Menu
▾
▴
linuxconsole-dev
|
From: Adam G. <ad...@ev...> - 2001-10-05 00:45:18
|
I started my SF page with the intent to make 2 items:
1. The low level kernel drivers,
2. the higher level userspace libraries and integration with Qt/GTK+.
Part 1 more rightfully belongs in the linuxconsole project, I think. They
have done an excellent job so far with all things USB, and I think they
will have something excellent for 2.5. They also have a force feedback
framework in place.
Part 2 doesn't exist yet, because we don't have Part 1... ;)
For now, it makes sense to focus resources on linuxconsole, because
that is where all the good stuff will happen. I looked at doing this,
but was somewhat stopped by two things:
a) The lack of a good way to send interrupt messages to USB devices
from userspace (like libusb, but someone suggested this isn't too
hard to fix),
b) the somewhat more troubling problem of where to fit the iFeel driver.
The iFeel mouse is a regular old HID device, and I think that the
HID driver does a wonderful job in controlling it. But the iFeel has
an extra interrupt endpoint which takes the force feedback commands,
and I wasn't sure the best way to put this nonstandard extension
into the current USB framework without rewriting or duplicating a
lot of code nicely implemented in the HID driver.
Problem a is not a huge problem, it just makes it easier to test and
develop the lowest level stuff.
Problem b is a more complicated issue, and I think this needs to be resolved
before we can move forward. I don't think it's a huge issue, but it needs
to be addressed. The place to do it is with the linuxconsole folks...
Yay, verbosity. Anyway, I hope that this gives you a better idea of
where I'm at, and what needs to get done before we get some progress.
The actual commands to send to the mouse are dirt simple, we just
need to get them there smoothly... :)
For now, my 'tactile' SF project is sleeping. Once we get the basic kernel
stuff going, I think it would be a good place to work on the higher level
stuff, perhaps in a more device independent way. (That's why I named it
'tactile' and not 'ifeel-linux'.)
Regards,
Adam
On Thu, Oct 04, 2001 at 09:03:36PM +0200, Johann Deneux wrote:
> On Wed, 3 Oct 2001, Adam Goode wrote:
>
> > Not really, the force feedback stuff has to go into the new Ruby
> > Linuxconsole code, and I haven't had a lot of time to hack around with
> > it...
> >
> > Maybe this weekend... ha!
> >
>
> It seems you opened a project on sourceforge dedicated to this device,
> btw. As it does not look very active, do you expect to use it in the
> future, or do you intend to use the linuxconsole project ? I am asking
> that, because I sometimes receive questions related to force feedback and
> IFeel mice. Should I redirect people to the linux-usb-dev ML, the
> linuxconsole one, or your project page on sourceforge ?
>
> Did you manage to get any information about the protocol itself ?
>
> --
> Johann Deneux
>
>
|
|
From: James S. <jsi...@tr...> - 2001-10-05 18:19:42
|
> Part 2 doesn't exist yet, because we don't have Part 1... ;) Well we pretty much do. > a) The lack of a good way to send interrupt messages to USB devices > from userspace (like libusb, but someone suggested this isn't too > hard to fix), By interrupt messages what do you mean exactly. > b) the somewhat more troubling problem of where to fit the iFeel driver. > The iFeel mouse is a regular old HID device, and I think that the > HID driver does a wonderful job in controlling it. But the iFeel has > an extra interrupt endpoint which takes the force feedback commands, > and I wasn't sure the best way to put this nonstandard extension > into the current USB framework without rewriting or duplicating a > lot of code nicely implemented in the HID driver. > Problem b is a more complicated issue, and I think this needs to be resolved > before we can move forward. I don't think it's a huge issue, but it needs > to be addressed. The place to do it is with the linuxconsole folks... Hm. Do you have documentation on the iFeel protocol to how we can fit things together? |
|
From: Johann D. <jo...@Do...> - 2001-10-05 21:39:45
|
On Fri, 5 Oct 2001, James Simmons wrote: > > > Part 2 doesn't exist yet, because we don't have Part 1... ;) > > Well we pretty much do. Indeed. At least, we have an API that should not be to hard to extend for mice support. Of course, we also need a driver behind the API ! > > > a) The lack of a good way to send interrupt messages to USB devices > > from userspace (like libusb, but someone suggested this isn't too > > hard to fix), > > By interrupt messages what do you mean exactly. > > > b) the somewhat more troubling problem of where to fit the iFeel driver. > > The iFeel mouse is a regular old HID device, and I think that the > > HID driver does a wonderful job in controlling it. But the iFeel has > > an extra interrupt endpoint which takes the force feedback commands, > > and I wasn't sure the best way to put this nonstandard extension > > into the current USB framework without rewriting or duplicating a > > lot of code nicely implemented in the HID driver. I guess you could simply extend the existing hid driver. There are callbacks in the input_dev structure, which are currently not used in hid. These callbacks are: * upload_effect to upload a new effect to the device or update an existing one * erase_effect to suppress an effect * event is used to play and stop effects > > > Problem b is a more complicated issue, and I think this needs to be resolved > > before we can move forward. I don't think it's a huge issue, but it needs > > to be addressed. The place to do it is with the linuxconsole folks... > > Hm. Do you have documentation on the iFeel protocol to how we can fit > things together? > I guess Adam knows pretty well the protocol. Anyway, I found this: http://moore.cx/out/ifeel/ -- Johann Deneux |
|
From: Vojtech P. <vo...@su...> - 2001-10-07 23:15:55
|
On Thu, Oct 04, 2001 at 08:44:32PM -0400, Adam Goode wrote: > I started my SF page with the intent to make 2 items: > > 1. The low level kernel drivers, > > 2. the higher level userspace libraries and integration with Qt/GTK+. > > > Part 1 more rightfully belongs in the linuxconsole project, I think. They > have done an excellent job so far with all things USB, and I think they > will have something excellent for 2.5. They also have a force feedback > framework in place. > > Part 2 doesn't exist yet, because we don't have Part 1... ;) > > For now, it makes sense to focus resources on linuxconsole, because > that is where all the good stuff will happen. I looked at doing this, > but was somewhat stopped by two things: > > a) The lack of a good way to send interrupt messages to USB devices > from userspace (like libusb, but someone suggested this isn't too > hard to fix), You don't need this, if you do b) > b) the somewhat more troubling problem of where to fit the iFeel driver. > The iFeel mouse is a regular old HID device, and I think that the > HID driver does a wonderful job in controlling it. But the iFeel has > an extra interrupt endpoint which takes the force feedback commands, > and I wasn't sure the best way to put this nonstandard extension > into the current USB framework without rewriting or duplicating a > lot of code nicely implemented in the HID driver. > > Problem a is not a huge problem, it just makes it easier to test and > develop the lowest level stuff. > > Problem b is a more complicated issue, and I think this needs to be resolved > before we can move forward. I don't think it's a huge issue, but it needs > to be addressed. The place to do it is with the linuxconsole folks... I have no problem with just adding an hid-ifeel.c driver with reasonable hooks into the hid-core and hid-input. > Yay, verbosity. Anyway, I hope that this gives you a better idea of > where I'm at, and what needs to get done before we get some progress. > The actual commands to send to the mouse are dirt simple, we just > need to get them there smoothly... :) > > For now, my 'tactile' SF project is sleeping. Once we get the basic kernel > stuff going, I think it would be a good place to work on the higher level > stuff, perhaps in a more device independent way. (That's why I named it > 'tactile' and not 'ifeel-linux'.) -- Vojtech Pavlik SuSE Labs |
|
From: Adam G. <ad...@ev...> - 2001-10-08 00:58:11
|
Excellent, this sounds like a great start. Can you just split off this file in CVS? Then I can sync up with it and (eventually, once I get my laptop net card) manually send you the patches. Sound good? Adam On Mon, Oct 08, 2001 at 01:15:38AM +0200, Vojtech Pavlik wrote: > I have no problem with just adding an hid-ifeel.c driver with reasonable > hooks into the hid-core and hid-input. > > -- > Vojtech Pavlik > SuSE Labs > |
×
Want the latest updates on software, tech news, and AI?
Get latest updates about software, tech news, and AI from SourceForge directly in your inbox once a month.
Thanks for helping keep SourceForge clean.
X