Try to load the Bootgod XML for this game in Nestopia and see how messed up it is. When using Nestopia's internal database, the game runs fine. When using a header built from the Bootgod database, it runs fine. Either Nestopia isn't reading the XML file correctly, or the XML file has an error in it somewhere for this game.
this happens because most emulators cheats with Konami VRC pcbs emulation.
VRC2 and VRC4 pcbs came with different wiring on the board, changing which bits (of suitable memory addresses) control PRG and VROM bankswitch. iNES headers handled this by shuffling stuff in different mappers (that's why you find VRC2 and VRC4 in different mappers, if you ever wondered... check mapper docs), NEStopia internal db identifies by crc the different cases and sets up the proper bits. in MESS software lists, we have some specific fields for these Konami boards, which prescribe the correct wiring.
I guess bootgod xml do not include anything about these settings, by default, given that no emu search for these except MESS and NEStopia (others rely on mapper numbers)
EDIT: It seems there is some interest in NES v2 support as the de-facto standard for No-Intro NES CRCs. I might (no promises!) bang together a version that can do NES v2 from an XML database as well.
EDIT2: Don't know if anyone has noticed but BootGod updated his XML on Oct. 22, 2012. That's a year worth of updates!
So, no-intro only concerns itself with the actual NES ROM data, excluding iNES headers?
I noticed when I tried to play Quest of Ki with a ROM from a no-intro set, and it was glitched in FCEUX. Tried the ROM from the GoodNES set and it worked fine.
Took me awhile to realize the only difference was the ROM in the "no-intro" set had the wrong mirroring flag in the header.
(though even with the bad header, both ran fine in Nintendulator and Nestopia. So it's probably a FCEUX emulation problem, as Quest of Ki is a MMC3 game, so the mirroring is supposed to be mapper-controlled anyways, meaning the iNES mirroring bit should in theory be irrelevant for MMC3?)
Since the No-Intro set doesn't have headers, I fail to see how you had a 'No-Intro' ROM with a bad header.
As long as the 'data' part matches the dat, you can have any 'mickey mouse' header you want. Therefore there could almost be an infinite amount of 'No-Intro' nes collections that would pass a dat scan.
To surmise, any No-Intro dat that uses the header skip feature means that there is no such thing as a No-Intro set of 'roms' for it.
I should have added that there are people (maybe within the No-Intro members) that have gone out of their way to add, tweak and fix headers so that many new dumps are playable in the 'older' emulators.
It is misleading to then share them as 'No-Intro' dumps though, which is probably why there are so many bad headered collections named as No-Intro.
