Page 16 of 37

Posted: Wed Apr 16, 2008 6:53 pm
by Shiru
AamirM wrote:2) Zero the Kamikaze Squirrel : Yes, there is little flickering in the sky and is still not fixed in latest Regen. I hope it happens on real hardware too :). Can anybody check please?
I don't have real console for now, but I'm almost 101% sure that this flickering is not happens on real console. I played this game much and I'd surely noticed if there was flickering. That also looks like bug, because many other games have similar BG scrolling without any flickering. From programmer's point of view, I even don't know how to make such flickering on SMD by mistake (there is hardware line scroll and no need to tricks with H-int), and it will be not so easy to make this intentionally.

Posted: Wed Apr 16, 2008 7:05 pm
by Sik
I've just checked the game in both Fusion and Regen and both have flickering... or at least what I think is the bug. I have the impression that the game is using raster effects, through not for the scrolling, but to make the gradient. It's a good way to increase the number of colors on screen, trust me ;)

Posted: Wed Apr 16, 2008 7:15 pm
by Shiru
Sik wrote:I have the impression that the game is using raster effects, through not for the scrolling, but to make the gradient.
You're right, I didn't noticed this - bottom BG layer uses only one palette, but has 22 colors.

Posted: Wed Apr 16, 2008 7:15 pm
by Captain Boomerang
AamirM wrote:BTW, My exams finished today :D so I now have resumed work on Regen and just finished up AVI recording support and ROM browser. Cheat searching and editing shouldn't take much time but I guess we'll see.
Hi,

Glad to hear it mate! I'm curious, what's the status on the bug reports and some of the other bug reports blokes have posted about here? Have you figured out how to fix pro action replay codes yet?

Thanks!

-Adam

Posted: Wed Apr 16, 2008 7:17 pm
by Eke
@Aamir: huh, this is strange, how could it be possible ??

about Striker, I guess I need to check my timings & hv values again (let's hope, for the last time ;-))

Posted: Wed Apr 16, 2008 7:46 pm
by AamirM
Hi,

@ Shiru

Ok then. I'll try to fix it.

@ Adam

I still haven't figured out the PAR. To be honest, I have not given much time to fix it :( . Steve Snake seems to be busy with other stuff otherwise I would've asked him as he knows lot more than I do. The problem is that some code will work perfectly while other which are no different won't. Like here is an example, these are the cheats for power rangers using my latest modifications to PAR:

Code: Select all

[0]
Name=Infinite Time *works*
0=Off
1=On,FFC110:0099

[1]
Name=Infinite Health P1 *works*
0=Off
1=On,FFD080:0060

....

[3]
Name=Max Special Bar P1 *doesn't work (needs to be byteswapped)*
0=Off
1=On,FFD082:0036

[4]
Name=P1 Can't Be Hit *doesn't work (needs to be byteswapped)*
0=Off
1=On,FFD089:00FC
About the bugs. If its on the list, it'll get fixed. Most of the bugs are fixed. The list is on ZSNES boards and also here.

@ Eke

Very strange indeed. Did you set the region to something or changed any other setting?

stay safe,

AamirM

Posted: Wed Apr 16, 2008 8:00 pm
by Sik
Them all are byte writes, not word writes! Man, all the values to write are < 0x100. This means that they're byte writes, not word writes! Ugh, how many times do we need to explain you? I've lost the count of times it has been told to you already >_>

Posted: Wed Apr 16, 2008 8:06 pm
by AamirM
Hi,

FYI, I am doing byte writes. I repeat, I am doing byte writes.

stay safe,

AamirM

Posted: Wed Apr 16, 2008 8:10 pm
by Sik
Oh, well, then you're doing something wrong when dealing with the bytes...

What do those variables do, exactly?

Posted: Wed Apr 16, 2008 8:11 pm
by Captain Boomerang
You handle word writes too, correct?

If you don't mind, could you post the portion of the source that handles PAR codes here? Maybe I or somebody else here can take a look at it and test it and see if we can help find a fix for it.

Posted: Wed Apr 16, 2008 8:31 pm
by AamirM
Hi,
Sik wrote:Oh, well, then you're doing something wrong when dealing with the bytes...

What do those variables do, exactly?
If I am doing something wrong, then how do you explain the other cheats working correctly which are technically same as the ones not working.
You handle word writes too, correct?

If you don't mind, could you post the portion of the source that handles PAR codes here? Maybe I or somebody else here can take a look at it and test it and see if we can help find a fix for it.
Yes, word writes are handled too. The code is not wrong. I am sure.

stay safe,

AamirM

Posted: Wed Apr 16, 2008 8:41 pm
by Sik
AamirM wrote:
Sik wrote:Oh, well, then you're doing something wrong when dealing with the bytes...

What do those variables do, exactly?
If I am doing something wrong, then how do you explain the other cheats working correctly which are technically same as the ones not working.
That kind of bugs normally affect some parts of the stuff (in this case codes) but leave the others working properly. I know what I mean, I get the weirdest bugs all the time, and normally they're caused by a stupid mistake that you would have never thought of. And trust me, this happens most of the time :P

Posted: Thu Apr 17, 2008 12:41 am
by King Of Chaos
Well, one could use deductive reasoning to try to figure out the problem. What's similar between the codes that do work, and do not work? That's a start I think. :)

Posted: Thu Apr 17, 2008 2:44 am
by Sik
First try changing the address of the codes that need byteswapping. In the given examples:
  • FFD082:0036 -> FFD083:0036
  • FFD089:00FC -> FFD088:00FC
First let's make sure the codes are OK, otherwise then it's just a waste of time :P

Posted: Thu Apr 17, 2008 7:02 am
by TascoDLX

Code: Select all

FFC110:0099
Yes, this code seems to be working. FFC110 = 99.

Code: Select all

FFD080:0060
No, Regen has this wrong. It's doing FFD080 = 0060. The correct adjustment is FFD080 = 60.

Code: Select all

FFD082:0036
No, same problem as the previous code. Regen has FFD082 = 0036 instead of FFD082 = 36.

Code: Select all

FFD089:00FC
No, Regen doesn't appear to be writing any value! Should be FFD089 = FC.

After a few tests, it seems Regen has major problems writing a given value to the correct address. Good luck with the bug hunt! :wink: