To Roman (or anyone that might have insight):

Being home sick today I decided, that I would try to resolve the whole Red Light / Green Light issue. I recently audited MAME32 .119 and still was getting a Red light, even though I was missing nothing AFAICT. But then, doing a separate scan of JUST samples revealed that I had 4 failing audit, even though they were passing while scanning Samples as part of the "Roms" audit.

So here's what I did, and here are the results:

1) Uninstalled CMPRO (even though I had latest version installed)
2) Reinstalled CMPRO (3.104b)
3) Re-downloaded DAT files to via WWW Mode using URL on CMPRO site (www.clrmame.com/mamedats.zip)
4) Created new profile for MAME32 .119 with a description of "M32_119" in
5) Double-clicked on the new profile:
--A database update was needed (OK)
--At 40% it said the Sample merge relationship will be removed for an alternative samplefolder (OK TO ALL)
--At 80% it asked about setting a missingle alternative samples folder to something else (YES TO ALL)
--Then it asked about removing a parent/clone relationship to a cloneset (OK TO ALL)
--Then it said 'Atomiswave Bios' and 'Naomi 2 Bios' aren't used by any set (OK)
6) Created settings for scanning (ROM Paths, Samples Paths, and Add Paths) and saved as Default
7) Went to Scanner and selected the following settings:
--You Want To Scan: (All checked by default)
--You Prefer (Split Sets)
--NoDump Handling (Hide Missing [even though at one point weeks ago I chose 'Create Missing'])
--Check (All checked by default)
--Fix (Clicked + to select all checkboxes)
--Options (Unchecked Ask Before Fixing)
8) Clicked Scan

Results:
Scan Time 1:43
Missing Sets: 0 of 6737
Wrong SetName Case: 0 [0 fixed]
Wrong Named Sets: 0 [0 fixed]

Missing Roms: 0 of 95596
Missing Bytes: 0 of 55gb
Missing BIOS Roms: 2 of 102
Wrong Named Roms: 0 [0 fixed]
Wrong RomName Case: 0 [0 fixed]
Wrong Sized Roms: 0 [0 fixed]
Wrong Date/Time Roms: 0 [0 fixed]
Wrong CRC32s: 0
Wrong MD5s:
Wrong SHA1s:
(OK)

9) Back to Profiler:
--Shows Red Light for Profile
--Version 0.119 (Sep 13 2007)
--Missing Sets 0 / 6737
--Missing ROMs 0 / 95596

Results for Other Datfiles:

MAME Artwork 20070909 (Passed, Green Light)
MAME Cabinets 20070205 (Passed, Green Light)
MAME Control Panels 20070913 (Passed, Green Light)
MAME Flyers 20070913 (Passed, Green Light)
MAME Marquees 20070913 (Passed, Green Light)
MAME Samples 20070913 (Failed -- 134 Wrong Sized Roms)
Mame32 titles V0.118 (Passed, Green Light)
Official Mame32 Cabinets .116 (Passed, Green Light)
Official Mame32 Icons .119 (Passed, Green Light)
Official Mame32 Screenshots .119 (Passed, Green Light)


Bottom Line: Both MAME32 .119 and MAME Samples aren't getting the green light.

I'm baffled that the Samples would pass audit during the regular ROMS scan but not for its own datfile.

Is anyone else (still) having this problem? If you did and resolved it, how did you go about doing it?

Samples sources have been changed yesterday. Get the samples from mameworld.net/samples.

The red light: You miss something. Be sure you scan for all sets/systems.
Have a look at the mame's profile .cmp file (settings folder), open it, look out for things like:

Misc_MissingSets = 0
Misc_MissingRoms = 0
Misc_MissingBIOSRoms = 0
Misc_MissingSamples = 0
Misc_MissingDiskImages = 0
Misc_WrongSizedRoms = 0
Misc_WrongChecksums = 0
Misc_MissingNoDumps = 0
Misc_MissingCompleteNoDumpSet = 0
Misc_MissingCompleteBIOSRomsSet = 0
Misc_MissingCompleteNodumpBiosSet = 0


Roman Scherzer

One WAV file is still "wrong size", but down to 1 from over 100 is a marked improvement. Thanks for the heads-up on that.

Regarding the CMP file in the settings folder, the only thing I could find out of whack was:

Misc_MissingBIOSRoms = 2

But the problem I'm having is, it's not telling me which ones are missing. The Scan Results window comes up empty. For grins I clicked the Set Information button and the Set Information window appeared for ALL the sets.

I understand that something changed from MAME 116u4 as far as the way BIOS roms are handled, but the technical stuff is over my head. I feel like I missed a step somewhere...

a) It's the mac2bios set. It's a full nodump set. Go and create it for now...
b) For which sample do you have a differnet size?


Roman Scherzer

> a) It's the mac2bios set. It's a full nodump set. Go and create it for now...
> b) For which sample do you have a differnet size?

I suppose it is the explos2.wav of Star Cruiser (starcrus). I checked with the official starcrus from Twisty's site, which should be reference. Dat information on that file and the actual file don't match (size in dat: 115942, actual size 106178). New recording or wrong in the dat file?

ah...had a discussion with Twisty about that file some days ago. His old one had a non-0 sample start while the one at mamedevs was fine. Looks like he replaced it. Or not. Will check that later.


Roman Scherzer

I have exactly the same problem with the Starcrus sample not passing the scan (missing rom: explos2.wav [size: 115942] [CRC32: e84965b1] [SHA1: 3f26184c8c7a7c8ce241aa611546784eefe19e47]). I have also download the file agian to make sure that it wasn't corrupt, but still no look. Also get the same problem with the 2 BIOS files.

Any ideas?

> ah...had a discussion with Twisty about that file some days ago. His old one had
> a non-0 sample start while the one at mamedevs was fine. Looks like he replaced
> it. Or not. Will check that later.
>
>
> Roman Scherzer
>



Roman Scherzer

for a), it will be fixed in the next version.

> a) It's the mac2bios set. It's a full nodump set. Go and create it for now...
> b) For which sample do you have a differnet size?
>
>
> Roman Scherzer
>



Roman Scherzer

> > ah...had a discussion with Twisty about that file some days ago. His old one
> had
> > a non-0 sample start while the one at mamedevs was fine. Looks like he
> replaced
> > it. Or not. Will check that later.
> >

The updated sample of Star Cruiser on http://www.mamechannel.it/pages/samples.php does match your previous dat before you updated it. The sample explos2.wav therein seems to be different from the one on Twisty's page (did not check, he may have changed that now as well). Do you happen to know, which one is the correct one for Star Cruiser and what is the correct CRC value then for the dat? I am pretty confused now to be honest.

old situation:
mame.net hold a clean (starting with a 0 sample) version of that explosion2 file
mameworld.net/samples had a different file which had a click at the beginning.

current situation:
- mame.net is gone
- mameworld.net/samples got that one file from the old mame.net set.


If you have a different one, try downloading the set again from mameworld.net. Most likely the old one was cached.

The 152f7bdf is the correct one as stated in the dat and that's what you currently download from Twisty's page.


Roman Scherzer

> The 152f7bdf is the correct one as stated in the dat and that's what you
> currently download from Twisty's page.
>

Thanx for the clarification. Downloaded the file from Twisty's page and everything is fine again. Not that I would notice the difference between the two files, which I had, to be honest. In particular I could not hear a click in the beginning of either of the files.

Will be a much nicer situation once we can get rid of all the samples due to perfect emulation of discrete circuits someday in the hopefully not too distant future. won't take any bets on when that will be, maybe never, but who knows. At least a few samples were removed not very long ago, because they are not needed anymore.

It is interesting to see all the decryption having been broken recently, but not so much happening on discrete circuit emulation (or is simulation more correct in this context, ah never mind). One should think that this task would be more straightforward, though certainly not a piece of cake, don't take me wrong.

The "tzz" noise at the beginning is hearable (it's very very very short though), although it depends a little bit on your soundcard as well. If you look at the waveform of the two files you see that one start with a zero sample, while the other doesn't and that non-0 value create the "tzz".


Roman Scherzer