Is this core on development or shall I put my hopes down for new features?
Would 
to see MIDI, save states and rewind-support in newer updates 
just keep going with the suggestions, probably post in github page too. someone might pick it up. a bounty would also help attract some more help
with the latest version of the core i canāt clear stage 2 in akumajou dracula, everytime i reach the part with the raft it disappears out from under me as soon as i strike the lion head statue. anyone else experiencing this problem? iāve played through the entire game before sometime within the past year so i know it worked properly at some point, i may try and git-bisect it later when i have the time but iād like to verify the problem exists for others before i waste any time looking into it.
the problem shouldnāt be related to either the bios or game files as iāve verified both the bios roms (iplrom.dat, cgrom.dat, and iplrom30.dat) and the game images (dimās) all match hashes from both MAME and TOSEC. i even tried an xdf version, it made no difference.
dont think the core has been updated since apr14. was this section of the game worked before? what version are you using? i mean is this an fdd dump or hdr
when i said the past year i meanāt within the past 365 days or so from today, not within just 2018. sorry if that wasnāt clear. Iām using floppy images, as stated Iām only using known good dumps, however just loading the games with px68k does modify the image files, though every time Iāve checked for the bug Iāve recreated them from fresh copies.
let me play this quickly. i do remember getting past this raft
@e-tank hmm weirdā¦ im sure i had no issues in this area before. i managed to work it once, and then each succeding tries this raft disappearsā¦
1 Like
although i failed to mention it i also was able to get it to work once as well on the latest version, but like 9/10 tries (far more often than not) it doesnāt appear and you just fall to your death. thanks for the verification, as i said Iāll try and run a git-bisect (when i have time) between the last commit and one from 365 days ago
I had it happen repeatedly on android a while back, then it didnāt happen and I could go stage 3.
I tried it again on a windows PC and it wasnāt happening at all.
So itās a bit random, not sure if itās platform specific.
looks like a problem with the cpu used. i managed to find my old repo of px68k with the original cpu it came with and it worked fine 10/10.
what platform are you using? i can compile for windows and linux 64bit maybe you check.
EDIT: well, i just compile new core for linux/win x64 maybe you can try if this does work better or notā¦
the machine iām testing on is fedora 27, so vanilla linux x86_64 + glibc, nothing fancy or out of the ordinary.
iāve done a lot of testing and i believe itās due to cpu timing issues. iāve found the raft is less likely to disappear the higher the cpu clock speed is set. at 10 mhz with the current 68k cpu core (or even the version from the initial libretro commit), in my tests the raft disappears more often than not. with the core you posted at 10 mhz i got the raft to disappear only once out of about 10 attempts. perhaps at this point in time it would be a good idea to change the default cpu speed of the core to 16 mhz from 10, at least until improvements can be made to the accuracy of its timings.
also, for anyone who wants to test this iād suggest racking up points for extra lives on the first stage via killing loads of fleamen behind the breakable wall, preferably with holy water for huge multiplier bonus. then at the raft kill yourself each time it appears successfully to try again in order to help determine your rate of success.
interesting find. iāve always run at 10Mhz as increasing it past that causes some minor issue especially for audio before when i was tinkering with core. It does work at 16Mhz though. good thing its not a big jump in Mhz.
No, donāt change the default 10MHz cpu speed as it will make plenty of games run too fast.
at the very least an issue on its github page should probably be opened outlining the problem in akumajou dracula, perhaps with a note that it may be due to cpu and/or memory access clock cycle timings being inaccurate (too slow compared to real hardware) in order to run the game properly at default speed, and advise users to increase the clock speed manually to side step the problem.
also, iād like to suggest an enhancement to the core here: rearrange the main loop so that vblank is handled first in the emulated frame slice (as iām pretty certain the code is currently organized in the opposite way, active first then vblank), which would of course reduce latency by 1 frame in many games. if thereās anyone here who has the time and interest in taking on such a project but doesnāt know where to begin iād be happy to provide links to previous commits on other emulator cores where this kind of code reorganization has been done before, which should be useful as a general reference on how one goes about solving this type of problem.
reducing 1 frame should be easy enough hereā¦
Hello everyone. I was surprised by this core because of how easy and good it is, I have tried several games and it is going very well but there is one that I do not know what to do to make it work:
Super Street Bomber X (19xx)(Dragon Studio)[Req Install].dim
When I load this game I get next message about Human.sys:
The file game name said āReq Installā, Must I install it? How can I do?
Thanks.
Probably you need to create a harddrive image and:
- Grab the Human68k (the operating system) install floppy and install it to the harddrive image
- Remove the floppy and boot into the harddrive image
- Insert Super Street Bomber X into the floppy drive and install it to the harddrive image
- Remove the floppy, then navigate to where you installed the game and run it
1 Like
having a heckuva time getting Akumajou Dracula to run. Ive got it seeing both discs, but it looks like it wants to install? is there a proper way of creating an HDD0 file?
Thanks!
Is this core still crashing on MacOS? Iām running Retroarch 1.7.5 since itās the latest working build with metal support. Whenever I try to load content with this core Retroarch crashes.
I just installed Retroarch 1.7.7 on macOS 10.14.5 to try Px68k core and it crashed.
EDIT: turns out I had to put BIOS files in ~/Documents/RetroArch/system/keropi then it booted up with no crash.
1 Like
Iām having problems with this core not starting any game since I installed Retroarch 1.7.8. Tested it on Windows and a Raspberry Pi. Bios are correct, game booted before. Can anyone confirm this, please?