Print

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

[nimrodeo]

Hi, I'm a long-time clrmamepro user, but just recently joined the forum.

This question is about the way nodump ROMs are tabulated and created.

Is it intentional for clrmamepro to not tabulate status="nodump" files when there is a size="0" tag included?  I have found in the MAME Software List hash files folder that there is an xml file that declares a nodump on the same line as setting the size to zero.  The result is a zero length file included in the .zip file that is created for the set.  The .xml file I am talking about is "amiga_a3000.xml", the set is "amix11", the zero-length file is "cass".  Since the size is zero, the fact that the "cass" file is a nodump is apparently ignored (and not counted) for the summary seen at the end of the scan.
 
There are two other files where the same thing happens, apparently due to a MAMEdev specifying a size="0" on a status="nodump" line: "jupace_cass\m_valkyr.zip" and "jupace_cass\triplep.zip"
 
In other MAME Software Lists hash folder files, when status="nodump" is declared, and there is not a size specified, the nodump is tabulated on the scan in the normal way.  And of course, a zero-length file is not created.
 
Perhaps this is intentional because developers need to declare zero-length files as placeholders for parent/child sets, or something?  Or is it just an oversight?  A possible work-around would be to ask the MAMEdevs to not place such conflicting tags in their .xml files.  Or to remove the size="0" tags on nodump lines by hand.  But this issue may come up again on future MAME relases.

I have another question, but I will make a separate post for it.

[Roman]

I guess I have to double check but I think a given size, even if it's 0 marks a file as valid dump. A nodump flagged file got no sha1/md5/crc32 and no size value. That's how I see it from the cmpro developer point of view. If MAME returns a size=0 and nodump flag I guess I have to add something for that (if it's not already in....) so that the nodump flag got a higher priority than the others.
Good that you already gave an example...
A nodump is indeed "ignored" (well it's listed in the set information screen) but cmpro will mark it as unneeded if you have a dummy file for it (unless you use other datfile specific settings to force nodump availability).

So yes, I will double check files with size=0 AND nodump flag....

[nimrodeo]

...

So yes, I will double check files with size=0 AND nodump flag....

Thanks!

[Roman]

fixed for next version

[nimrodeo]

Thank you very much for looking into this and fixing it.

I tested this on the new build, and I found it mostly works.  The part that works is that zero-length files are no longer created, which is good.

The part that doesn't quite work as expected is that some of the nodumps are no longer tabulated on the stats results of the scan.

I found that these lines created nodump tabulations:
    <rom name="cass" size="0" status="nodump" offset="0" />

And these lines did not create any nodump tabulations:
    <rom name="cass" status="nodump" offset="0" />

So that part is reversed, formerly the lines with the extra size="0" tags weren't counted and the plain nodump lines were.  in the new build, it's the lines with the extra size="0" tags that are being counted and the plain nodump lines are not.

If this is the intended behavior, just let me know.

[Roman]

http://mamedev.emulab.it/clrmamepro/binaries/cmp20150816.rar

that should solve your nodump issue

[nimrodeo]

Yes, that works as expected, both kinds of nodumps are tabulated, and neither one creates a zero-length file.

Thank you so much for working this out.