We stop using a buffer so it's less memory intensive and allows for larger files to be sent.
With current KIO has CPU usage issues, there's a patch in review that fixes them.
BUGS: 378488
Summary:
QSslSocket is already a QIODevice so just use that.
Should fix the issue of transfering *big* files.
Unfortunately this seems to trigger a bug in KIO and CPU usage goes through the roof, so haven't really been able to test it does actually fix things.
Please don't merge/approve yet
Reviewers: apol, albertvaka, kdeconnect
Subscribers: #kde_connect
Differential Revision: https://phabricator.kde.org/D6039
Summary:
There is currently an issue where the device doesn't send the correct IP address for sshfs when a VPN is present.
Instead of asking the device to find and send its address, we can store it from when the device link is created then reuse it.
Test Plan:
All unit tests pass.
In these different situations:
- without a VPN
- with a VPN running
- with a VPN started then stopped (the tun interface might still be there)
Try to remotely browse the device. It should work in all cases.
Reviewers: #kde_connect, albertvaka
Reviewed By: #kde_connect, albertvaka
Subscribers: albertvaka, apol, nicolasfella, sredman
Tags: #kde_connect
Differential Revision: https://phabricator.kde.org/D6730
If we're ignoring an ssl issue, at least show it.
On connected show whether the socket is valid, otherwise we get remote
devices disconnected without any message. (In this case HostNameMismatch)
Also fixes a typo
CCBUG: 381542
Notifications are not persistent, and once gone there is no
way to accept the request. Clicking the notification in the
notification history does nothing in plasma and gnome.
There make them persistent and close them after the pairing timeout,
after which they are no longer useful.
BUG: https://phabricator.kde.org/T5002
If both devices are in "discovery mode" (ie: both want to keep links
established) and the last one to create the link goes out of discovery
mode, it will close the link when it shouldn't (because the other end still
was "using" it to display it to the user).
Introduces a big fat buffer :(
Actually test the trasfers :)
Takes QSslSocket causistic into account, for some reason QNAM refuses to
mark as finished when the QSslSocket (through QIODevice) closes.
It would be good to look into dropping the QBuffer, doing so with the test
in place will help.
Makes it possible to specify the different properties sent at once,
rather than one by one as we used to do.
Also port whenever possible to the initializer-list syntax.
REVIEW: 128269
