nesdev.com

Hey, it runs Slalom correctly, without the glitchy "weave" effect seen in many other emulators.
But it fails to run Dragon Quest 1 or 2, or Door Door correctly, since those are the ONLY games which use the APU Frame IRQ feature.

_________________
Here come the fortune cookies! Here come the fortune cookies! They're wearing paper hats!

Strange, I tried Dragon Quest 1 and Door Door (japan version and then run on fnes NTSC version) and they work well here. What is your problem?

Without emulating APU frame IRQ, Dragon Quest 1 and 2 play no music, and also crash after a battle, possibly because something is waiting for the music to finish playing.
I think Door Door just doesn't play music without Frame IRQ. I don't know offhand of any crashes for that game.

I only mentioned those games because they are the only games I'm familiar with that use APU Frame IRQ.

_________________
Here come the fortune cookies! Here come the fortune cookies! They're wearing paper hats!

Ok, I understand. I'll begin immediately to work on it, many thanks for testing

I was only joking :)
You never know. FCEUX comes from FCEU, which comes from Bero's FCE, which was what I described previously. But for every FCE, there are 100+ LochNES'es.

I personally find it amazing how many people attempt NES emulators. There's millions of hardware edge cases, hundreds of custom mapper hardware, dozens of controllers, custom sound packs for some games, years and years of work in a scene that is 95% dominated by Nestopia already, which has every feature under the sun. And yet every time I visit here there's a new project starting out.

That someone can walk in knowing their project is 99% doomed to obscurity is really awe inspiring. It really shows you're doing it for the experience and not for the fame, which I really respect.

I certainly wish your project the best, and hope you'll come up with some really novel, never-before-seen feature or twist to stand out from the crowd :D

On the other hand, there is a core set of a few mappers and the basic sound hardware, and a subset of PPU behavior that works well for a large portion of games.

Exactly. People make them because it's enjoyable, and a way to hone skills. I don't think many would even want their emulator to be popular, given the burdens of having to respond to lots of users and problems.

Every feature except a debugger, which now appears to be dominated by FCEUX. If NES emulators were structured as libraries, taking a ROM and button presses and spitting out a series of RGB frames and blocks of audio samples, possibly with some GDB-style hooks for low-level control, it might be easier to add features to a front-end. Look at Maxthon: it adds new user features to the engine of Internet Explorer, made possible by the meat of IE being a COM DLL.


Unless you're trying to make the first emulator for a platform that can't run standard C++, such as SWF (ActionScript bytecode languages only), Java SE applets (JVM languages only), Java ME MIDlets (JVM languages only), Silverlight (CLR languages only), or XNA (CLR languages only).

- Who cares if Nestopia or any other NES emulator has 99% of the scene? I have my project rolling since 1998 (almost together with NESticle) and still in the scene. Too bad that Alexandre has left the RockNESX project, otherwise, it would be a great software, specially for its nice GUI. I still receive e-mails asking about it, but all I can say is "I'm sorry".

- What really matters is cooperation here.

Like this?

Not GDB-style hooks yet but...damn good idea. =] There's plenty of hooks there that could be turned into said GDB-style hooks!

edit: Fixed borked link due to Gitorious repository movement.

I don't mean any offense, but from reading your past responses, I assumed you would be such a person. Seems I have really misread you, sorry.

I'm a bit that way myself, but it's more to do with what I view as stagnation, not as much that I want to be #1. If I were in the NES scene, it would be the iNES format that would be driving me insane right now, and I would want Nestopia to take charge toward eliminating it.

In a way, I would not want to be on top because I feel the top player(s) have an obligation to help advance the scene as a whole. Sadly, none of the top players in any scene really seems to agree with me there.



Precisely! This is what I find most valuable. When someone comes into the scene and actively tries to improve things by doing their own research and then sharing their findings with everyone. That is the true spirit of emulation.

I think this is what ultimately bothers me about there being so many emulators. So many people show up, want all kinds of attention and help, make a tiny bit of progress in basically implementing our existing findings, and then they get bored and abandon everything.

That, to me, is a colossal waste of everyone's time and energy.

But just because someone only has "mapper 0" emulated doesn't mean they can't make the next big hardware quirk breakthrough.

I started to implement emulation of APU, is still at an early stage (no sound yet) but at least Dragon Quest 1 (and Dragon Quest 2 i hope, but i can't try it now) no more freeze after a battle. Links updated (v0.7).

Hi all, I'm back with a new version. See the topic for more info.

Cool, it now passes the Slalom test.
But it doesn't run my Chu Chu Rocket game (MMC3) correctly, CHR bankswitching and IRQs don't work, and the "Fast Joystick Reading" code I'm using doesn't seem to run correctly.

_________________
Here come the fortune cookies! Here come the fortune cookies! They're wearing paper hats!

You are right, I don't have implemented MMC3 yet but now that I'm quite satisfied with the level of emulation of the CPU and PPU reached, I can begin to work on APU and other mappers (starting with the MMC3).

New version added.

Changelog:
0.12:
Implemented reading of $2004 during the rendering.
(thx to Quietust for the info and for the read2004.nes test rom).