[libhid-discuss] hid_force_open failed with return code 12 again

2008-03-19 Thread comicinker
I've the same problem like described here:
http://lists.alioth.debian.org/pipermail/libhid-discuss/2007-March/000104.html
I don't want to run my app as root to access the device.

I tried the solution like already discussed in

http://lists.alioth.debian.org/pipermail/libhid-discuss/2007-March/000110.html

However, I'm new to udev and all that stuff, even though I slightly
understand what it does, I couldn't get my device to work. I added a
file

49-hid-development.rules

with following content:

# udev rules file for development of usb hid device driver experiments
# to be replaced by a script!
# use @RUN_AS_GROUP@ tag for setting the rw permissions for the correct
user id


# play around with the Microsoft Wireless mouse
SYSFS{idVendor}==045e, SYSFS{idProduct}==00e1, MODE=664,
GROUP=1000

# for the BiSP gateway 
SYSFS{idVendor}==03eb, SYSFS{idProduct}==2034, MODE=664,
GROUP=1000

I (UID 1000) want to have read and write access for two devices to play
around with them. 

As I've read in the udev readme, a reboot should not be nevessary after
saving that file in the udev/rules.d directory. What am I doing wrong?


Regards,

comicinker

-- 
my public key is on
hkp://pgp.mit.edu:11371
or
ldap://keyserver.pgp.com
___
libhid-discuss mailing list
libhid-discuss@lists.alioth.debian.org
http://lists.alioth.debian.org/mailman/listinfo/libhid-discuss

Re: [libhid-discuss] Parsing hid_dump_tree()

2008-03-29 Thread comicinker

Am Freitag, den 28.03.2008, 21:01 -0400 schrieb Charles Lepple:
 I wouldn't advise parsing the output of that function - instead, take  
 a look at the hidparser code that hid_dump_tree() references.

Thanks, good hint. Now I also found http://libhid.alioth.debian.org/doc/
which is nowhere mentioned as documentation. But with that documentation
it makes more sense to have a look at the code.

 It looks like you may have to look at the tree.
 
 (libhid was originally designed for HID PDC UPSes, which have fairly  
 strict requirements on the format of the HID element tree - so a  
 fixed path would work there.)
 
  AND: How can I get the RECV_PACKET_LEN for this address automatically?
  Or is it always the same length for each address (6)? The  
  test_libhid.c
  doesn't tell where this value is detected.
 
 
 It is in the interface descriptor.


It seems like I have to learn more about USB and HID (besides reading
the specs - sigh).

Thank you and sorry for my clumsy behavior in this list.

Bests,

comicinker
___
libhid-discuss mailing list
libhid-discuss@lists.alioth.debian.org
http://lists.alioth.debian.org/mailman/listinfo/libhid-discuss

Re: [libhid-discuss] Parsing hid_dump_tree()

2008-03-31 Thread comicinker
I slowly understand how the structure of the libhid is. This is really
great work, thanks!

Am Freitag, den 28.03.2008, 21:01 -0400 schrieb Charles Lepple:
  AND: How can I get the RECV_PACKET_LEN for this address
 automatically?
  Or is it always the same length for each address (6)? The  
  test_libhid.c
  doesn't tell where this value is detected.
 
 
 It is in the interface descriptor.
 

However I still wonder how to determine the RECV_PACKET_LEN for specific
addresses, which is required for hid_get_input_report()? I found 

hidif-hid_data-Size

which returns the size (8) of the current Usage in Bits. I know that
this Usage is wrapped together with other Usages in one report with the
ReportID 17 and this report is in complete 6 Bytes long. So can I
determine the length of this report? Do I have to loop through the whole
report parser and add up the sizes belonging to that report ID myself?

Bests

comicinker

-- 
my public key is on
hkp://pgp.mit.edu:11371
or
ldap://keyserver.pgp.com
___
libhid-discuss mailing list
libhid-discuss@lists.alioth.debian.org
http://lists.alioth.debian.org/mailman/listinfo/libhid-discuss

Re: [libhid-discuss] Parsing hid_dump_tree()

2008-04-21 Thread comicinker

Am Freitag, den 28.03.2008, 21:01 -0400 schrieb Charles Lepple:
 
  AND: How can I get the RECV_PACKET_LEN for this address
 automatically?
  Or is it always the same length for each address (6)? The  
  test_libhid.c
  doesn't tell where this value is detected.
 
 
 It is in the interface descriptor.

I'm still looking for this entry in the interface descriptor. Can
somebody give me a hint where it is hidden? Is it really HIDData.Size?

Thanks in advance

comicinker

-- 
my public key is on
hkp://pgp.mit.edu:11371
or
ldap://keyserver.pgp.com
___
libhid-discuss mailing list
libhid-discuss@lists.alioth.debian.org
http://lists.alioth.debian.org/mailman/listinfo/libhid-discuss

Re: [libhid-discuss] Parsing hid_dump_tree()

2008-04-22 Thread comicinker
Am Dienstag, den 22.04.2008, 08:31 -0400 schrieb Charles Lepple:
 On Apr 21, 2008, at 8:21 PM, comicinker wrote:
 
  Am Freitag, den 28.03.2008, 21:01 -0400 schrieb Charles Lepple:
 
  AND: How can I get the RECV_PACKET_LEN for this address
  automatically?
  Or is it always the same length for each address (6)? The
  test_libhid.c
  doesn't tell where this value is detected.
 
 
  It is in the interface descriptor.
 
  I'm still looking for this entry in the interface descriptor. Can
  somebody give me a hint where it is hidden? Is it really HIDData.Size?
 
 HIDData.Size is the size of one element, I think. (Sorry, I don't  
 have time to look through the code.)
 
 I was mistaken - I was thinking of wMaxPacketSize in the endpoint  
 descriptor (not the interface descriptor). I usually either just  
 hardcode that for a given device, or use control transfers instead.
 

Do I understand it right that you use control transfers for each single
data package? I think I won't tell you something new, but (depending on
the chip architecture and vendor) this causes more operating time on the
chip. I must refrain from using control transfers because my controller
is not fast enough to deliver data and process all control transfer
requests. So my device does not support the get_input_report request. I
must use bulk or interrupt transfers. Oh I just see I should use the
hid_interrupt_read function. Da***

However, I would like to determine the length of a specific report with
a certain report ID, or the maximum length of the biggest report. Do I
have to do it by myself by looping through the whole HIDParser?

Bests,

comicinker


-- 
my public key is on
hkp://pgp.mit.edu:11371
or
ldap://keyserver.pgp.com
___
libhid-discuss mailing list
libhid-discuss@lists.alioth.debian.org
http://lists.alioth.debian.org/mailman/listinfo/libhid-discuss

Re: [libhid-discuss] Windows support?

2008-09-07 Thread comicinker
Peter, thanks for the translations.

Sorry for the German version, I thought with this amount of errors,
there must be a small problem in the very beginning and is easy to
solve.

I attached now the MSVC errors in almost complete English. I also
attached the config.h which I am using now, with some explanations
added.

As I said, I just added some header files to the paths, which I
arbitrarily copied from different systems and programs (see comments in
config.h), however, some need also the libraries for the OS, so it is
basically nonsense what I did. That's why I need some help, as I don't
know how to provide all the needed declarations and definitions.

Thanks in advance,


C



Am Samstag, den 06.09.2008, 09:54 -0400 schrieb Charles Lepple:
 On Sep 6, 2008, at 3:53 AM, comicinker wrote:
 
  How did you manage to compile libhid?
 
  I'm using Visual Studio 2005. I receive 240 Errors. I included a
  config.h from Linux, but that doesn't make sense at all. See the
  attached files.
 
 I don't have a Windows machine to test this with, but you may get a  
 little bit more help from others on the list if you temporarily set  
 the system language to English, then post that build log. Please  
 compress the log as well (gzip or pkzip).
 
 It looks like the ERROR() definition is conflicting with something in  
 wingdi.h:
 
 c:\dokumente und einstellungen\simeon\eigene dateien\projects\libhid 
 \libhid\include\debug.h(21) : warning C4005: 'ERROR': Makro- 
 Neudefinition
  c:\programme\microsoft visual studio 8\vc\platformsdk\include 
 \wingdi.h(98): Siehe vorherige Definition von 'ERROR'
 
 I'm not sure which would take longer for me, installing Windows then  
 downloading all of Visual Studio Express and the platform SDK, or  
 learning enough German to decipher that log file :-)
 
___
libhid-discuss mailing list
libhid-discuss@lists.alioth.debian.org
http://lists.alioth.debian.org/mailman/listinfo/libhid-discuss

Re: [libhid-discuss] Windows support?

2008-09-07 Thread comicinker
Forgotten zip file. Sorry for traffic...


___
libhid-discuss mailing list
libhid-discuss@lists.alioth.debian.org
http://lists.alioth.debian.org/mailman/listinfo/libhid-discuss