2-week+ timer set [Re: Integrate libsane and sane into Solaris[PSARC/2008/170 FastTrack timeout 03/11/2008]]

[Tzongyu Paul Lee]

Hi all,

1. This open-source import project integrates sane and libsane to 
support usb scanner devices.
It uses libusb API from an unprivileged process. It does not introduce 
any device drivers or device nodes.

2. There is an issue needing further investigation in the context of 
device_allocation support and usb,
it is logged by "CR 6676059 Non mass-storage/audio devices need to be 
allocatable".

3. For the above two issues, this project would neither contribute to it 
nor make it worse.
 Since this fast-track case was extended in the 3/12 PSARC meeting,
 I would propose to explicitly set the timer two weeks from the original 
proposal 3/11 to
 3/25 to incorporate further discussions related to this case.
 And with hope we can converge based on the same or compatible 
understandings of the issues raised
 by this case.

Best Regards,
Paul

Gary Winiger wrote:
>> On Tue, 2008-03-18 at 17:35 -0800, Gary Winiger wrote:
>>     
>>>         The issue deals with both data (documents
>>> 	left in the scanner) and state.  While this might be a general
>>> 	USB libusb/ugen() issue, I'd not noticed it before. 
>>>       
>> This is pretty clearly a bordering-on-intractable USB/libusb/ugen issue;
>> there are no constraints on what sort of state can be retained by USB
>> devices, and no generic USB-level "go clean yourself" action that is
>> guaranteed to do what a device clean operation would be expected to do.
>>     
>
> 	I agree in the general case, USB seems to have no uniform
> 	architecture, however for Sun supplied device libraries, I
> 	believe we can do better than say: too bad if you use USB devices
> 	there is no hope for providing Object Reuse.
>
> Gary..
>   


-- 
Tzongyu Paul Lee, Tzongyu.Lee at Sun.Com or Paul.Lee at Sun.COM
BJS05 7225, x84343
http://blogs.sun.com/tpaullee/  Feel the Pulses of China
http://tpaullee.blogspot.com/