Kimplul

> If I apply this with 0x0005, the Bootloader becomes active. This generally means that the firmware on the wheel crashed, which I guess I'm not entirely surprised about. If...

Appreciate the info, we saw something along those lines with the T128 over in #67, though the init method was for whatever reason not the same and `hid-tminit` had to...

I pushed cafd0f46535df3674098754c6c693cd34274810c that got FFB working in *Speed Dreams*, could you check if it also fixes your problem? The T128 is still in another branch so you'll probably have...

Kind of interesting that the FEdit program I used to do most of the reverse engineering never even touched the saturation. I had a cursory glance at code, looks pretty...

> It can be made even simpler if we only use the 0x0e message type, which results in losing the equivalent behavior with the Windows driver and adds an extra...

> I'm currently wondering about why does the duration updated with every update request, even it is unchanged Heh, now I'm a bit embarrassed. I thought that re-sending the duration...

> Out of curiosity, do you know any driver that does the ff-memless way of the duration update and does not depend on ff-memless? Logitech wheels are the only ones...

From the LKML thread at https://lore.kernel.org/all/[email protected]/ > As you've seen, the effect model should very closely match the DInput model (as that is what the devices tend to support), mostly...

> He was talking about the effect model, which is pretty close already, while this discussion was about semantics, or so I understand. I thought a model defines the semantics,...

> What we are speaking about is not just one parameter of the effect, it's all the parameters that changes with time like delay, attack length/force, fade length/force, ramp start/end...