without a lightgun shooters in mame are pretty hit or miss and most of the later era ones need a real beast of a comp to even play at a decent frame rate ... so my guess would be yea but i wouldnt bet the experience would be the best..gnargoyle11 said:sorry, havent messed with mame emulation much in the past, so im really sorry if this is a stupid question:
will this be able to run shooters like house of the dead?
With a touchscreen, playability should improve.mmalficia said:without a lightgun shooters in mame are pretty hit or miss
GunPei2X said:With a touchscreen, playability should improve.mmalficia said:without a lightgun shooters in mame are pretty hit or miss
[written by someone who duct taped a plastic Uzi to a joystick (and switched its up/down inputs) to improve the Op Wolf "experience" on a C64...]
grin i actually have a big touch screen on my cab while its better there can be some response lag not sure if thats mame or the hardware thou .. that being said my preferred method before i got my guns was to use a modded happs trackball .. my other deciding factor was with a big touchscreen like that it never fails one a your pissed up Friends/family starts wailing away on the screen after the fighting gets intense and sorry for as much $$$ as ive got invested in my baby thats a nono :>
If we're given a decent framebuffer to write to, then SDLMAME should be relatively easy to port and run quickly. The killers for performance are the artwork, overlays and any scaling done. Spitting the output into an 1:1 resolution without processing it in any way massively improves performance.Dunny said:Not sure about using the latest MAME sources (or at least, porting them to Pandora). It might be a better idea to develop using multiple versions as a reference. A while ago, the MAME team removed all the tricks used to get decent framerates, with the goal of "documenting" the hardware rather than games being playable.
As a result, sound samples were abandoned in favour of emulation - caused a huge FPS hit in a lot of games, with no real noticeable benefit to the player. Stuff like the Voodoo hardware emulation is purely software, as allowing your PC's 3D hardware to take the brunt of the work was frowned upon. As Slaanesh says, there were a number of other issues which caused some awful slowdowns (but increased accuracy in a small number of cases).
Unfortunately that's not the case for most emulators, and MAME is probably the worst for it of them all. In general, as emulators progress they get more accurate - accuracy is probably more important than speed in environments where CPU power isn't much of a factor. As accuracy improves, the core usually gets slower. I recall in the days of JPP where a spectrum emulation would run in a few mhz (slowest I've seen is 14mhz) but that was with very poor emulation - nowadays you need a minimum of about 200 - 300mhz to emulate a Speccy at all well. Most games will play in a bog-standard "runs on a 486" emulator, but most demoscene productions will fail hard.joshwaan2k said:So as they go up in version some games get slower hey?
I thought it be the other way around people would tweak driver set to make things faster.
Dunny said:Unfortunately that's not the case for most emulators, and MAME is probably the worst for it of them all.joshwaan2k said:So as they go up in version some games get slower hey?
I thought it be the other way around people would tweak driver set to make things faster.
Exactly! This is why going to the latest version of MAME is in my opinion not really desirable. The focus on total accuracy was not as strict in the earlier days of MAME development. This is evident in the source code as there are plenty of hacks/shortcuts/speedups. I just think we need to find the suitable version of MAME for Pandora. I will starting doing some speed tests of various versions.
slaanesh said:Dunny said:Unfortunately that's not the case for most emulators, and MAME is probably the worst for it of them all.joshwaan2k said:So as they go up in version some games get slower hey?
I thought it be the other way around people would tweak driver set to make things faster.
Exactly! This is why going to the latest version of MAME is in my opinion not really desirable. The focus on total accuracy was not as strict in the earlier days of MAME development. This is evident in the source code as there are plenty of hacks/shortcuts/speedups. I just think we need to find the suitable version of MAME for Pandora. I will starting doing some speed tests of various versions.
http://benchmark.mameworld.net/
its a bit of excessive information to push through, but still useful.
Go for it! Unless the Pandora is much more powerful than I thought, it sounds like going for the most recent version would be a waste.slaanesh said:I just think we need to find the suitable version of MAME for Pandora. I will starting doing some speed tests of various versions.
PoisonedV said:http://benchmark.mameworld.net/
its a bit of excessive information to push through, but still useful.
This is interesting but unfortunately uses an inconsistent means of running the tests. ie. Because they are use submitted results, the hardware and environment that was used to conduct the tests will vary enormously.
It's possibly useful for seeing if a particular game is playable for a particular version of MAME assuming that you have similar spec hardware to whoever did the original test.
I'm going to setup a more rigorous test where I use several versions of MAME running games on the same hardware.
If you are using 0.53, you may as well use 0.62... newer and very similar performance...joshwaan2k said:.53 looks like a good version to use