Linus Torvalds
2018-05-13 19:36:56 UTC
On Sun, May 13, 2018 at 12:00 PM Willem Ferguson <
and we never even get to the point where we can talk to it.
Why _that_ happens, I have absolutely no idea. There is simply not a lot of
information to go on. It's like the device isn't paired.
We simply do
serial_port->socket->connectToService(remoteDeviceAddress, ..
and then wait for the socket to go into ConnectedState. Which it obviously
doesn't do. IOW, this isn't a subsurface issue, it's something else.
Now, you *could* try the attached hacky patch, that tries to avoid using Qt
for this, and uses the libdivecomputer rfcomm functionality instead.
This patch requires that you have libbluetooth installed
(bluez-libs-devel), and honestly, I don't think that the Qt rfcomm code is
the problem. But I guess it's easy enough to test this patch.
Other than this patch, I would suggest just trying to unpair and then
re-pair the device. I assume you've done that a number of times already and
it didn't make any difference.
I guess at some point I should just get one of those stupid rfcomm devices.
Linus
Thank yo so much for your time, Linus. Nope, same error. Error bar at
bottom of Subsurface screen "Unsupported operation".
Yeah, that's just the "none of the protocols we tried worked" error case.bottom of Subsurface screen "Unsupported operation".
Failed to connect to device 00:13:43:5B:8F:BE .
Device state QBluetoothSocket::UnconnectedState .
Error: QBluetoothSocket::UnknownSocketError
So it's simply that Qt doesn't successfully connect to the device at all,Device state QBluetoothSocket::UnconnectedState .
Error: QBluetoothSocket::UnknownSocketError
and we never even get to the point where we can talk to it.
Why _that_ happens, I have absolutely no idea. There is simply not a lot of
information to go on. It's like the device isn't paired.
We simply do
serial_port->socket->connectToService(remoteDeviceAddress, ..
and then wait for the socket to go into ConnectedState. Which it obviously
doesn't do. IOW, this isn't a subsurface issue, it's something else.
Now, you *could* try the attached hacky patch, that tries to avoid using Qt
for this, and uses the libdivecomputer rfcomm functionality instead.
This patch requires that you have libbluetooth installed
(bluez-libs-devel), and honestly, I don't think that the Qt rfcomm code is
the problem. But I guess it's easy enough to test this patch.
Other than this patch, I would suggest just trying to unpair and then
re-pair the device. I assume you've done that a number of times already and
it didn't make any difference.
I guess at some point I should just get one of those stupid rfcomm devices.
Linus