R800 assembly is Z80 assembly plus a handful of extra instructions.
OK. Now I only need
- Turbo-R
- reason to get interested in Turbo-R programming
Well a good reason would be for BASIC programmers to make games for the turbo-R.
the problem of aiming for turboR is that you also aims for a little public % that actually has a TR.
@pitpan, Dont get me wrong with this topic, I like the idea of the Dev contest. Again, i'm not blaming anyone, but to me its so surprising that, after 20 years, the focus is on the MSX 1.
@mas, Indeedio correctomundo. I'm indeed from Holland. Within this topic I forgot the demographic release of the msx 2, something yzi pointed out correctly.
@yzi I have two 8250 left, in case you want some more rolls royce's (for the same price? )
Again, dont get me wrong in this topic, it's not to bash not to blame anyone.
I was at the fair @ Zandvoort when the Swiss Demo was released and it was awesome (at that time) and if i'm correct Terra Nova was then also released.
But after that it seems like *pooof* almost nothing. Again, I like all the initiatives, especially Space manbow 2 and Goonies 2 and hardware like the Megaflash rom, midipac, IDE interface etc..
(although I never understood why some hardware isn't migrated into one card, but that has to do with powermanagement, copyrights, demands etc.)
@fly, I know there is a smal %-tage of TR's here (is this true?, reading the TR serialnr discussion there are more then a few. Perhaps there should be a rule, one TR per family )
Reading this site and watching youtube I think there are relative more demo's / games / experiments for the moonsound and the 9990 card then for the 2+ / TR. Again, no bash or blaming intended, just a friendly observation.
I'm waiting for someone to make a killer native R800 Turbo-R demo. It would be interesting to see if it could have kicked the Amiga 1200's butt!
No chance... Amiga's are entirely different beasts, with 32 bit CPU, very powerful video & sound hardware, many, many expansion options etc. IMO a big difference is that in MSX machines almost everything needs the CPU (Z80 / R800) to do actual work. Where Amiga's have some specialized hardware 'co-processors', like: fill memory buffer with sound data, set a few parameters & hardware does actual sound output while CPU moves onto other tasks.
MSX has some of such hardware too, but is much more CPU-intensive design. Meaning: Amiga has way more potential to do impressive stuff if programmed well. Oh and btw Amiga 1200 is a later generation with improved video hardware, faster CPU (68020 IIRC) etc.
As much as I like MSX in general (including Turbo-R), in my mind there is little doubt that Amiga >> Turbo-R if you talk about what is possible on these systems.
the problem of aiming for turboR is that you also aims for a little public % that actually has a TR.
Don't get me wrong, but this URL seems to prove the opposite: There's a high percentage of *active* users that do own a Turbo-R. :)
About pcm audio playback, the SCC and SCC+ have a big potential, even higher than the pcm on TR
Pcm in TR has a single byte as buffer, thus the playback has to be cycle accurate making pcm audio in TR not useful for sfx during actual gameplay
SCC and SCC+ allows pcm on the ISR spending very low cpu time, this thanks to the wave buffers that provide 4x32 = 128 bytes per frame or 5x32= 160 bytes per frame for independent playback
At 60Hz, SCC+ can output 8 bit samples at 160*60=9600Hz, that is good for bass line added to psg music, for spoken messages and for many sfx
If you stay with SCC, the playback frequency is 7680Hz, there is some static noise due to and HW bug of this chip, but except for music (where the Hw noise can be annoying), it can still be used for vocal messages and sfx (it is perfect for explosions)
Due to the fact that the data cannot be re-sampled on fly, all this stuff is best suited on msx2, where you can set 60Hz as ISR frequency.
It would be very sad to publish two roms one with 60Hz pcm data and one with 50Hz audio data
PS
About me, I am considering msx2 seriously, but the scarce free time (and two children, one of them newborn) does not leave a lot of room for coding at all.
With Huey we started the port of MOAM to msx2, the game is screen4, it has no special visual effects, except tons of awesome animated graphic and great scc music and pcm sfx by JohnHassinks (who created also the settings and the Gothic plot).
The port stopped on the sprite engine, maybe I wanted too much from the vdp (willing to implement sprite rotation and to use and access to 64K of vram for sprite animations). We will see if cutting down something we could restart the project.
Another msx2 project of ours is Terror Revelation. It is a top down shooter (like Alien Syndrome) but with light/shadows, field of view and walls changing perspective while the screen is being scrolled.
We stopped at the AI implementation of the enemies. Pathfiding can be a good addition for the project.
I think A* pathfinding can become real time: the key is in the definition of the "nodes".
Not all the squares of the map have to be nodes, but only those occurring at turns and crossings
In this way, a map of say 64x64 tiles, can be reduced to not more than 20, 25 nodes according to the complexity of the level
Even the crappiest C implementation of A* can work on these figures on fly. More about pathfinding soon.
We had also a TR project, the 3d Maze project implements a 3d raycating on r800.
If I find the time for the scalers it could become a real game
Yeah. Sprites on MSX2 are the main bottleneck (rotation/mirroring to avoid 9th sprite issue). If only they did something similar as on the Sega Master System VDP (mode 4). Sigh.
Did anyone consider programming on the SMS? I bought myself a SMS console + everdrive and have done some code tests (porting MSX1 megarom to mode 4) and coding for SMS feels just the same as for MSX
What is so special about the Swiss Demo and Zone Terra that makes it turbo-R specific? What parts would not work on real MSX2-es? (except the MSX2+ images and PCM samples)
I remember laughing my ass of seeing ANMA's turbo-R demo running on a normal MSX2 with only missing the PCM sample.