Fwd: BLE on macOS and many requests for testing

Benjamin nystire at gmail.com
Sat Sep 23 00:18:26 PDT 2017


Resending to the entire list due to not enough coffee.

---------- Forwarded message ----------
From: Benjamin <nystire at gmail.com>
Date: 23 September 2017 at 10:18
Subject: Re: BLE on macOS and many requests for testing
To: Dirk Hohndel <dirk at hohndel.org>


Good morning
Due to the Mediterranean not cooperating, I'm testing software instead of
diving today :(
Running the latest version from the daily builds - 4.6.4 - 839 with a
Petrel 2.
Windows 10 64 bit - Bluetooth works fine
macOS 10.13 Beta (17A360a) - can't seem to connect to the dive computer.
The logfile consists of the following 2 lines:

INFO: Open: name=00:13:43:0C:56:29

ERROR: Failed to open the serial port. [in ../../src/shearwater_common.c:46
(shearwater_common_open)]

This behaviour under macOS is consistent even after removing the Bluetooth
pairing and reassociating the dive computer to the computer.

Benjamin

On 18 September 2017 at 02:41, Dirk Hohndel <dirk at hohndel.org> wrote:

> I finally found some time to look more into BLE on macOS. Things there are
> rather different because Qt doesn't give us BT addresses. Or, actually, it
> never gives us BT addresses for BLE devices, it seems, it sometimes (but I
> can't quite figure out when) gives us BT addresses for BT classic devices.
>
> Anyway, our whole BLE setup was designed around the idea that we'd have a
> BT address to use in order to create our QLowEnergyController. Sadly that
> doesn't work at all on macOS (or iOS), and is deprecated on Linux (I
> haven't checked if it will still be in 5.10, though).
>
> Instead, what you are supposed to do is to create that controller using
> your QBluetoothDeviceInfo. But sadly I haven't figured out a way to get
> this unless I am scanning for BT/BLE devices. So what this means is that
> at least in the BLE case you now need to scan for and pick your dive
> computer every time you want to download. :-(
>
> At first I tried to use that newer method on all OSs, but it turns out
> that that was a bad idea. For one thing, our Android implementation
> doesn't use Qt to scan for devices, so a lot more code would have to be
> rewritten there. And additionally, the new interfaces were only introduced
> in Qt 5.7, so the Linux distros with Qt 5.6 would fail to build Subsurface
> after I made those changes.
>
> So for now I went back to the old code for all the non-Apple OSs.
>
> So what has been tested (and works) so far:
>
> BLE download on Mac, right after scanning for and selecting the dive
> computer appears to work with Shearwater Petrel AI, Shearwater Petrel 2,
> and Suunto EON Steel.
>
> More testing (especially with other BT and BLE dive computers) would be
> extremely useful:
>
> http://subsurface-divelog.org/downloads/test/Subsurface-4.6.
> 4-830-gae209a3d9f8e.dmg
>
> BLE download on Linux (Fedora with a manual install of Qt 5.9.1) works
> with a Shearwater Petrel AI. I couldn't get the Petrel 2 to work because
> BT and BLE code somehow stepped all over each other there (this is a dual
> mode dive computer). I need to investigate a bit more what's going wrong
> there.
>
> I have NOT checked Windows, yet; obviously, BLE isn't supported there at
> all, but I would love to know if BT still works or if I inadvertandly
> broke something there.
>
> http://subsurface-divelog.org/downloads/test/subsurface-4.6.
> 4-831-g753c00a4936f.exe
>
> I did a quick test on Android and things seem to continue to work there.
> Again, there's a new test APK and testing would be appreciated.
>
> http://subsurface-divelog.org/downloads/test/Subsurface-mobi
> le-4.6.4.831-arm.apk
>
> And I haven't had enough time to try to build a new iOS binary based on
> this latest code. This will require me to write some Qt based BLE scanning
> for Subsurface-mobile (which might eventually allow us to migrate the
> Android binary to the new way of doing things).
> But in general, I'm hoping that we might be getting closer to having
> working BLE download for iPads and iPhones.
>
> Please test and let me know.
>
> Thanks
>
> /D
>
> _______________________________________________
> subsurface mailing list
> subsurface at subsurface-divelog.org
> http://lists.subsurface-divelog.org/cgi-bin/mailman/listinfo/subsurface
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.subsurface-divelog.org/pipermail/subsurface/attachments/20170923/07f1e2b9/attachment.html>


More information about the subsurface mailing list