01666: mooncrst: Side-by-side test.

ID	Category [?]	Severity [?]	Reproducibility	Date Submitted	Last Update
01666	Side-by-side	Minor	Have not tried	Apr 3, 2008, 18:34	Nov 13, 2017, 01:28

Tester	Gary Walton	View Status	Public	Platform
Assigned To		Resolution	Open	OS
Status [?]	Confirmed			Driver	galaxian/galaxian.cpp
Version	0.37b8	Fixed in Version		Build
		Fixed in Git Commit		Github Pull Request #

Summary	01666: mooncrst: Side-by-side test.
Description	Setting both mame and PCB off together, everything is fine up to and including the sequence following the docking( i.e. docking and then the bug-type aliens ). During the next sequence, the PCB always dies first, leaving five (or six - it's hard to tell) aliens left. Mame dies a couple of seconds later leaving three aliens. This results in the next sequence being totally different as the PCB has ships II and III whilst mame has ships I and III. The demo sequence is not random and is the same each time it plays.
Steps To Reproduce
Additional Information	This report is derived from 01023 in 0.124u1. Video:
Github Commit

Flags	Verified with Original
Regression Version
Affected Sets / Systems	mooncrst

Attached Files

No.14431 Haze Senior Tester Nov 12, 2017, 22:07 edited on: Nov 12, 2017, 22:10	there's some additional evidence (including videos) for this one in the Mameworld post http://www.mameworld.info/ubbthreads/showthreaded.php?Cat=&Number=366400 strange thing to have broken since 0.37b8 (3 October 2000), but it really looks like that's the last time Moon Cresta actually worked properly.
No.14432 hap Developer Nov 13, 2017, 01:27	Nope, this isn't a regression. MAME Moon Cresta attract mode when compared to PCB has never worked properly. 0.37b8 is same as it is now. 0.36 is same as it is now too. JonBoyRetro at MAMEWorld said it's fine on Mame4All PSP v4.9r2, which is based on 0.34. But 0.34 is still not right. While it does show 2 'docking sessions', the demo play not identical to the PCB. If you change vsync to exactly 60hz in current MAME, attract mode will be same as in MAME 0.34. It's possibly video(interrupt) timing related. Or something more obscure, like number of CPU cycles from power-on until 1st IRQ.

No.14431

Haze

Senior Tester

Nov 12, 2017, 22:07

edited on: Nov 12, 2017, 22:10

there's some additional evidence (including videos) for this one in the Mameworld post
http://www.mameworld.info/ubbthreads/showthreaded.php?Cat=&Number=366400

strange thing to have broken since 0.37b8 (3 October 2000), but it really looks like that's the last time Moon Cresta actually worked properly.

No.14432

hap

Developer

Nov 13, 2017, 01:27

Nope, this isn't a regression. MAME Moon Cresta attract mode when compared to PCB has never worked properly.
0.37b8 is same as it is now. 0.36 is same as it is now too.
JonBoyRetro at MAMEWorld said it's fine on Mame4All PSP v4.9r2, which is based on 0.34. But 0.34 is still not right. While it does show 2 'docking sessions', the demo play not identical to the PCB.

If you change vsync to exactly 60hz in current MAME, attract mode will be same as in MAME 0.34.
It's possibly video(interrupt) timing related. Or something more obscure, like number of CPU cycles from power-on until 1st IRQ.