Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[oddi]

Hi Roman , see :
full Mame 0.159 sets ( unchecked - sets, roms, sampes ), only enable chds :



ignor warning window about uncheked sets:

and see results  : Line - Fixed Wrong Name



0/16 fixed
Maybe cmp bug about new bios chds. Tnx

P.s - with checked all options -. sets, roms, samples, chds , same result, 0/16 fixed

and - if reopen dat and used button "scan " not "new scan"- CMP report correct results - 0/0

[B2K24]

I always scan with official MAME binary and can't repo this behavior. I'm unsure why you would scan with just CHDs checked and leaving sets unchecked 

http://imgur.com/ayEbLm6

http://imgur.com/CCzIycA

[Roman]

nobody really cares about the stats screen .. ;-)
if the values are gone after a 2nd scan it might be some garbage left over....low prio...I will look into it some day..

[oddi]

OK Roman , statistics is low prio

but this where is problem ...

Create full mame dat from exe with clrmamepro , if scan - all is ok

If create dat from "Set information" --> "Export" and load this dat in CMPRO and scan...



+few other same warnings

and






little fragment from first warning :



whats u think , buggy mame code ?

[Roman]

well...you're exporting the full-merged subfolders....what do you expect when you load such a dat....

[oddi]

OK, forget it, Tnx

[Roman]

nah...I will look into this...actually the export should not export the subfolders...

[Roman]

ok...found (and fixed) the reason for the 0/16 count....
however I was not able to reproduce your export (with subfolders) issue...

[oddi]

hmmm, strange job, settings:

Misc Profiler Options -> Parsing Options -> Parser ROM 'merge' tags: enabled   (default)
Misc Profiler Options -> Parsing Options -> Parser DISK 'merge' tags: enabled  (default)
Misc Profiler Options -> Hash collision -> Naming Pattern: %f\%1   (default)
Misc Profiler Options -> Hash collision: All Clones in Relationship

Settings -> Full Merge Mode: Hash Collision Name

and same issue :
update mame source code --> load exe in cmp --> set information --> export
load new dat and again issue .
Test with 4.17 and your new build from other thread (from yesterday )

Roman, here upload dat generated from export:
https://www.sendspace.com/file/ij1bbz

Tnx

[Roman]

I can now reproduce the export problem with the latest official cmpro but I cannot reproduce it with yesterdays build (http://mamedev.emulab.it/clrmamepro/binaries/cmp20150310.rar)

I can even trigger the appearance of the issue with one and the same profile when I simply change the cmpro exe files...so yesterday's build from the Merged sets and Super Kaneko Nova System BIOS thread does fix the export problem....

I somehow doubt your results with that build...be sure you have cleaned the cache..and use the correct exe.

The latest build does export the dat without the "/" subfolders (e.g. for "sknsj1.u10" in "cyvernj")...and loading such a fresh exported set should not show the problem.

Your provided datfile is of course bad...and will stay bad when loaded...even with yesterday's build..

[oddi]

U are right, i little mistake with previous test, now with the last forum build  export is correct
Many tnx