pauses the blitter, as the command engine I guess.
I imagine copy command then, internally, works line-by-line and updates the registers after every line and will only actually be halted after one line was finished? (how does this work with drawing commands?)
And to "actually" abort a command, whatever its uses, would require to manually zero all involved registers.

بواسطة ARTRAG

Enlighted (6923)

صورة ARTRAG

07-06-2011, 17:23

@Hrothgar
Corruption occurs when sprites are active, but even whe they are off. The ral conditions when this happens are unclear to me.
The fact is that now it occurs and I do not know why (try on real HW my supermario demo, you will see some black dots here and there that were almost invisible with other tiles thanks to the black background)

The idea behind my screen8/5 scrolling is to let as much as possible the vdp work in parallel with the cpu

The game needs to adjust the r#18 at least twice per frame to guarantee scrolling and screen split for scorebar.

The whole set of copies have to be sliced into 16 steps to guarantee border masking.

In this framework, it is required that the ISR halts the vdp while it is copying, set r18, resume the last operation where it was

بواسطة Hrothgar

Champion (479)

صورة Hrothgar

07-06-2011, 19:21

Thanks Artrag, I googled for it but http://map.grauw.nl/articles/vdp_guide.php#r18cmd is apparently incorrect then.

صفحة 2/6
1 | | 3 | 4 | 5 | 6