Kale's Mame Wip » mj-8956

Mixing chocolate and milk

Kale — Mon, 30 Mar 2009 19:07:39 +0000

I’m been busy with the Sega Naomi HW stuff with Haze and ElSemi. Thanks to the latter, I’ve implemented a JVS unimplemented i/o command used by the later games of this hardware.� Now Ikaruga (and others) can be coined up and are playable (although still in wire-frame)…

As a bonus track: thanks to some internet research there, I’ve fixed one of the most long standing bugs that afflicted Idol Janshi Suchie-Pai Special with the back layer pens.

Priorities still need fixing in that driver though (needs some side-by-side tests).

Last but not least: thanks to my site admins for the new blog skin.

I wanted to be the first kid on my block to get a confirmed kill

Kale — Sun, 26 Oct 2008 20:56:43 +0000

I’m currently doing some major work to the jalmah.c driver,in order to fix most of the issues in it and in search of new protection issues.Some WIP shots:

Mahjong Kakumei

Mahjong Daireikai

Mahjong Channel Zoom-In

Urashima Mahjong

Bring out the girls and let’s have a mud fight

Kale — Wed, 08 Oct 2008 20:12:03 +0000

Ok,here’s the story: the D2 register wasn’t consistent so I had to write a new ASM m68k function for this issue.In about two hours to write + debug boring opcodes in machine language I’ve done and in a uncalculated time (in between 1 and 5 minutes I think) I’ve hooked up a sort of DMA/protection thing on which the game accesses hoping that nobody will be so smart to notice that

Now I just need to clean-up the whole driver and it’s ready for the submission.

�

Next to come only on this radio station: write debug code for Alien Command,to let it finally boot.

When you hear hoofbeats, think horses, not zebras

Kale — Mon, 06 Oct 2008 19:51:53 +0000

Found the problem with it,I tried every possible thing on Earth with it and finally the simplest solution pops up: it seems that the protection jump at $f03c0 a.k.a. the palette ram sub-routine is called with the wrong vector.Differences between this call and all the others is that the D2 register is not equal to zero but is equal to 0x0b00000.My 0,50 � guess is that the programmers added this extra “feature” to protect the most interesting thing of this game.

Now I need to check the other two girls and if there’s consistency I have to add this to the sub-routine…

Where the grass is green and the girls are pretty

Kale — Sun, 05 Oct 2008 10:35:20 +0000

Fixed the bankswitch bug in Urashima Mahjong.It uses a ram location at $8001c and some combinations between the two bg roms.

�I hate to say but it happens that when a problem is fixed another one pops up.Needs to know why when you win the colors are screwed up like this:

�

Color bankswitching?

Swinging swords like Shinobi

Kale — Wed, 01 Oct 2008 22:05:56 +0000

As I’ve expected previously,this game simply uses different addresses for the video ram and two planes instead of four.It appears to be an earliest revision than the other games of this hardware,namely:

urashima -> daireika/mjzoomin -> kakumei/kakumei2 -> suchipi

Needless to say,the game now “somewhat” works.Still to do: fix/find the tile banking (as you can see in snap 4) and double check the scrolling registers.

Can it be all so simple then…?

Kale — Sun, 28 Sep 2008 16:19:54 +0000

Got a look into the jalmah.c driver again,fixed a regression which caused an hang bug into daireika,mjzoomin and urashima and fixed a protection issue with the random number generator into Mahjong Daireikai.Game now fully playable with some annoying video issues here and there.

Next episode only on this channel: get Urashima Mahjong to work.