trying to make progress with Subsurface-mobile

Dirk Hohndel dirk at hohndel.org
Sun Mar 19 12:34:48 PDT 2017


> On Mar 18, 2017, at 5:10 PM, Dirk Hohndel <dirk at hohndel.org> wrote:
> 
> I spent a day trying to fix the selection bug in Subsurface-mobile that
> Willem and others have reported. When you tap on a dive in the dive list,
> you often (but not always) get the next dive down the list (i.e., the dive
> one before the dive that you wanted, given the ordering of the dive list).
> 
> In the process I think I managed to clean up some of the code and remove
> some obsolete code, but unfortunately the problem itself still isn't
> fixed. Certain cases that used to reliably fail now work at times, but it
> still isn't "simply correct".
> 
> I thought I understood how our use of the QML ListView works, and I'm
> baffled why the wrong item is shown. Printing debug output isn't all that
> helpful because this is not a procedural language, so it's hard to make
> sure that the debug code actually runs when you want it to run.
> 
> Any help would be welcome. I'm running out of time today, I'll try to make
> new beta apps for Android and iOS so people can test.

A new Android apk is in daily: http://subsurface-divelog.org/downloads/daily/Subsurface-mobile-4.6.3.132-arm.apk

I also have a new iOS build, but that I need to send to the Beta channel 
because I can't create a .ipa that will install on a random iOS device (thanks, 
Apple). Then again, we have only ~250 users on iOS, so this isn't really
a popular platform for us.

In my testing so far I see two major differences:

a) it seems to reliably show the correct dive (good)
b) when you tap on a dive that's later in time, it sometimes renders ALL 
of the dives (in the background) and seems to be hung for a long time 
(depending on how many dives are in your dive list, I guess), before
it finally shows the right one (REALLY bad)

I guess that's progress? Not sure.

Testing and help in fixing the problem would be very welcome

/D


More information about the subsurface mailing list