Maarten ter Huurne
Maarten ter Huurne
Cross-compilation from Linux using mingw-w64 does work, but I've heard that native compilation can run into problems when trying to build pkg-config. Note that the easiest way to build is...
Is this true for all user files that we write? Things like settings and ROM pool caches should stay on the machine that wrote them. But for example save states...
As far as I know, 7MHz conversions included some circuitry to add wait cycles after I/O to prevent too-fast access. So unless we emulate that as well, `z80_freq_locked` is not...
I think it would make sense to not persist the value of `z80_freq_lock`.
I have some reservations about using random numbers. For TAS replays and network play (hopefully some time in the future) we need deterministic randomness. But if the randomness is deterministic,...
> Returning 00H would be better than FFH, I would be ok with that solution. I guess we should have pulled-up, floating and pulled-down at the XML level and then...
Our current host MIDI event timing is not equal to the MSX timing, see [this comment](https://github.com/openMSX/openMSX/blob/39f53618/src/serial/MidiOutCoreMIDI.cc#L22). The current code will just send out messages as quick as possible, which combined...
For some other hardware, we have Tcl hooks that print a warning if they're accessed too fast or in another invalid way (like the PSG I/O directions being set wrong)....
> I've given up on option 2. Instead of pitch correction, would it be an option to cut fragments out of the sound? In other words, if we're at 125%...
From the error message, it looks that the patching itself is not the problem, but the order in which mapper detection and patching are done, as Manuel suggested. @MBilderbeek: If...