Wiki editing discussion

Page 51/65
44 | 45 | 46 | 47 | 48 | 49 | 50 | | 52 | 53 | 54 | 55 | 56

By sd_snatcher

Prophet (3246)

sd_snatcher's picture

18-12-2019, 21:00

gdx wrote:

My JoyMega is made with a transistor. I know there may be problems with OCMs / Zemmix but I did not note any problem on my 3 MSXs (MSX1, MSX2 and Turbo R). Which MSXs does this cause problems with?

During the development of the HIDlib, I tested with a plethora of different adapters, controllers and model combinations. And I remember that the JoyMegas made with transistors had problema with many combinations, including the original Sega Japanese 6-button controller.

Quote:

is Sega Multitap compatible with the Ninja Tap or only usable for specific programs?

No, they’re totally different devices, with specific protocols.

HIDlib doesn’t support the NinjaTap because back then (1) I couldn’t find any information about its protocol, and (2) I don’t have this device for testing.

By Wild_Penguin

Hero (590)

Wild_Penguin's picture

19-12-2019, 09:29

sd_snatcher wrote:
gdx wrote:

My JoyMega is made with a transistor. I know there may be problems with OCMs / Zemmix but I did not note any problem on my 3 MSXs (MSX1, MSX2 and Turbo R). Which MSXs does this cause problems with?

During the development of the HIDlib, I tested with a plethora of different adapters, controllers and model combinations. And I remember that the JoyMegas made with transistors had problema with many combinations, including the original Sega Japanese 6-button controller.

I think this has something to do with the timing (The transistor is not fast enough?). [EDIT: It may be bouncing, however, I can not stress enough the following:] However I'm not electronic expert, so I could be totally wrong on this timing theory / misremember / misunderstood the discussion. This has been discussed to death elsewhere on this forum.

I can confirm I get spurious input just sometimes when using a transistor-based JoyMega (I bought on eBay) but the inverter-based (7402 or similar, I've used two different chips) work much better, on all MSXes I own (once I figured the cheap PADs I were working on had problems of their own). EDIT: MSXes I own include three MSX1:s, two MSX2:s and the Zemmix Neo; though I must admit I didn't do conclusive systematic testing with the adapters and gamepads, just got a rough feeling with games any joytest.

By gdx

Prophet (3320)

gdx's picture

19-12-2019, 09:30

sd_snatcher wrote:

Zero lag

Do you know an adapter that causes perceptible lags?
Micro controllers react in a few nanoseconds. That can have an impact only on very very rare games badly programmed which would change the state of pin 8 too often (eg Choro Q). Only wireless controllers with bad transmission can cause perceptible lags.

By sd_snatcher

Prophet (3246)

sd_snatcher's picture

22-12-2019, 03:37

Lags can be caused by many factors, so it all depends on the design.

But a middleware software translator will very rarely be able to avoid the lag problem, no matter how fast the MCU is, just because of the constraints involved.

For example, Nintendo and Playstation controllers use serial protocols. This means it’s not possible to read all buttons+directions directly at the time that the Z80 reads the buttons in parallel via General Purpose ports, as the serial protocol is much slower. This means that the adapter has to latch the pad buttons from time to time, and depending on how out-of sync the two processes (on the MSX and on the adapter) go, you could be pressing the button at the moment that the MSX performed the read, but it wasn’t still latched yet by the adapter. One entire frame will be lost (16.6ms on NTSC) until that info is finally latched by the adapter and then read by the MSX.

Sure, for those Micro Cabin games this won't make any difference. But on Aleste 2 and Space Manbow, a frame can make the difference of being alive or not.

By gdx

Prophet (3320)

gdx's picture

22-12-2019, 09:49

sd_snatcher wrote:

One entire frame will be lost (16.6ms on NTSC) until that info is finally latched by the adapter and then read by the MSX.

Yes but Microcontrollers instructions take 50~100 times lower than 1ms. For example, in the case of the SNES controller, I think we can read the state of the 16 bits several times during a frame so no frame should be skipped if the program is well done and if the Microcontroller is well chosen.

By sd_snatcher

Prophet (3246)

sd_snatcher's picture

22-12-2019, 14:24

Yes, as I mentioned, it depends on many factors, including the protocols chosen. For example, bluetooth adapters tend to have a much higher lag than pure RF transmitters. And the more adapters you daisy chain, the more the lag will sum up.

With a JoyMega adapter the lag is always zero. This means that you can connect a wireless Megadrive adapter to it, and then total lag will be exactly the same as a Megadrive would have. This is not the case for the Nintendo and Playstation adapters.

By gdx

Prophet (3320)

gdx's picture

23-12-2019, 01:50

sd_snatcher wrote:

With a JoyMega adapter the lag is always zero. This means that you can connect a wireless Megadrive adapter to it, and then total lag will be exactly the same as a Megadrive would have. This is not the case for the Nintendo and Playstation adapters.

It's wrong. As I mentioned above for wireless adapter, it depends on the transmission quality. I use it with a JoySNES and I noticed lags only when battery is low or when I go too far. Also avoid placing the transmitter on the ground. About bluetooth, yes lags come more easily. Bluetooth adapters are not ideal even on PC.

By gdx

Prophet (3320)

gdx's picture

28-12-2019, 11:36

About 8bit4ever and RBSC Carnivore2, devices in the wiki are not classified by developer but by the maker/vendor.
So 8bit4ever and RBSC Carnivore2 should either have their own page or a general page without 8bit4ever or RBSC in the title.

By tfh

Prophet (2068)

tfh's picture

29-12-2019, 14:59

Hi... Maybe some stupid questions from my side, but I was wondering about the following:

* Why are the groups/persons a "catagory". It also looks weird, mentioning catagory in front of every entry.
* The yellow links are not that easy to read:

By mars2000you

Enlighted (5564)

mars2000you's picture

29-12-2019, 16:32

It's how the Wiki can function in an optimal and flexible way. All Fony products are for example in the Fony category. Without that, then it's impossible to connect the Fony products together.

Maybe it's less obvious when it comes to specific persons but it's difficult to define and use there a perfect general rule, especially for several reasons:

* some groups are actually only one person (the most obvious case is Xelasoft) - it's the main reason why persons and groups have been merged in an unique big category, and it's the equivalent for the MSX scene of the category MSX Companies for the official MSX commercial side
* some persons want to have a page about all their work in the MSX scene and many persons have worked in more than one group - if i take John van Poelgeest as example, the Wiki page about him, currently linked only to the Groups/Persons category, could also be linked to other categories: Federation Against Commodore and First Class Software - the mentioned merging has also be thought in this perspective, if such extended linking seems to be requested in the future

I agree about the yellow links, but it's something that only the admins can change. Btw, gdx had also asked that the direct links Scene Members and Scene Groups would be merged in the Wiki submenu, but the admins don't have changed that until now.

Page 51/65
44 | 45 | 46 | 47 | 48 | 49 | 50 | | 52 | 53 | 54 | 55 | 56