Microhertz
Monday, 13th February 2006
They say that pictures are what make a journal interesting... but I don't really want to show the pretty pictures I have, as they'd spoil the surprise. Of what, you might ask? Well, I've been developing a TI-83/83+ scene demo, entitled "Microhertz" (for no particular reason). You can download it (and some screenshots, if you really must!) here.
Things have been very busy - a new Latenite beta, riddled with all new exciting "features" (*cough*), has been released... I'm slowly fixing them as and when I can 
It's a bit rubbish for debugging - you can start, pause, and stop a debugging session, but that's about it. Hopefully it won't be too long until I get around to actually sorting out all the debugging gubbins, but I'd rather get the main IDE and the primitive debugging more stable first.
One other TI project I'm working on is updating QuadPlayer. I have a 1.1 version which can play songs from the archive as well as RAM, has a loop function and other little fixes and tweaks. To complement it, I've written a simpler script for creating your own songs, a better Windows-based player for testing (which simulates QuadPlayer's internal sound generation system and has accurate timing) and (best of all) am working on a program that takes PSG VGM files (BBC Micro, Sega Master System &c) and converts them to QuadPlayer songs. So far, any songs that do "exciting" things - vibrato, for example - end up sounding horrible in high octaves (I'm having to limit 10-bit to 8-bit periods), but I have some very decent sounding QuadPlayer files based on Sega VGMs. Some songs use lots of short, fast notes - which sounds great on dedicated hardware, but when you have a delay as QuadPlayer stops outputting a square wave to fetch the next note, sounds crap.
YM2413 (OPLL) Emulation
Monday, 30th January 2006
As (yet another) side project to all this Z80 work, I've also decided to have a stab at the Japanese Master System's FM chip. It's a YM2413, and the documentation on it is fairly tricky to get to grips with - not only thanks to it being in typical Japanese-manual English.
So far, I have this - the VGM player is a bit buggy and extremely primitive (unfinished), but should be enough to demonstrate the current state of the OPLL. You'll need an FM VGM to try it with - the Space Harrier demo from the BIOS sounds pretty good.
If anybody has had any experience with the YM2413, I'd be interested to hear if you had any helpful tips...
In other news, I've been adding multipage program support to Brass, and Latenite has had a lot of debugger integration work done on it. It's still a long slog before it's in a presentable state, sadly.
Debugging...
Thursday, 12th January 2006
CoBB has added an non-interactive mode to his emulator. I'd never thought about using stdin/stdout to pass information between executables; but it works well (once I've ironed out the bugs at my end, that is!) All I need to do is pass it instructions via stdin then read back the results (such as a screen dump) through stdout. This can then be, hopefully, used to pass instructions back to Latenite ("Highlight line 43 on file source.asm", for example, to show where the PC is).
TI Z80 Development Gets Cool
Tuesday, 10th January 2006
This excites me greatly - integrating a debugging emulator into Latenite is very, very awesome. CoBB's emulator is incredibly good as a standalone, being able to step through your code instruction-by-instruction with breakpoints and a watch window or two is just amazing!
I spent most of the last few days working on Brass - adding for loops to assemble blocks of code multiple times and a few file operations for people not happy with the behaviour of .incbin. I also remembered to release an XML help file for Latenite.
The Marble Madness engine project has been plodding along in the background - I've written a fast masked, clipped, 16x16 sprite routine and have started tackling the big problem of mapping the ball's 3D coordinates to the screen (easy) and then to a heightmap (hard) for Physics.
Have a marble rolling off a ledge then bouncing off the right screen boundary. 
Scrolling Marbles, Fire Track ROM
Friday, 6th January 2006
Fire Track
Prompted by a PM from evolutional, I rebuilt a working-ish version of Fire Track.
Keys should be fairly self-explanatory; hold down button 1 and 2 as the ROM loads to get access to a cheat screen. Of course, you'll need an emulator; Emukon is a highly advanced, but suprisingly fast emulator (fast in that it lets me run GG ROMs at full speed on an old Pentium PC).
I will be uploading all the source and tools in a zip at some point so you can experiment with some of the other features if need be.
Marble Madness
I started work on a 4-way smooth scrolling tilemapper last night for use with the Marble Madness project (for the main view) and have come up with a fairly simple, but sufficiently fast mapper.
The GIF is a little wasted as the level is only 128 pixels wide, and on a 96 pixel wide display this doesn't give much room for showing off the smooth x-scrolling. Having never written a mapper like this (only basic, slow horizontal scrolling and vertical scrolling, never together) it was quite an experience, but not as hard as I'd thought.
I have cobbled together a basic editor that allows me to edit the map, and that takes a folderfull of GIF files and exports a binary sprite resource file. Hooray RAD with C#!
It looks like I'm going to need a lot of sprites... 
As for this crashing and burning - Fire Track failed because of my absolute lack of music skill; also, the game was too hard to be any fun.
Marble Madness is just a set of ideas; I'm not comitted to it (hence no "official" announcement to any TI people, kv83!)