This should never happen, since our interface is bassically synchronous,
but it could happen with delayed replies that came in just after we
decided to re-transmit a command.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Our model of waiting for 100ms before re-checking if we got a packet
over BLE resulted in potentially horrendously bad latency for received
packets.
That isn't just a possible performance issue, it actually seems to cause
IO errors with my Suunto EON Core. I'm not entirely sure why, but it
might simply be some timing interaction, particularly since the IO
errors seemed to primarily happen when the dive computer itself was also
busy updating the screen (ie if you pressed buttons on the dive computer
to switch to compass mode, for example).
So replace the silly hardcoded 100ms "waitFor()" function with a
WAITFOR() macro that checks the provided expression every time through
the loop, which gets us a much lower latency (we basically check every
ten milliseconds).
The macro is not beautiful, but it WorksForMe(tm).
This makes a huge difference to the reliability of the download for me,
and might matter for some other dive computers too.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This is based on something that Anton Ludin sent to the mailing list.
Reading through the code it seemed that there were scenarios in which
DC_vendor and DC_product were not updated correctly. That's one of the
problems of the declarative approach in QML - it can be very hard to
figure out which code is run when in certain situations.
This may help address the issue with FTDI downloads no longer working on
Android.
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
The asynchronous load seemed to be (at least one of) the culprit(s) of
the banner occasionally not showing up.
Making the font for the cloud ID smaller looks better (and works much
better for long email addresses).
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
We want to allow people to keep dives they collected without a cloud
account. The code was mostly there, we just got confused about the
existing status because we ran through this twice (no cloud -> unknown
-> verified). This way we explicitly remember this kind of transition.
Fixes#1404
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
That change clearly would have benefited from better code review.
This is a superset of a change proposed by Jan Iversen.
Closes#1406
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
MAX_TANK_INFO is defined in dive.h but is not
used in add_cylinder_description() or when
allocating 'tank_info'.
Use MAX_TANK_INFO instead of the literal 100.
Signed-off-by: Lubomir I. Ivanov <neolit123@gmail.com>
Instead of a constant or a macro for the maximum
number of 'ws_info' elements the 100 literal was used.
Define MAX_WS_INFO in dive.h and use it everywhere.
Also clamp loops that iterate `ws_info' to MAX_WS_INFO.
Prevents potential out-of-bounds reading, similarly to
the previous commit about 'tank_info'.
Signed-off-by: Lubomir I. Ivanov <neolit123@gmail.com>
In a number of places the global 'tank_info' array
is being iterated based on a 'tank_info[idx].name != NULL'
condition.
This is dangerous because if the user has added a lot of tanks,
such loops can reach 'tank_info[MAX_TANK_INFO]'. This is an
out of bounds read and if the 'name' pointer there happens to be
non-NULL, passing that address to a peace of code that tries
to read it (like strlen()) would either SIGSEGV or have undefined
behavior.
Clamp all loops that iterate 'tank_info' to MAX_TANK_INFO.
Signed-off-by: Lubomir I. Ivanov <neolit123@gmail.com>
In dive picture shift time dialog when pressing the up or down arrow
of the timeEdit widged there is some risk of double triggering because
the function called after this UI action ("updateInvalid()") can have
quite some runtime.
Suppress any potential double triggering by disabling the timeEdit
widget after each change until the code is processed.
Signed-off-by: Stefan Fuchs <sfuchs@gmx.de>
Apple typically forces a much more detailed review if the version number
changes. Let's get this taken care of now as we prepare for release.
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
Correctly use gettextFromC::instance()->tr(); instead of a simple
tr(); to translate the dive mode names.
This goes on top of 0bc9edf855
and finally makes the whole thing work.
Signed-off-by: Stefan Fuchs <sfuchs@gmx.de>
For some reason Kirigami.Icon mess up icon display when filename
extension is omitted. Because of this a perfectly good, scalable svg
show up as a low resolution scaled up icon.
Signed-off-by: Murillo Bernardes <mfbernardes@gmail.com>
calls to savePreferences was moved to prefs. in
b8eb348f54, but the corresponding
C++ code was not merged.
Revert call to savePreferences to manager.
Signed-off-by: Jan Iversen <jani@apache.org>
The commit secured that plotDive was not called before actually being used.
However our (rather fragile) C++ qml interface did not work correctly (ony sometimes).
Revert the previous commit.
Signed-off-by: Jan Iversen <jani@apache.org>
Kirigami appears to have a bug that makes it fail to show our icon.
With this we can be much more flexible in what we show in the top area
of the global drawer.
Fixes#1331
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
We want to be able to roll our own banner image, logo, title, and other
information in the GlobalDrawer - but Kirigami adds an ugly margin
around that. This attempts to remove that margin (but for some reason
there is still a margin on the left side).
This requires the patch command to be installed, but because of the
context sensitivity of the changes, I couldn't figure out how to do
this with sed or perl (which we already require).
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
git clone/checkout code was mostly the same copied over and over.
Move code to a single common function
Signed-off-by: Murillo Bernardes <mfbernardes@gmail.com>
The list of known dive computers was stored in a multi-map indexed
by the device name. Turn this into a sorted QVector. Thus, no
map-to-list conversion is needed in the device editing dialog,
which distinctly simplifies the code.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
Remove the explicit constructor in DiveComputerNode: Just use
classical C-style struct initialization. Moreover, remove the
empty constructor and destructor of DiveComputerList.
The variable DiveComputerList::dcWorkingMap was unused. Remove.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
update() was only called in conjunction with init(). No point in
having two functions. The separation between both functions seemed
arbitrary anyway.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
Enable translations for dive event name "modechange".
But clearly don't do this for the string which is used internally
but only push "modechange" to the translation system to be able to
translate it in the UI.
Signed-off-by: Stefan Fuchs <sfuchs@gmx.de>
Giving the string ": %1" to translation makes no sense.
And it's also not consistent with code a few lines below this line.
Signed-off-by: Stefan Fuchs <sfuchs@gmx.de>
We shouldn't need to manually set it, but it appears we do. To avoid
constant resizing, let's only grow it - and let's set the size before we
update the text.
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
According to Qt's documentation, the application exits if all windows
with the WA_QuitOnClose attribute are closed. This attribute was cleared
for three dialogs. This seems not necessary because:
1) The application can't be closed as long as the modal dialog is shown.
2) The flag only concerns primary windows, which these are not.
See: http://doc.qt.io/qt-5/qguiapplication.html#quitOnLastWindowClosed-prop
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
Commit b8eb348f moved credentialStatus but missed one spot.
When starting from a fresh install, clicking "No cloud mode” fails because of this.
Signed-off-by: Murillo Bernardes <mfbernardes@gmail.com>
