February 2006 - Bluetooth-dev

Follow up: Device Pairing Re-revisited
by Joseph Kelly 10 Feb '06

10 Feb '06

site_archiver(a)lists.apple.com Delivered-To: bluetooth-dev(a)lists.apple.com _______________________________________________ Do not post admin requests to the list. They will be ignored. Bluetooth-dev mailing list (Bluetooth-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/bluetooth-dev/site_archiver%40lists.… This email sent to site_archiver(a)lists.apple.com

1 0

Re: Device Pairing Re-revisited
by Joseph Kelly 08 Feb '06

08 Feb '06

site_archiver(a)lists.apple.com Delivered-To: bluetooth-dev(a)lists.apple.com Well, I'll go and implement this, and report on whether this works. joe _______________________________________________ Do not post admin requests to the list. They will be ignored. Bluetooth-dev mailing list (Bluetooth-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/bluetooth-dev/site_archiver%40lists.… Are you using UUIDs or channel IDs? If you use UUIDs how do you know the RFCOMM channel ID? This information has to come from the SDP! You're absolutely correct -- I'm using -getServiceRecordForUUID: whose documentation says "This method only operates on services that have already been queried. It will not initiate a new query." I'm quite embarrassed about this, as additionally I do not NULL-check the result of that operation which would've saved me a few days of work. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Problem removing Serial Port Service
by Tico Ballagas 08 Feb '06

08 Feb '06

site_archiver(a)lists.apple.com Delivered-To: bluetooth-dev(a)lists.apple.com Thanks -Tico On Jan 25, 2006, at 11:22 AM, Tico Ballagas wrote: Thanks -Tico This email sent to ballagas(a)cs.rwth-aachen.de _______________________________________________ Do not post admin requests to the list. They will be ignored. Bluetooth-dev mailing list (Bluetooth-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/bluetooth-dev/site_archiver%40lists.… I've been doing a little more investigation on this problem. I ran "lsof" as root to list the open file handles since the bluetooth serial port is implemented through the /dev/tty.* file descriptor. "/ dev/tty.<my bluetooth serial port>" is not listed indicating that the device is not currently in use by a process (in contradiction to the error message). I think the problem is that there is a filesystem lock left over that I need to delete. However, I've been looking around in the /var/ directory and can't seem to identify which file it might be. Could anyone help me with this? I'm having a strange issue where I am unable to remove a Bluetooth Serial Port. I get the following error: "The service cannot be modified. You are attempting to remove a port which is in use. Quit the application using this port and try again." However, this happens even after a fresh boot with a fresh user (no login items). How do I check which process is using a particular serial port. Is there an alternative way to remove a port? _______________________________________________ Do not post admin requests to the list. They will be ignored. Bluetooth-dev mailing list (Bluetooth-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/bluetooth-dev/ballagas% 40cs.rwth-aachen.de This email sent to site_archiver(a)lists.apple.com

1 0

Re: Device Pairing Re-revisited
by Alexander Traud 08 Feb '06

08 Feb '06

site_archiver(a)lists.apple.com Delivered-To: bluetooth-dev(a)lists.apple.com User-agent: Microsoft-Entourage/9.0.6.8320 >> If the Bluetooth service needs authentication, than a pairing is requires. > Bluetooth service? I call RFCOMM and L2CAP "channels" a service (because there is a bit more than the channel itself; the SDP record etc, etc.) >> Since Mac OS X v10.4.3 since can be done programmatically without the Apple >> Bluetooth Assistant. > How? It is called "Inquiry" - just do a search on the Mac OS X v10.4 headers you are using. > I have never bothered to check the services, because I know my particular > device only does one thing. Are you using UUIDs or channel IDs? If you use UUIDs how do you know the RFCOMM channel ID? This information has to come from the SDP! If you are using channel IDs directly, make sure the service is on that ID. I recommend to look up all returned error codes once again. Perhaps you always connect to a specific RFCOMM channel without knowing it. _______________________________________________ Do not post admin requests to the list. They will be ignored. Bluetooth-dev mailing list (Bluetooth-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/bluetooth-dev/site_archiver%40lists.… This email sent to site_archiver(a)lists.apple.com

1 0

Re: Device Pairing Re-revisited
by Joseph Kelly 08 Feb '06

08 Feb '06

site_archiver(a)lists.apple.com Delivered-To: bluetooth-dev(a)lists.apple.com joe _______________________________________________ Do not post admin requests to the list. They will be ignored. Bluetooth-dev mailing list (Bluetooth-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/bluetooth-dev/site_archiver%40lists.… Wallowing in the mires of Bluetooth Ignorance, I would like to add this question: Is it requisite, perhaps, to call -performSDPQuery on the device, beforehand? I have never bothered to check the services, because I know my particular device only does one thing. I noticed that Bluetooth Setup Assistant merely needs to "Gather Some Information" about the device, and if I quit outright at that point, the device works (and a "Services" entry appears in the device's dictionary in blued.plist). This email sent to site_archiver(a)lists.apple.com

1 0

Re: Device Pairing Re-revisited
by Joseph Kelly 08 Feb '06

08 Feb '06

site_archiver(a)lists.apple.com Delivered-To: bluetooth-dev(a)lists.apple.com How? (Dammit, I'd hate if I missed a new SDK.) Again: It depends on the security setting of the remote Bluetooth device, if a pairing is required. The resulting error code should have been quite informative for such a case... joe _______________________________________________ Do not post admin requests to the list. They will be ignored. Bluetooth-dev mailing list (Bluetooth-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/bluetooth-dev/site_archiver%40lists.… On Feb 7, 2006, at 2:39 PM, Alexander Traud wrote: If the Bluetooth service needs authentication, than a pairing is requires. Bluetooth service? I'm using a simple RFCOMM channel (serial port service class). Is there something more I should be doing? Since Mac OS X v10.4.3 since can be done programmatically without the Apple Bluetooth Assistant. We've tried adjusting the security. It's now at the 0 level (lowest -- no passcode, etc). And it still fails. There is no resulting error. I make the call to - openRFCOMMChannelAsync: which returns immediately with 0, and then - rfcommChannelClosed: gets called almost immediately after. - rfcommChannelOpenComplete:status: never gets called. If the device is in range and closes the connection correctly (shut down, turning off), it is quite immediately. If the device is out of range, the event fires reasonable. I am now getting -rfcommChannelClosed: messages quite happily. I had forgotten to add all the delegate methods. I'm probably going to DTS this one, but I'm hoping someone could say "oh yeah, we had that problem and we fixed it like this..." This also could be a problem with the device itself, but since the Bluetooth Setup Assistant makes everything just work, then dammit, I want This email sent to site_archiver(a)lists.apple.com

1 0

Re: Device Pairing Re-revisited
by Alexander Traud 07 Feb '06

07 Feb '06

site_archiver(a)lists.apple.com Delivered-To: bluetooth-dev(a)lists.apple.com User-agent: Microsoft-Entourage/9.0.6.8320 I am not Apple, however I think I can answer this one, too... > must all devices be made known to the Mac via the Bluetooth Setup Assistant > in order to use them No. > if so, is there a programatic way to avoid this nastiness? If the Bluetooth service needs authentication, than a pairing is requires. Since Mac OS X v10.4.3 since can be done programmatically without the Apple Bluetooth Assistant. Again: It depends on the security setting of the remote Bluetooth device, if a pairing is required. Again: This can be done programmatically, however you should consider to use the inquiry user interface in the BluetoothUI package which does the same on your request - although it has worse interfaceÝ than the Apple Bluetooth Assistant. ÝWhy is a Bluetooth passkey shown as bullets here. Does this still happen in Mac OS X v10.4? Anyway, this is the way it was in previous versions. > I'm seeing that I am unable to open RFCOMM channels on certain of our > bluetooth devices The resulting error code should have been quite informative for such a case... > do I rely on the delegate methods -- say of IOBluetoothRFCOMMChannel -- for > instance will -rfcommChannelClosed: be invoked when a device disappears? In case you did not get any answer to this one, my two cents: I am quite happy with the *events* of the callbacks when it comes to channel closing. Not sure what you are talking about. There should be *no* polling with Bluetooth. You could poll on a cable link, but I would not poll on a Layer 2 secured link like Bluetooth if you have access to its API like on the Mac. If the device is in range and closes the connection correctly (shut down, turning off), it is quite immediately. If the device is out of range, the event fires reasonable. There is no way to get the information more soon - I doubt that even an additional layer 2 protocol on the link will be faster. However, I am not sure if this meets your requirements. "Near real time" - what is this? ;-) _______________________________________________ Do not post admin requests to the list. They will be ignored. Bluetooth-dev mailing list (Bluetooth-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/bluetooth-dev/site_archiver%40lists.… This email sent to site_archiver(a)lists.apple.com

1 0

Device Pairing Re-revisited
by Joseph Kelly 07 Feb '06

07 Feb '06

site_archiver(a)lists.apple.com Delivered-To: bluetooth-dev(a)lists.apple.com Thanks. joe _______________________________________________ Do not post admin requests to the list. They will be ignored. Bluetooth-dev mailing list (Bluetooth-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/bluetooth-dev/site_archiver%40lists.… Once again, I'm seeing that I am unable to open RFCOMM channels on certain of our bluetooth devices w/o first running the Bluetooth Setup Assistant. The device opens and authenticates okay, but then when I issue an RFCOMM open command, I get an immediate - rfcommChannelClosed:. It's all remedied, however, after running through the S.A. So I gotta ask the question again: must all devices be made known to the Mac via the Bluetooth Setup Assistant in order to use them, and if so, is there a programatic way to avoid this nastiness? Since it's intermittent, I have this glimmering hope that maybe it's something I'm doing wrong, and blah blah, but the official word from Apple would really help us firm up our documentation (so we can advise our users that they must run through the S.A.). This email sent to site_archiver(a)lists.apple.com

1 0

Bluetooth via Rosetta
by Geoffrey Schmit 06 Feb '06

06 Feb '06

site_archiver(a)lists.apple.com Delivered-To: bluetooth-dev(a)lists.apple.com My ppc-only application that uses Bluetooth does not work on an Intel-based iMac. In the process of determining the issues involved, I compiled the RFCOMMClientSample example as a ppc-only application and attempted to run it on the iMac. While running, the following messages were displayed in the system log: Feb 6 16:47:20 icbm kernel[0]: IOUserClient: failing rosetta access to IOBluetoothHCIUserClient Feb 6 16:47:27 icbm kernel[0]: IOUserClient: failing rosetta access to IOBluetoothDeviceUserClient and in the console log: 2006-02-06 16:47:26.778 RFCOMMClientSample_ObjC[384] Error: 0xe00002c0 - unable to open RFCOMM channel. 2006-02-06 16:57:14.964 RFCOMMClientSample_ObjC[387] Service selected 'binary Chat Server' - RFCOMM Channel ID = 5 The application fails to connect to the RFCOMMServerSample example running on my other system. What steps are required such that an application running in Rosetta can leverage the Bluetooth framework? <radar://4435210> Thanks! geoff -- Geoffrey Schmit Senior Software Engineer National Instruments geoffrey.schmit(a)ni.com www.ni.com _______________________________________________ Do not post admin requests to the list. They will be ignored. Bluetooth-dev mailing list (Bluetooth-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/bluetooth-dev/site_archiver%40lists.… This email sent to site_archiver(a)lists.apple.com

1 0

Re: return value of getDeviceClassMajor
by Elliotte Rusty Harold 01 Feb '06

01 Feb '06

site_archiver(a)lists.apple.com Delivered-To: bluetooth-dev(a)lists.apple.com User-agent: Thunderbird 1.4 (Macintosh/20050908) Am 31.01.2006 um 17:33 schrieb Elliotte Harold: -- Elliotte Rusty Harold elharo(a)metalab.unc.edu XML in a Nutshell 3rd Edition Just Published! http://www.cafeconleche.org/books/xian3/ http://www.amazon.com/exec/obidos/ISBN=0596007647/cafeaulaitA/ref=nosim _______________________________________________ Do not post admin requests to the list. They will be ignored. Bluetooth-dev mailing list (Bluetooth-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/bluetooth-dev/site_archiver%40lists.… Christian Klein wrote: What is the legal range for values returned by getDeviceClassMajor in IOBluetoothDevice? Is it 1-31 or 0 to 7968 or something else? In other words, is it an unsigned 5-bit int or unsigned 13-bit int, or something else? I've seen both used in practice in other APIs for working with Bluetooth. // +---------------+---------------+---------------+ // | octet 3 | octet 2 | octet 1 | <- Octet Transmission Order // +---------------+---------------+---------------+ // <------ 11 bits ----->< 5 bits ><- 6 bits -> // +---------------------+---------+-----------+-+-+ // | Service Classes | Major | Minor | | | // +-+-+-+-+-+-+-+-+-+-+-+ Device | Device |0|0| // | | | | | | | |*|*|*| | Class | Class | | | // +-+-+-+-+-+-+-+-+-+-+-+---------+-----------+-+-+ // | | | | | | | | | Yes, but there's more than one way to interpret that layout. The 5 bits of the major device class could be represented as a simple 5 bit number or the whole class ID could be masked off and the entire thing treated as a 32-bit number of which only the major class part can be non-zero. I've seen both done elsewhere. Which does Apple's Bluetooth API do? (Or does it do something completely different?) This email sent to site_archiver(a)lists.apple.com

1 0