The problem with this change is that speed is shown in bytes, not in bytes/s.

I mentioned in D14835 that the total number of received/transmitted bytes is more useful than the current speed.
One can see the current speed on the graph.

ping

https://clang.llvm.org/extra/clang-tidy/checks/modernize-avoid-bind.html

std::bind can be hard to read and can result in larger object files and binaries due to type information that will not be produced by equivalent lambdas.

Maybe bump up required Qt version?
Qt 5.9 is the current LTS release and the next LTS release Qt 5.12 is coming soon.

rename slot

Why not use include(ECMEnableSanitizers)?

The constant 0.001 is taken from libzip. It means that the callback will be called each time when compressing more than 1/1000 part of a file.
At least it doesn't change the old behavior.

Lambdas are more readable.

D16248

simplify, leave only invokeMethod() -> QTimer::singleShot() change

Yes, but gwenview requires at least 5.9.
And code is longer with QMetaObject::invokeMethod, compare:
QTimer::singleShot(0, this, &FileOpsContextManagerItem::updateActions);
and
QMetaObject::invokeMethod(this, &FileOpsContextManagerItem::updateActions, Qt::QueuedConnection);

BTW, should I create a review request to reformat code with astyle-kdelibs?

more reinterpret_casts

fixed

fixed