f18a 60 rows, 80 columns mode showing Stevie

[+retroclouds]

Here's a thing I've been playing with, and thought about putting it in a separate thread.

With the help of @Asmusr I tweaked the f18a emulation in my local clone of js99er, so that it can now run in 60 rows, 80 columns mode.

The idea was to see how I need to setup VDP memory to get stevie running in 60 rows, 80 columns mode and do refactoring for the different video modes (24x80, 30x80, 60x80, ...)

This is still a work-in-progress and will never be officially released, but it does prove that it's doable.

As a starting point I took the stevie build I made for classic99 where I use a character cursor instead of a sprite cursor.

(Sprite cursor is possible, but will only get halfway the screen, due to 1-byte for Y in the sprite attribute table, so of no real use here)

 

Anyway, the build for classic99 was also a hack, because I refresh the frame buffer each time the cursor blinks.

That clearly shows here, because I'm dumping quite a bit of data to the VDP and it gets a bit slow.

So, next thing to do for me is to get that cursor routine a bit smarter.

 

You never know, maybe the f18a mk2 will support a 60x80 mode.

I'm still wondering if it'd be doable on the f18a mk1 with a firmware update, but I think Matthew said it won't work because of VHDL already being chuck-full.

16K of VDP memory is enough though, as my demo shows.

 

Enjoy the demo (VDP corruption at the end when turning on the ruler., but that's part of the fun. I still need to fix some things in stevie).

 

 

 

 

 

 

 

 

 

 

Edited January 23 by retroclouds

[+TheBF]

Now you are showing off.  

Very nice.

[HOME AUTOMATION]

Can it core a apple?🍎

 

[+TheBF]

You wanna go to da moon?

[+retroclouds]

7 hours ago, TheBF said:

Now you are showing off.  

Very nice.

 

Thanks, I’m kinda hoping we’ll get such mode in real hardware, perhaps in the mk2 😀

[matthew180]

On 1/23/2023 at 3:02 PM, retroclouds said:

You never know, maybe the f18a mk2 will support a 60x80 mode.

 

Depends.  This means full scan line generation in 32us vs 64us.  With the external SRAM in the MK2 I lose the FPGAs dual-port access to VRAM, so there is less time on the MK2 than on the original F18A.  This is one of the main reasons for having to rework the HDL.  If it were *just* a text display then that would not be a problem.  However, much of the F18A's capabilities are usable in despite the "mode", i.e. things like sprites, the bitmap layer, the 2nd tile layer, etc., all add time to generate a scan line.