Fwd: Re: Subsurface-mobile downloadfrom divecomputer

[Dirk Hohndel]

On Sat, Feb 13, 2016 at 08:55:27PM +0200, Willem Ferguson wrote:
> >>1) I have now gone beyond the last patch that I submitted (I think 3 days ago) doing some first groundwork by removing some code from the usual binaries and putting it the downwnloadmanager class. Looks like that patch has not been implemented yet. So I can now submit a larger patch that INCLUDES the previous one and that, in addition, populates the vendor and product comboboxes.
> >That sounds excellent. Yes, please send a new patch set - as I'm sure you've noticed, I've been very focused on getting this first release out and spent all my time on cleaning up the issues found in the alphas and the beta (frankly, surprisingly little feedback on that - as usual, I guess (so why am I surprised?)).
> >
>  I need to filter the above dive computer names to those with FTDI
> interfaces. In libdivecomputer the dc_descriptor_t contains a lot of
> information, for instance the transport definition. In addition there is
> dc_type that also has information. Which is the mechanism to detect
> which compuyers have an FTDI interface?

I don't think there's a way to tell from their descriptor. We may have to
white-list the ones that we know works. Manually.

> >>At the moment I am figuring out how to get communication between C++ and the progressBar.
> >I may be wrong but I think the way you'd do that is to have a Q_PROPERTY on the C++ side that you update with the progress value and then use that property on the QML side for the value in the progress bar. It should automagically be kept in sync as long as your setter on the C++ side raises the corresponding ...Changed signal every time it updates the value. Does that make sense?
> 
> It makes very much sense, but the problem is that QComboBox takes a
> variable of data type 'real' to indicate progress. No indication on

You lost me. Do you mean the QML type ProgressBar?

Not sure where a QComboBox needs a real.

> whether this is 16-bit, 32-bit or even longer. In c++ I translated that
> as float but this is currently not working.  The combobox value is not
> overloaded for other formats. Calling the setter  that provides a float
> from C++ currently creates a segfault. But I will get there.

Can you post a piece of code?

> I am looking at updating the mobile user manual, but am waiting for a UI
> that does not change rapidly. Looks like we are close.

Yes. With my apologies for the thrash, but I think this is worth it. It
feels MUCH better now. But I'd give it a few more days.

/D