EMULAB Forum

Please login or register.

Login with username, password and session length
Advanced search  

News:

The new forum is online, hope you enjoy it!

Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Messages - jclampy

Pages: [1]
1
In regards to using Mame on *PC's* perhaps I will retract and hold back on my judgement for now.
Just tried a Mame version from three years ago and noticed some of the minor graphical glitches in games. So yeah, it is best to check if more modern Mame version's have improvement to driver's you are interested in.
Also the 'slowness by bloatiness factor' with modern Mame versions may have been from using MameUI/32 or other frontends. Just tried Mame 0.148 command line binary and it seemed to not be effected. Although "the 'regressions' that have killed performance in certain game drivers" still exists unfortunately.
It will probably take me a while to find that 'sweet spot'. I think I'll start researching the individual driver improvements.

Addition1:
~~~~~~~
I have run some tests with the 'currently new' Mame 0.148 and found that it can run quite well with the MamepGUI frontend. The MamepGUI frontend has recently been through some version updates and it now works very well with the modern Mame builds. Using the 'currently new' versions of MamepGUI does get rid of the slowness/bloatyness factor of other 'MameUI style' builds.

Mame 0.148 does have performance problems with certain game 'drivers' which you may have to use an older build for, or an alternative emulator if possible. I have not tested alot yet, but I know CPS1, CPS2, CPS3, Sega Model 1 and System/Multi 32, all work fantastic and may arguably make alternative emulators for those particular games obsolete. So far, some game 'drivers' that don't work well are Konami GX, Sega Model 2 and Model 3. But I know there will be more.

To get 100% performance for Konami GX you would have to go back to Mame 0.129u3 or previous. Although you may also be able to use other emulators like Nebula/etc if supported?
There are Sega Model 2 and Model 3 standalone emulators which are good alternatives if you have a decent enough computer.

I think that should be enough 'food for thought' for now.

2
As I tried to hint at previously where I have mentioned Mame versions that were used as a base for various ports of Mame to portable devices. I never got around to trying emulators on some old portable devices I have and thought I might like to try them for fun.

On the PC side maybe there is no point going so 'vintage' (although I have a PIII 600MHZ I would like to try with as well) but I can see people using certain older versions of Mame due to 'regressions' that have killed performance in certain game drivers. I know of one driver regressions 'of-the-top-of-my-head' where using a Mame build from beginning of 2009 (and not any build from then to now) is the only option for best experience.

To throw a curve ball, this PC build looks interesting, although I have not tried it yet; DDEmeMame 0.37b16.99.9(2011.02.13)
Based on 0.37b16 but with some modern drivers and bells & whistles added.

All comes down to the individuals wants and needs.

My computer is getting older. Mame is getting more bloated and slower.
But that is not as simple as it sounds. If you think back when Mame could run on a PIII and tried that version on a four or five year old computer today, what would have been the CPU usage? Maybe 10% whereas, modern Mame versions probably run towards 100% CPU usage. So going back to Mame versions a couple of years ago might put you around 50% CPU usage. Having lower CPU usage means you could fire up a game when ever you feel like no matter if you have tasks running in the background, etc..

I won't be updating my PC in the foreseable future either. It can do whatever I need it to do, so not necessary to upgrade.

Anyway, not 'hating' on the modern Mame versions. I just stopped 'following' them is all. I'll still probably look at one on the odd random occasion, I'll most probably be busy elsewhere.

3
While looking something else up I came across the Mame version number for change to 'baddump' flagging;

--------------------------------------------------------------------------------
Since MAME v0.68, a new method of specifying these ROMs has been used; MAME now uses a ROM flag of 'baddump' to indicate that the ROM is incorrect/incomplete/bad. Data files for ClrMamePro should now use the 'flag baddump' syntax in addition to the actual CRC but RomCenter and many other emulators still need to use complemented CRCs.
--------------------------------------------------------------------------------

4
Hey thanks, that's great. Was worried I was going to trip up somewhere sooner or later.

I'm using the 'force nodump' header in the .dats at the moment because I like to be able to match what I see in a Mame Audit to a Clrmamepro Audit. Do you know if the "Support Inverted CRC32" might have an impact there when I get up to the Mame .dat versions that don't require it?

5
Hi Roman,

Do you know off the top of your head around what Mame/.Dat Version we would toggle from "Support Inverted CRC32" enabled to disabled?

Just a vague idea or perhaps some 'keywords' I should go and search for in the MAME previous release changelogs would be of help.

Thanks.

6
I have now tested enough to confirm that my 'method' is working perfectly.
I have it setup as one lot of files that is used both as an archive and at same time fully useable in Mame.
Currently is about four folders deep 'piggybacked' signifying each version jump and I can just add another folder per version that I am interested in.
For the Mame version that I run I just link to pertaining version folder and the preceeding version folders all the way back to my set 'beginning/foundation' folder. Each Mame version could see a complete romset and the various folders would be setup as 'diffs' between the relevant version and all previous version jumps. A 'diff' doesn't have to be each version; you can create your own 'diffs' that could be spread out by six months or a year for example. Also possible to branch out at any time for an unofficial Mame build wouldn't be a problem. Just make a 'diff' and folder to 'branch' to that unofficial Mame build.

The three caveats listed in my previous post still apply.

I'll close by saying that after using Clrmamepro for a while now and getting used to how it works I am very impressed. It is really powerful and now that I am more comfortable at using it I notice it is designed very well. At first I liked Clrmamepro because it seemed easy to use. Now I also see how extremely functional it is and the good amount of thought and wisdom that has gone into it to get it where it is today.

Cheers.

7
Here is an example of mamediff.exe output;

game (
   name aafb
   description "All American Football (rev E)"
   year 1989
   manufacturer "Leland Corp."
   rom ( name 15603-01.u4t size 65536 crc cdc9c09d )
   rom ( name 15604-01.u5t size 65536 crc 3c03e92e )
   rom ( name 15605-01.u6t size 65536 crc cdf7d19c )
   rom ( name 15606-01.u7t size 65536 crc 8eeb007c )
   rom ( name 24000-02.u3 size 8192 crc 52df0354 )
   rom ( name 24001-02.u2t size 65536 crc 9b20697d )
   rom ( name 24002-02.u3t size 65536 crc bbb92184 )
   rom ( name 24002-02.u8t size 65536 crc 3d9747c9 )
   rom ( name 24004-02.u67 size 16384 crc cd7a3338 )
   rom ( name 24005-02.u68 size 16384 crc 8ea75319 )
   rom ( name 24006-01.u69 size 16384 crc 6a576aa9 )
   rom ( name 24007-01.u70 size 16384 crc 40e46aa4 )
   rom ( name 24008-01.u90 size 16384 crc 4538bc58 )
   rom ( name 24009-02.u91 size 16384 crc b857a1ad )
   rom ( name 24010-01.u92 size 16384 crc 78705f42 )
   rom ( name 24011-02.u93 size 32768 crc 71f4425b flags nodump )
   rom ( name 24012-02.u94 size 32768 crc b2499547 flags nodump )
   rom ( name 24013-02.u95 size 32768 crc 0a604e0d flags nodump )
   rom ( name 24016-01.u13 size 65536 crc 6997025f )
   rom ( name 24017-01.u14 size 65536 crc a48bd721 )
   rom ( name 24018-01.u15 size 65536 crc 76eb6077 )
   rom ( name 24019-01.u25 size 65536 crc 9e344768 )
   rom ( name 24020-01.u26 size 65536 crc 0788f2a5 )
   rom ( name 24021-01.u27 size 65536 crc 94081899 )
   rom ( name aafbu58t.bin size 65536 crc fa75a4a0 )
   rom ( name aafbu59t.bin size 65536 crc ab6a606f )
)

game (
   name aafbb
   description "All American Football (rev B)"
   year 1989
   manufacturer "Leland Corp."
   rom ( name 24014-02.u58 size 65536 crc 5db4a3d0 )
   rom ( name 24015-02.u59 size 65536 flags nodump )
)

game (
   name aafbd2p
   description "All American Football (rev D, 2 Players)"
   year 1989
   manufacturer "Leland Corp."
   rom ( name 26000-01.u3 size 8192 crc 98c06c63 )
   rom ( name 26001-01.2t size 65536 crc f118b9b4 )
   rom ( name 26014-01.58t size 65536 crc 79fd14cd )
   rom ( name 26015-01.59t size 65536 crc 3b0382f0 )
)

Clrmamepro locks merge setting to 'non-merged sets'.

I noticed the output also doesn't have an 'header' section. I tried to add one myself but failed probably because I was trying to add an listxml header to an listinfo dat. (If I understand correctly). Whether forcing 'merge' type via 'header' of listinfo.dat is possible or would have worked in Clrmamepro on this occasion I don't know.

So, that is why I used datutil to convert the .dat over to listxml format so I could add the 'header' forcing message which gets me around this issue I was having with Clrmamepro.

I suppose you can probably see what the *real* cause of the problem is though. I suspect it is something to do with no tags like 'clone of' and 'parent of' in the .dat, but again I am not that clued up, so...

8
Ok, I have been busy away from the computer for most of the last couple of days but I have spent quite a bit of time thinking about my original idea.

I am now about 50% of the way through confirming that my 'method' can work with a few caveats.

#1 Looks like Mame has *no issues* as long as my newest 'diff' romset folder is listed first and then the rest follow in order back to the 'base/beginning/foundation' romset folder. (so rom folders are listed in reverse order of pertaining version number).

#2 Clrmamepro doesn't work as Mame does above so all 'diff' romsets must be handled individually via there respective .dat file to make sure *no issues*.
As you noted "they keep the differences in separated rompaths and use separated profiles."

#3 So when it comes to auditing you would use Clrmamepro to scan 'diff' romsets for errors and you would use Mame to do an 'overall' scan. So far I am getting perfect results running these scans.

As I said, I am only about 50% of the way through checking my method is fully sound. At the moment it looks like my method can be used as an archive and also for normal use, which is great since it saves future recompiling.

I am using 'split' merge as well if anyone wants to know.

9
mame034.exe -listinfo using WindowsXP cmd appears to work fine here.
Shame won't work inside Clrmamepro, but no worry.

Also, I think you are right about "you could run a mame -listinfo manually and redirect the output in a file and load that (most likely it's incomplete)...." as that looked the case.

As you say; I'll have to live with it.

The precompiled .dats from progettasnaps seem to work great anyway, so no problem.
Thanks.

10
Solved;

Well I'm only new to this, but I have found a solution.
Might be connected to differences between listinfo and listxml formats?

I ran my 'diff' .dats through datutil.exe and output in listxml format and now I can force by 'header' of the .dat which merge type I wanted in Clrmamepro.

I think makediff.exe outputs in listinfo format and maybe it is a limitation of that format?

Cheers.

11
Hi,

I have just noticed that whenever I make a 'diff' .dat using makediff.exe and then in Clrmamepro 'Scanner' or 'Rebuilder' I am forced to use 'non-merged' option. The other two choices are greyed out.

Is there no way I can unlock the other options when using a 'diff' .dat?
Was it only possible in older versions of Clrmamepro but not now?

I tried adding forcemerging="split" to the datfile header but no difference.
I have tried with 'diffs' .dats I have made of either new or old mame versions and no difference.

Thanks.

12
Hi,

I am trying to create/load new profile with DOS Mame034 .exe and Clrmamepro v4.09a but a error message pops up when it is going to start creating a database;

I select mame.exe and make a description
I select OLD_MAME emulator setting and create profile
click [ok]

Under 'options' box I tick "Support Inverted CRC" and "Hide Common DatFile Errors"

I select the profile in the profiles window and click 'load/update' button
Message box appears saying to create database which I click [ok]

running: c:\mame034\mame.exe -listinfo

Then straight away error box appears:
Warning: Found an incomplete datfile or a bad exe-output!
Last correctly imported set:
name: none
description: none

Any ideas how to get it to work?

Could it be some old entries in the registry from previous clrmamepro usage?

Additional Info:
I downloaded mame034 binary from mamedev website.
Also tried old clrmamepro all the way back to 3108 and still same result.

Thanks.

13
Kind of like;
Option A) First scan all specified rom paths and then afterwards if required roms are not found then start popping up error messages. Instead of popping up error messages whilst scanning only the first of the three folders.
Option B) Just slipped my mind but I think it was something to do with thought below...

Does Mame (maybe the more modern versions) work with 'multiple paths' where any duplicated roms in the 2nd path overides the 1st path, and any in the 3rd path overides the 2nd path, etc,etc?

Or, maybe mame and clrmamepro aren't made for, or don't like it when the romsets in multiple paths *overlap* each other?

I dunno, it's nearly 1am here so I am going to call it a day.
I'll check back tomorrow when I am fresh.

[Maybe my idea above is only good for archiving purposes and then I should create a 'new' folder for a compile of a *single* full romset for a particular mame version when I want to run 'that final scan audit' or to actually boot the games?]

Thanks.

14
Hmm, I thought it seemed like a good idea to keep them seperated like that at the time. Saves space from having three different full romsets and also keeps the old rom versions that would be lost otherwise. Plus needed the seperate versions for the three listed emulators that I want to try out.

But yes, I am starting to get the 'unusual' feeling as well. Because it appears to me that maybe Mame (especially these older versions) and also Clrmamepro probably are getting confused by my setup.

Am I right in thinking that there is no way or settings for Clrmamepro to be able to handle this:
Quote
15. Got mame v0.037b5 .dat from progettasnaps.
16. Completed 'New Scan' on *all three previously created folders at same time and results in a whole lot of problem messages popping up.
without the problem messages popping up?

15
Hi everyone I am new user and having a few difficulties using Clrmamepro.

I use default settings.
When I first load a .dat file in Clrmamepro I select 'no to all' questions.
I compile romsets using split method.

Here is what I am trying to do:
01. Created folder for mame v0.034 romset.
02. Used mame v0.034 .dat from progettasnaps with Clrmamepro to compile romset.
03. Completed 'New Scan' and showed up 0 missing.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
04. Created folder for 'difference of' mame v0.034 to v0.036 romsets.
05. Got mame v0.036 .dat from progettasnaps.
06. Used mamediff to create 'difference of' .dat between mame v0.034 to v0.036 .dat files.
07. Used newly created 'difference of' .dat with Clrmamepro to compile romset.
08. Completed 'New Scan' and showed up 0 missing.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
09. Created folder for 'difference of' mame v0.034 to v0.036 to v0.037b5 romsets.
10. Got mame v0.037b5 .dat from progettasnaps.
11. Used mamediff to create 'difference of' .dat between mame v0.034 to v0.037b5 .dat files.
12. Used mamediff to create 'difference of' .dat between ['difference of' mame v0.034 to v0.036] and ['difference of' mame v0.034 to v0.037b5] .dat files.
13. Used newly created 'difference of' .dat with Clrmamepro to compile romset.
14. Completed 'New Scan' and showed up 0 missing.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
15. Got mame v0.037b5 .dat from progettasnaps.
16. Completed 'New Scan' on *all three previously created folders at same time and results in a whole lot of problem messages popping up.


Is there a way to do the last part so that the problem messages don't pop up? I assume it starts throwing messages up while scanning the v0.034 folder even though it still has the other two (v0.036 and v0.037b5) folders to scan after.

Maybe I have to change some settings in Clrmamepro to reflect my use of multiple paths?

Also, am I using the best method for what I am trying to do above?

Thank you for all help/advice.

PS:
v0.034 is for PCMamev0.4a for PPC/WM
v0.036 is for PCMamev0.5wip for PPC/WM
v0.037b5 is for Mame4all and various iterations

Pages: [1]

Page created in 0.133 seconds with 21 queries.