Cheat Engine Update

[ianpatt]

New cheat engine update: mostly bugfixes and additions to the watch feature.

Tell me if there's something I forgot or you want added.

source

[stephh]

New cheat engine update: mostly bugfixes and additions to the watch feature.

Here is another little bug report related to the new features ...

When you modify the "Add value" of a watch in the "Edit Cheat" menu, then decide to add the watch, the "Add value" isn't transmitted : it is ALWAYS 00 !

The "Address shift" only shifts right (>>) but not left (<<), so I can't "convert" the addresses given by the cheat engine to addresses I can use in the debugger As I don't think it is needed to shift the address up to 31 bits, coudldn't it be possible to shift it up to 15 bits right (values 1 to 15) and up to 15 bits left (values -1 to -15) ?

What is the use of "Data shift" ? Where could it be needed ? IMO, it would be better to have a "XOR value" (1 byte only) ...


And now a lttle proposition to end this post ...

When you define an ASCII watch with multiple bytes (elements > 1), there shouldn't be a "space" after each character ...

For example, the watch displays "U P S " while IMO it should display "UPS" ... Can you fix that ? This would allow to display more text on the same line ...


That's all I found at the moment ... Let me know if you need more ...

Steph from The Ultimate Patchers

Visit The Ultimate Patchers' site

[Bugfinder]

Here's a small bug to report: when you have a ROM cheat with "restore previous value" attribute on, and you assign that cheat to enable automatically, if you turn it off the previous value isn't restored.

[ianpatt]

.... I take back what I mentioned on MAME General/Serious. These fixes didn't take much time to implement, so I'm making an exception to my "no coding for a bit" rule.

Anyway, everything in stephh's post is now fixed.

What is the use of "Data shift" ?

For converting TMS addresses stored in memory to logical addresses. XOR added also.

The issue with auto-enabled restore-on-disable ROM cheats was present for all types of auto-enabled restore-on-disable cheats. It has been fixed also.

source

[stephh]

.... I take back what I mentioned on MAME General/Serious. These fixes didn't take much time to implement, so I'm making an exception to my "no coding for a bit" rule.

Thanks for doing that ! When I posted the previous message, I thought that MAME 0.62 was about to be released ... I couldn't guess that blah blah blah ...

Anyway, everything in stephh's post is now fixed.

I'll make some tests, and I'll let you know ...

What is the use of "Data shift" ?

For converting TMS addresses stored in memory to logical addresses.

I thought it was "Address shift" ... Or did I understand something wrong ?

XOR added also.

I guess that "Bugfinder" will be happy now ...

Steph from The Ultimate Patchers

Visit The Ultimate Patchers' site

[Bugfinder]

XOR added also.

I guess that "Bugfinder" will be happy now ...

Steph from The Ultimate Patchers

I'm the one who didn't get the idea now. Why will I (specially) be happy with this XOR data shift? I have not tried the new cheat engine version yet.

[ianpatt]

I thought it was "Address shift" ... Or did I understand something wrong ?

Data shift changes the value displayed in the watch, address shift only changes the address input.

In pseudocode: ShowWatch(ReadData(address >> addressShift) >> dataShift);

Why will I (specially) be happy with this XOR data shift?

(random guess) CPS2 ROM cheats?

[Bugfinder]

Why will I (specially) be happy with this XOR data shift?

(random guess) CPS2 ROM cheats?

Hmm, now I get it... Really cool! Thanks for the nice addition

[Bugfinder]

I was assigning keys to cheats, and find out that some codes present in cheat.c don't correspond to the key actually assigned. As an example, the codes for "keypad +" and "keypad del" are respectively 53 and 54 according to the source, but these are the codes for Up and Down. You need to add 5 to these codes to get the correct codes for [+] and [DEL].

[ianpatt]

I was assigning keys to cheats, and find out that some codes present in cheat.c don't correspond to the key actually assigned. As an example, the codes for "keypad +" and "keypad del" are respectively 53 and 54 according to the source, but these are the codes for Up and Down. You need to add 5 to these codes to get the correct codes for [+] and [DEL].

Fixed table:

Code: Select all

Key Index List:

	A		00	Q		10	6		20	F3		30	[		40	PGDN	50	RCTRL	60
	B		01	R		11	7		21	F4		31	]		41	LEFT	51	LALT	61
	C		02	S		12	8		22	F5		32	ENTER	42	RIGHT	52	RALT	62
	D		03	T		13	9		23	F6		33	:		43	UP		53	SCRLOCK	63
	E		04	U		14	[0]		24	F7		34	'		44	DOWN	54	NUMLOCK	64
	F		05	V		15	[1]		25	F8		35	\		45	[/]		55	CAPSLCK	65
	G		06	W		16	[2]		26	F9		36	\		46	[*]		56	LWIN	66
	H		07	X		17	[3]		27	F10		37	,		47	[-]		57	RWIN	67
	I		08	Y		18	[4]		28	F11		38	.		48	[+]		58	MENU	68
	J		09	Z		19	[5]		29	F12		39	/		49	[DEL]	59
	K		0A	0		1A	[6]		2A	ESC		3A	SPACE	4A	[ENTER]	5A
	L		0B	1		1B	[7]		2B	~		3B	INS		4B	PRTSCR	5B
	M		0C	2		1C	[8]		2C	-		3C	DEL		4C	PAUSE	5C
	N		0D	3		1D	[9]		2D	=		3D	HOME	4D	LSHIFT	5D
	O		0E	4		1E	F1		2E	BACKSP	3E	END		4E	RSHIFT	5E
	P		0F	5		1F	F2		2F	TAB		3F	PGUP	4F	LCTRL	5F

The tabs work out in Visual SlickEdit, really they do. Just replace " " (four spaces) with "\t".

I've already submitted the latest updates, so this will be included in the next batch. (ie. probably not in .62)

[kelvSYC]

So, on a Mac, what value corresponds to the Command keys?

[ianpatt]

So, on a Mac, what value corresponds to the Command keys?

I don't know off the top of my head, but if you add this line of code:

Code: Select all

usrintf_showmessage_secs(1, "%.2X", code);

after line 4419 in the current source (it should look like this when done)

Code: Select all

...
					else if(	(code != CODE_NONE) &&
								!input_ui_pressed(IPT_UI_SELECT))
					{
						entry->activationKey = code;
						entry->flags |= kCheatFlag_HasActivationKey;
						
						usrintf_showmessage_secs(1, "%.2X", code);

						editActive = 0;
					}
...

then the cheat engine will put up a messagebox with the code of the key you pressed when you assign an activation key.

[stephh]

Ian, there are still some problems with address shifting

This is because

Code: Select all

	address >> -1

is DIFFERENT than

Code: Select all

	address << 1

So I guess that the following line (in Edit Watch function) is the cause of the problem :

Code: Select all

	sprintf(buf[total], "%.*X", cpuInfoList[entry->cpu].addressCharsNeeded, entry->address >> entry->addressShift);

Could you please confirm this and tell me if there are some other places where some other changes are required ? Thanks by advance for what you can do ...

Steph from The Ultimate Patchers

Visit The Ultimate Patchers' site

[ianpatt]

Missed one when copy/paste/replace.

Code: Select all

sprintf(buf[total], "%.*X", cpuInfoList[entry->cpu].addressCharsNeeded, DoShift(entry->address, entry->addressShift));