>>"Inescapable"? Unaccelerated Ignition, WC4 and WinQuake are
>>faster in Win than DOS.
>No, they are not.
Yes they are. Maybe you should try them out yourself sometime? I have.
The fact that WinQuake is definitely faster than Dos Quake has been
reported all over the net, it is not just me either.
With current machines (P133+) this extra load is insignificant, causing
like 1 fps slowdown. Whoopee! And sometimes the Win95 versions of games
(like Quake, Wing Commander 4, Ignition and UAE Amiga emulator) are
even faster than their MS-DOS counterparts.
Or maybe not, considering controller problems in MS-DOS games are
daily. Screamer Rally is a recent example, it fails to work even with
normal 2 axis analog joysticks, let alone some T2.
DirectInput already does damn good job with a SINGLE joystick port on a
regular sound card. For example: eight buttons, throttle, rudder, eight
way hat, all this without any restrictions at all.
Is that why such MS-DOS game like Screamer Rally forces people to play
it with a keyboard? A driving game? I have seen much more DOS
controller failures so far.
It certainly does change the way that single joystick port on my sound
card can (not "must") be utilized. Much more efficient than the MS-DOS
controllers with their button restrictions (CH Flightstick Pro is
especially so poor in this), big latency, vibration when you are not
even touching the joystick etc.
DirectInput is technically quite superior to any controller support we
have seen in MS-DOS games to date. Show me one MS-DOS game or
controller, where you can use a 2 axis analog joystick, rudder,
throttle, eight buttons and EIGHT way hat, *** all simultaneously
without any restrictions whatsoever, using only _one_ normal SB16 game
port ***. DirectInput has handled that for quite some time already, and
has less latency and better accuracy than any MS-DOS controller in any
MS-DOS game too.
Too bad you have bought peripherals which are not supported by the
hardware manufacturer. I see the mere way these older MS-DOS
controllers work as a hardware defect.