New release of EightyOne available

[sirmorris]

Did it work before? I don't think anyone's been near the TZX stuff. Could you please try with an earlier version?

[Moggy]

I have done on 1.2 and it's the same that's why I tried this version just for a giggle.
It's no biggy from my point of view just thought I'd point it out.

[GCHarder]

LOADing a binary file will sometimes disable the ZXPand ROM.

Regards;

Greg

[sirmorris]

There is a feature in ZXpand which disables the overlay if the loaded file is read-only. I'm thinking I should remove this as it tends to cause nothing but FUD

Could you please check the flags on the data file and see if it's read only? If not then there's an issue.

*edit - data file 2 appears to have its write protect flag set. That'll do it!

[1024MAK]

It will make everyone's life easier if when submitting a bug report, you state if the problem occurs on versions earlier than 1.2.

Mark

[Moggy]

It will make everyone's life easier if when submitting a bug report, you state if the problem occurs on versions earlier than 1.2.

Mark

Unfortunately I can't get EO less than v1.00 to run on win7 64 bit That's why I mentioned nothing less than v1.2.

[1024MAK]

Old PCs running Windows 95/98/2000/XP * etc aren't just door stops...

They can find a new life pretending to be something useful, like a ZX81...

Mark

* PS yeah, I know I left other versions out, but I don't want to be here all day...!

[olofsen]

A few cents, without a solution however, on AY (the implementation from Fuse):

In EightyOne's source there are the lines

Code: Select all

machine.tperscanline=207;
machine.tperframe=312*207;

312 should quite possibly be 310 (viewtopic.php?t=1690#p17413) and 207 could in rare cases be 208 (viewtopic.php?t=558#p12105), and tperframe is different in FAST mode, and it could very well be somewhat different with an M/C program...

Note that in EightyOne, the Hsync counter is reset by Vsync rather than INT ACK; possibly this line in zx81.cpp

Code: Select all

machine.tperscanline=HWidthCounter;

handles an apparent inconsistency such as encountered with ASZMIC.

Also, in z80.c there is a line in "int z80_nmi( int ts )":

Code: Select all

waitstates=(ts/2)-machine.tperscanline;

which suggests that the line in main_.cpp

Code: Select all

j-= machine.do_scanline(BuildLine);

could in principle always subtract tperscanline from j during NMIs. But on the other hand there is a "borrow" variable so that tperframe may not always be reached.

Anyway, code such as the following line in sound.c

Code: Select all

newpos=(frametstates*sound_framesiz)/machine.tperframe;

assumes that every 20 ms that "AnimTimer1Timer" is called in main_.cpp, and therefore "sound_frame();", also tperframe tstates have occurred. But I think the actual number (maximum of frametstates) is currently somewhat variable, leading to a form of jitter with for example the beeper sound...

[Fruitcake]

In EightyOne's source there are the lines

Code: Select all

machine.tperscanline=207;
machine.tperframe=312*207;

312 should quite possibly be 310

This change was made in version 1.3 and does seem to improve the sound of the Beeper output.

[olofsen]

Indeed. However, in file "zx81config.cpp" there is still a 312?