Logiqx

[kazumi213]

Hi Logiqx.

I've recently decided to check the xml DAT output from latest official MAME executable:

mamepp -lx > m128b.dat

I then used latest DatUtil to cleanse it and remove all clones, as I'm only interested in possible parent differences.

datutil -f lx -k -r -v m128b.dat

The log reports correct duplicated ROM and sample removals (thanks for DatUtil!), but it also reports the removal of 6 "sampleof" declarations involving the sample archive "alphamc07". The affected games are:

equites
gekisou
bullfgtr
kouyakyu
splndrbt
hvoltage

I understand why "alphamc07" is removed by DatUtil. It is incorrectly referenced, there is no "alphamc07" game. "alphamc07" should be named like one of the previous games.

Your MAME DAT is also missing "alphamc07" sample set reference. However, the sample files in the "alphamc07" archive are still there, so I think it should be included, at least by manually adding the removed "sampleof" declarations to your DAT.

Finally I used the cleansed DAT (with manually fixed sampleof) to audit my collection using latest CMPro. The only difference that I found was regarding the game "chinagat", for the following 4 ROMs:

a-12
a-13
a-14
a-15

Your DAT contains hashes for them and marks them as baddump. DAT from MAME executable marks them as "nodump", hence CMPro reports them as unneeded.

Just wanted to let you know. Please sorry if you have your reasons to keep these differences.

[Logiqx]

I'll have to be quick as I don't have a lot of time...

The references to "alphamc07" are removed because they are invalid. DatUtil removes the references so that the output dat does not contain errors or ambiguities that could be handled differently by CMPro or RomCenter - that's what DatUtil does. I don't see why it is worth manually editing the datafile afterwards, making it invalid again. The problem can be resolved in MAME itself by choosing a more sensible "sampleof".

The "chinagat" thing is because DatUtil finds the CRCs in a related clone. At the same time, it marks them as "bad" in the parent so that it is clear that they may not be the correct ROMs for the parent (n.b. they are correct for two of the clones though). This is a design decision and intended to remove ambiguity in the datafile. Occurences like this could be handled in a variety of ways by CMPro and RomCenter but I made DatUtil generate datafiles that are most likely to behave the same in both tools.

[kazumi213]

Understood. And I agree. Thanks for your work and the time you put on it Logiqx