nanogui: Thread: Re: Generic Input Event question


[<<] [<] Page 1 of 1 [>] [>>]
Subject: RE: [nanogui] Generic Input Event question
From: Junior ####@####.####
Date: 10 Jun 2008 20:40:09 -0000
Message-Id: <90E3CDE00B0.000005F7ejr@inbox.com>

Could I assume no one has ever lost an input events (once it is delivered by the driver) regardless of the input driver (i.e just me)?.

Jr.

> -----Original Message-----
> From: ####@####.####
> Sent: Mon, 9 Jun 2008 11:22:29 -0800
> To: ####@####.####
> Subject: [nanogui] Generic Input Event question
> 
> Hi All,
> I've been periodically tracking a problem I've been having for some time
> now but haven't been able to zero in on it.
> I've been using tslib with nanoX for well over a year now but sometimes
> it is as though nanoX ignores some of the touch events.
> It is not consistent so putting a handle on it has been difficult. Hence,
> I'm not looking for a solutions but rather some
> thoughts on what be a better place to track it. I have some debug
> statements in the driver that prints me every input event
> occured,and although I see every event for some reason getnextevent will
> not pick it up. It is as though nano-X dumped it in the bit bucket.
> Any ideas/thoughts?
> 
> Addition: I have not been able to determine what filter_transform really
> does. It doesn't seem to be doing
> anything useful after looking at nxtransform.c. Is this input device
> specific code?
> To what extent can the driver handle jitter(devmouce.c)?
> 
> Thanks,
> Jr.
> 
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: ####@####.####
> For additional commands, e-mail: ####@####.####
Subject: Re: [nanogui] Generic Input Event question
From: "Greg Haerr" ####@####.####
Date: 10 Jun 2008 23:27:56 -0000
Message-Id: <013901c8cb51$d86e5cc0$6401a8c0@winXP>

> Could I assume no one has ever lost an input events (once it is delivered 
> by the driver) regardless of the input driver (i.e just me)?.

There are three distinct layers for mouse events.  They start
with the driver, which directly interfaces with the kernel,
using either a select (preferred), or polling interface with
the nano-X main loop.  These are in drivers/mou*.c.
The event is then passed to the engine layer, which
handles system wide transformations (all the transforms
and filters should be in this layer, but aren't due to hacking).
The engine layer determines whether which "event" is to
be generated (for either win32 or nano-X APIs), and
will sometimes add information (like which buttons where down
during mouse move, etc).  This is engine/devmouse.c.
The upper API layer is then called which handles
API system-specific activities for the event (in nano-X
this is usually nanox/srvevent.c and nanox/srvutil.c).

So there can be many areas to look when talking about
"dropped" events.

Regards,

Greg 

[<<] [<] Page 1 of 1 [>] [>>]


Powered by ezmlm-browse 0.20.