@cyclair Probably not enough. The link clearly asks for some logs and config files so that we aren't pulling things out of thin air. To be clear, with pretty much thousands of people running and playing vertical games on the official image, especially on the RPi3B that's been around for quite a while, I would vouch for the fact that what you're facing is pretty much configuration issues from that exact image.
But do share more details per the link shared above - type of screen you're using, etc.
@mitu Thank you. I had read that guide several times. I can’t believe I overlooked that. I rebuilt a full non merged set with clrmamepro and then used the same batchfile I mentioned earlier. Now it works.
@m2306 - I was thinking about this over the weekend and there might be another way but requires a bit more work. I don't think it would be to much really and only initially. If you're set and good to go then no worries.
Take CPS games for example. Basically if you had a list of parent ROM names in a text file with one per line like so:
With a list like this it would be pretty easy to pull parent ROMs only or parent + clones.
The advantage being is you only need to track parent ROMs and not clones as they are automatically linked to the parent.
So basically by listing the 70 or 80 CPS games you could generate new DATs after an update with very little changes after the initial setup.
I don't believe parent ROMs change to often and if they do it would be only a matter of changing the name in the list like sf2 to sfi2 for example. The clones will be automatically linked.
The clones make up a much larger part of the games list by several hundred so it's nice that you only need to track the parent.
I think a lot of filter lists and batch files that copy ROMs are based of specific DATs in time and are outdated with the next release whereas this would allow you to be a bit more dynamic with new releases and generate a current correct ROM set.
You could also stack lists to make a master list. Like combine the CPS and Neo Geo list or CPS, Neo Geo plus favorites, etc.
I'm not sure of that makes sense but I believe it would work fairly well.
Im hoping the pi 4 has better performance with run ahead
Supposedly it has 2x-3x the performance of a pi3, so yeah i guess it will help a lot, but i own an odroid n2 which clearly outperform the pi4 (more recent cpu with higher frequency) and it's still unable to run some games full speed iirc (and i'm not talking about the special case of killer instinct)
@mitu - Not a big deal but shouldn't Core Providedfill the screen on the x or y axis but not necessarily Integer Scaling which would be more of a whole number multiple of the original resolution less than the display resolution or am I reading that wrong?
Yes. Some games you can manipulate by changing setting to 2 coins/1 credit if the game in question has that option, some do not. I have tried re-installing mame 2010/ 2003 but same outcome. Will probably do a fresh install now that the pi-4 is out anyway. Thanks for your input.
Good in theory, but AFAIK the retroarch devs have no intention of implementing said feature in the short to medium timeframe. lr-fbalpha having it would be a lifesaver, there are a lot of 2, 3 and 4 buttons per player arcade cabinets out there that badly need it. What's the last lr-fbalpha version that fully supports it?
.44 is synced with current mame though a "non-merged" romset would have more roms
Careful with that statement, it's mostly true but there are exceptions, as an example i know a lot of pgm games won't work if you use a mame romset without converting it to a fbalpha one through clrmamepro (every few months i have people who report issues saying "pgm games don't work" thanks to that). FBAlpha also includes a lot of hacks from hbmame, some are fun.
@chubsta Sorry, i don't provide "no-clone" dat files, and i don't recommend it either :
a few parents aren't working and are there only to get their clones working
idk about 1943 specifically, but some games are different from a clone to another, i think that's especially true for a few shmup where you get different foes/pattern, i noticed that while working on a bug with raiden2 last year.