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

I haven't looked into the actual score but I could imagine that both USA sets (the demo and the normal one) might get the same score value and then maybe the last one is prefered for no real reason. I guess the algorithm was more for a scenario where you only got distinct regions within one parent/clone relationship.

Got it. Funny, it probably shouldn't be picking a demo ROM in this case (as there's a retail one available). I'll look into the algorithm. I also think DAT-o-matic's DATs might be to blame here, as they omit a lot of information in the XML (like language, for example) and maybe they're also to blame for the algorithm picking up a demo ROM. That could explain why it's missing from my output set: I don't have Demos, Samples, Prototypes, etc.

I'll look closer into the scoring system to find out why it's picking up that demo instead of a retail ROM.

I've checked that 'tutorial' and I don't see a problem with it. I used their description on the Nintendo datfile (used their region ordering USA/CAN/AUS/EUR....) and I still get Mario Kart listed. With the USA regio as highest prio, the algorithm picks Mario Kart DS (USA) (Demo) (Kiosk).nds / eb26155d as 'the chosen' set for that parent/clone relationship.

If people say "may game xxx is not appearing at all" then normally 3 cases might happen:
1) the game belongs to a region which is not enabled and so the set is filtered out
2) due to the scoring mechanism and the arrangement in the dat, a different clone/parent is picked as you might expect (see above Mario Kart DS example, I personally would have picked the "Mario Kart DS (USA)" set and not the "Mario Kart DS (USA) (Demo) (Kiosk)"....but of course scoring algorithms aren't perfect and depending on the datfile it picks what scores best and not what you think is the best guess ;-)...in such a case you might better filter out demo/kiosk versions before.
3) your xxx clone is actually used but gets a different naming due to the definitions in the datfile

So again....the scoring algorithm was designed by Logiqx and is used in several rom managing tools and runs for a decade now....every now and then users come up and are not satisfied with what the algorithm picks...but that's the way it was designed. Surely you can write your own routine...but you cannot compare it to what is used in various rom managing tools.

So I made a quick test....downloaded the Nintendo - Nintendo DS (Decrypted) (Parent-Clone) (20200126-032320).dat with that DAT-o-matic, loaded it in cmpro, enabled the 1G1R option, enabled EUR as region and I get Mario Kart DS (Europe) (En,Fr,De,Es,It) listed (see screenshot)
When I select region USA I get the mario usa set with checksum eb26155d...

So everything works fine here and I don't see any bug.

As mentioned before, Logiqx' 1G1R algorithm (which uses a scoring algorithm) is unchanged for over a decade and the results highly depend on a) your enabled regions/languages, plus b) the order of the enabled regions/languages, c) the original datfile and its "release" elements....and sometimes users simply have a totally different understanding what 1G1R is all about (how the setnaming and set selection works).

In your Python code I don't see Logiqx' algorithm implemented but surely you can point me to it in your code....
I guess part of Logiqx' code can be found on old forum posts like https://forum.no-intro.org/viewtopic.php?f=2&t=544

I'll dedicate some more time to this, but meanwhile, what I did was to follow the procedures from here:
https://forums.launchbox-app.com/topic/40303-rom-filtering-guide-for-clrmamepro-filter-all-usa-games-plus-all-exclusives-from-all-regions/

And I wasn't the only user who reported it excluding some games. Well, maybe the problem is in the steps in this link, then? I tried a number of other things but I failed to make it produce a set that included Mario Kart DS at all.

And I didn't implement his algorithm. I made my own (basically, it's also a scoring system and sorting a list of candidates based on it).
I'll check his algorithm and see where they differ, too.

So I made a quick test....downloaded the Nintendo - Nintendo DS (Decrypted) (Parent-Clone) (20200126-032320).dat with that DAT-o-matic, loaded it in cmpro, enabled the 1G1R option, enabled EUR as region and I get Mario Kart DS (Europe) (En,Fr,De,Es,It) listed (see screenshot)
When I select region USA I get the mario usa set with checksum eb26155d...

So everything works fine here and I don't see any bug.

As mentioned before, Logiqx' 1G1R algorithm (which uses a scoring algorithm) is unchanged for over a decade and the results highly depend on a) your enabled regions/languages, plus b) the order of the enabled regions/languages, c) the original datfile and its "release" elements....and sometimes users simply have a totally different understanding what 1G1R is all about (how the setnaming and set selection works).

In your Python code I don't see Logiqx' algorithm implemented but surely you can point me to it in your code....
I guess part of Logiqx' code can be found on old forum posts like https://forum.no-intro.org/viewtopic.php?f=2&t=544

The 1G1R algorithm hasn't changed since 2009.

If you face issues with the selection it's usually simply based on the used datfile or a misunderstanding what 1G1R actually does.
So if you're sure that Mario Kart DS should be selected, send the datfile and your settings and I'm pretty sure I can tell you what is wrong on your side or in the dat.

Hi there,

I have tried to generate a 1G1R set for the Nintendo DS using DAT-o-matic's Parent/Clone XML (with default parent data) and I observed a few issues with it.

ROMs like "Mario Kart DS" would be simply missing, despite being available for all regions I selected (USA and EUR, in that order), among others. I see the full data for them in the DAT, including region data, but the ROM file is simply never picked up for selection.

For GBA, also, a number of Animal Crossing titles were missing, as well as a few other ones here and there.

Again, I checked the DAT and all the Parent/Clone data was there for those games.

I'll generate a better bug report with more detailed information, as all that happened a few weeks ago.

I created a small Python application that generates what the ROM selection is supposed to be like, from a Parent/Clone DAT, as a means of validating the selection clrmamepro makes (as well as creating my own tool for that because I wanted the sets immediately back then, instead of reporting the bug and letting someone fix it, wait, etc )

You can find it here: https://github.com/andrebrait/1g1r-romset-generator

you mean for example a51mxr3k for area51mx? a51mxr3k is a clone of area51mx which uses the same chd (but different roms).

Thanks Roman, that worked.  I unbound the systems and re-ran the full scan.  Did not get the renaming options prompt this time around, so things look good to go.
What were those names that it was suggesting? They weren't the setname and they weren't the chd names, something internal?

Using sysdefpaths is only possible if you split your sets by (bios/software lists) systems (e.g. neogeo, naomi, etc).
Having a chd folder means that you have a rompath setup which uses the chd folder ("one folder") for multiple systems, so sysdefpaths won't work.
But of course your idea of keeping chds in one folder is fully valid.
So actually (as long as you don't use software lists), you should turn off the sysdefpath usage. This can be done by accessing the sysdefpaths (via "systems") and remove (unbind) all.

After that you should run a full scan again to see if this already resolves your rename issues, too. If not, you should send me your cmpro.ini, the *.cmp file from the cmpro settings folder so I can try to rebuild your scenario.