Bingo Party Multicart (Rev B) (M1 Satellite board) [zozo]
this is one of the satellite boards from a much larger system, unfortunately no other parts of it are dumped. We have bingoc.c but I think that's a different part of a similar system (possibly one of the 'link' boards but for an earlier) we also have 'pontoon' in segas18.c which is likely to be the video part of a similar system, but again there we're missing all the other boards. Finally we also have Royal Ascot (X-Board) which could also be a component of one of these setups (we know Caribbean Boule which has M1 satellite boards uses an X-Board for the communal display)
maybe one day we can actually get hold of a complete set of PCBs for one of these things....
In both Windows and SDL builds, format of -resolution can now be any of:
<width>x<height>
<width>x<height>x<depth>
<width>x<height>x<depth>@<refresh>
<width>x<height>@<refresh>
Omitted values are defaulted to 0, at the point of parsing (as before).
output [Fabio Priuli]
out of whatsnew:
I fear no proper and clean solution is possible as long as the
driver has such a privileged role in the MAME core structure,
because stuff that is not attached to a root_device (=driver)
is hard to parse. Anyway, there should be no more missing
roms from -lx output (until people starts adding them as
sub-sub-subdevices... ;) )
-showusage says, and the Windows version looks for,
"<width>x<height>[@<refreshrate>]", but the SDL version was scanning for
"<width>x<height>[x<depth>][@<refreshrate>]", with the effect of
silently ignoring refreshrate if depth was omitted. And if given, the
depth didn't appear to be used anywhere anyway.
ist certainly more modern than Gaussian elimination and Gauss-Seidel.
However, more the current maximum matrix (KidNiki, 89x89) a combination
of Gauss-Seidel to solve for maximum one step to catch quasi-stable
conditions and fall-back to optimized Gaussian elimination (for sparse
matrix) outperforms GMRES by up to 100%. [Couriersud]
