mame 2003 gamelist.xml with data from the binary

markwkidd

@grant2258 said in mame 2003 gamelist.xml with data from the binary:

@markwkidd said in mame 2003 gamelist.xml with data from the binary:

@grant2258 said in mame 2003 gamelist.xml with data from the binary:

rungun does work you just need to hold the service key in then reset the game.

I have done some research. Later versions of MAME include a rungun.nv in the rungun.zip romset which avoids the need to hold down the service key and reset. Unfortunately MAME 0.78/mame2003 does not look for .nv files within the .zip -- although it could. Backporting that functionality is a bigger project than I want to take on at the moment.

Look forward to talking with you again soon.

I think we can just add a section to the mame2003 docs for games that need a special setup.

We cant really be messing with romsets in this version. The only other option would be to copy the nvram file but i think that would be a copyright issue in itself

Ehh.. we can totally mess with romsets in this version, if there is a good reason :). I can't wait for the expected merger with mame2003-plus, that will bring in quite a few changes to the romsets but lots of fixes and a few new titles.

The other option I've been pondering is just as you say to copy the nvram itself. This is basically a memory map just like cheats.dat or hiscore.dat. Kind of like a savestate file. It's generated by MAME and only has meaning in the context of MAME. don't think there are copyright issues.

The rungun.nv could be added to a new folder in github called, let's say, assets/nvram which scripts like RetroPie-setup could pre-install.

That's kind of ugly and doesn't work as easily for other retroarch platforms but it's true to the way MAME 'fixed' the issue in mame2010 I think it would be an option. It would be easy.

Well I know that game is not the focus of the project now! Sorry to digress!

grant2258

@markwkidd

Well its all related to be honest but we know the status is working for that roms its just the best way to get the information to people.

Well what i want to do is get all the current catver.ini unplayable that match a good status with game flags in the mame source (that was them 5 games listed). and update the source accordingly or remove the unplayable from catver.ini. After we that we can automate it without loosing any changes we have so far.

Right now the magicmame does what its supposed too it will list all the none working games as unplayable as per mame driver flags and add the Lightgun as well. What its doing extra at the moment is parsing the the catver.ini unplayable added manually and printing a please verify if its marked good in the source.

What i do need from you if possible is the original catver.ini with all the genres set right if you have that. So if i game gets fixed in the source flags the script will automatically set it right.

I had a think about the best way to do this and have a plan and i will updated it and explain it 100% when its finished

markwkidd

Was I correct in that diehard should be marked as GAME_NOT_WORKING? I couldn't find any indication that there was a 'service button' trick, or any other kind of trick, to get it to boot.

grant2258

well im not familiar with stv bios but its not working right now so we can change the region in the bios somehow it probably would work.

there is a -bios swith but retroach wont support that anyway at the command line

this is for neogeo but the same syntax should apply

http://dammit.typepad.com/blog/2011/03/setting-the-neogeo-bios-in-mame.html

grant2258

never mind the above mark your right i pressed the service button and it worked lol

markwkidd

I'm going to submit a PR that will display this messagethe first time rungun is executed. I think this kind of popup should be used sparingly, but this is a perfect case for it.

0_1521907148012_Screenshot 2018-03-24 11.54.21.png

I'm also adding the rungun.nv memory map to the metadata folder.

grant2258

@markwkidd said in mame 2003 gamelist.xml with data from the binary:

I'm going to submit a PR that will display this messagethe first time rungun is executed. I think this kind of popup should be used sparingly, but this is a perfect case for it.

I'm also adding the rungun.nv memory map to the metadata folder.

nicely done mark

markwkidd

@grant2258 said in mame 2003 gamelist.xml with data from the binary:

never mind the above mark your right i pressed the service button and it worked lol

We will have to redeem diehard in a bit and also give it a custom warning message if possible.

What is the performance like in-game? It's taking quite a while to self-test here and the video seems like it might be off.

Hmm... it seems pretty messed up.

grant2258

mame diehard -bios us

is how start it in mame078 dont think you need the nvram set i can double check though. I think your probably going to need a i7 for this one

should be able to make a diehard.ini and just add
bios us

i not sure if we support ini files though another thing i would need to check

markwkidd

@grant2258 said in mame 2003 gamelist.xml with data from the binary:

mame diehard -bios us

is how start it in mame078 dont think you need the nvram set i can double check though. I think your probably going to need a i7 for this one

should be able to make a diehard.ini and just add
bios us

i not sure if we support ini files though another thing i would need to check

It's pretty playable here in mame 0.78 with -bios us on whatever intel processor this old laptop has to offer. Lots of graphical glitches, but if we can get it working in this condition in mame2003 it's definitely not unplayable.

No there is no ini support in mame2003, by design. I have not yet encountered this scenario of needing a mame2003 equivalent of -bios flag.

Unless you have any leads on how to do that, I think that someone needs to summarize the investigation so far and start a github issue for making diehard playable.

Would you like to write it up?

grant2258

@markwkidd said in mame 2003 gamelist.xml with data from the binary:

@grant2258 said in mame 2003 gamelist.xml with data from the binary:

mame diehard -bios us

is how start it in mame078 dont think you need the nvram set i can double check though. I think your probably going to need a i7 for this one

should be able to make a diehard.ini and just add
bios us

i not sure if we support ini files though another thing i would need to check

It's pretty playable here in mame 0.78 with -bios us on whatever intel processor this old laptop has to offer. Lots of graphical glitches, but if we can get it working in this condition in mame2003 it's definitely not unplayable.

No there is no ini support in mame2003, by design. I have not yet encountered this scenario of needing a mame2003 equivalent of -bios flag.

Unless you have any leads on how to do that, I think that someone needs to summarize the investigation so far and start a github issue for making diehard playable.

Would you like to write it up?

You know more about the mame system and how it works with retroarch than I do I think it would be best if you wrote it up.

int determine_bios_rom(const struct SystemBios *bios)
{
const struct SystemBios *firstbios = bios;

/* set to default */
int bios_no = 0;

/* Not system_bios_0 and options.bios is set  */
if(bios && (options.bios != NULL))
{
	/* Allow '-bios n' to still be used */
	while(!BIOSENTRY_ISEND(bios))
	{
		char bios_number[3];
		sprintf(bios_number, "%d", bios->value);

		if(!strcmp(bios_number, options.bios))
			bios_no = bios->value;

		bios++;
	}

	bios = firstbios;

	/* Test for bios short names */
	while(!BIOSENTRY_ISEND(bios))
	{
		if(!strcmp(bios->_name, options.bios))
			bios_no = bios->value;

		bios++;
	}
}

debugload("Using System BIOS: %d\n", bios_no);

return bios_no;

}

in common.c

i guess a selector within retroarch in options perhaps off hand

markwkidd

Alrighty, issue logged.

Seems like there is a lot of developer interest in mame2003 this week. My dream would be someone else sees that and instantly knows how to fix. Regardless, it's a good start.

grant2258

thanks mark only meant to fix the catver didnt relise we would uncover these bugs along the way

markwkidd

@grant2258 said in mame 2003 gamelist.xml with data from the binary:

thanks mark only meant to fix the catver didnt relise we would uncover these bugs along the way

Yep, that's how it goes! (This has already been great for the core.)

From your perspective, what is the next step on the catver.ini update?

grant2258

well we have taken the steps to fix any issues that werent marked as bad in the source from carver. It will do what it should do generated a gamelist by genre and unplayable like you did accept it will be automated when you run mamemagic from a freshly dumped xml. I will need a updated list thats not marked unplayable and ill finish up. No more updating catver.ini manually we update at the source level that way things can get fixed.

There is also the gamelist data that is created with information in it handy to test games and see there status see the screenshots above with the information it gives you.

This will work in lr-mame plus as well according to its flags being set in the source so we should be all good to go and the catver will be as accurate as the source information.

see post 31 great for debugging and checking if things are right information wise. I just wanted a nice games list its amazing how this all happened. I put it down to the work youve put in mark

markwkidd

Well, as always I have discovered 'one more thing' that probably should happen before the catver.ini.

Back in the dark ages of a couple of months ago, I made a series of manual edits to the MAME 0.78 XML.dat to fix bad merge tags. Those changes are traceable here: https://github.com/libretro/mame2003-libretro/commits/master/metadata/MAME 0.78 XML.dat

Now that I can so easily generate an XML DAT from mame2003-libretro (although not mame2003-plus-libretro -- sigh) I see that I need to make these same edits to the source so that the generated XML is correct.

I'll do this, but it may be a few days. One more thing, one more thing.

mediamogul

@markwkidd said in mame 2003 gamelist.xml with data from the binary:

Back in the dark ages of a couple of months ago

I just can't believe the way we dressed back then. What were we thinking?

grant2258

@markwkidd

this can probably be automate as mame has a function to list bad merges it probably something that could be added to mamemagic. Id need to look into it a little more though

C:\Users\grant\Downloads\mame078b>mame.exe -lwm
CRC(2066a0b7) SHA1(6d4ccc27d6be185589e08aa9f18702b679e49a4a)
chg1 puckmana
5e.cpu newpuc2
CRC(2066a0b7) SHA1(6d4ccc27d6be185589e08aa9f18702b679e49a4a)
chg1 puckmana
5e.cpu newpuc2b
CRC(3591b89d) SHA1(79bb456be6c39c1ccd7d077fbe181523131fb300)
chg2 puckmana
hm11.5h hangly3
CRC(9d027c4a) SHA1(88e094880057451a75cdc2ce9477403021813982)
hm1.6e hangly3
np2b1.bin newpuc2b
CRC(3591b89d) SHA1(79bb456be6c39c1ccd7d077fbe181523131fb300)
hm11.5h hangly3
chg2 pacheart

i wont list them all

grant2258

@markwkidd said in mame 2003 gamelist.xml with data from the binary:

Well, as always I have discovered 'one more thing' that probably should happen before the catver.ini.

Back in the dark ages of a couple of months ago, I made a series of manual edits to the MAME 0.78 XML.dat to fix bad merge tags. Those changes are traceable here: https://github.com/libretro/mame2003-libretro/commits/master/metadata/MAME 0.78 XML.dat

Now that I can so easily generate an XML DAT from mame2003-libretro (although not mame2003-plus-libretro -- sigh) I see that I need to make these same edits to the source so that the generated XML is correct.

I'll do this, but it may be a few days. One more thing, one more thing.

well this one is fixed just need to look into the other two remaining

edit:

all fixed no need to edit files

markwkidd

I'm just posting to say that Twinaphex merged the XML generator into standard mame2003 today. I hope to play around with it more once I have time.