| Age | Commit message (Collapse) | Author |
|---|
|
If the events server is initialized in a different process than the backend,
the txzmq socket raises an "zmq.error.ZMQError: Interrupted system call"
exception during the events server initialization. Despite that, communication
seems to work flawlessly after the initialization.
Moving the events server initialization to the same process as the backend
causes the exception to not be raised during events server intialization.
|
|
- Related: #6359
|
|
* Instead of checking if soledad is still syncing for the timeuot cancel
the delayed call.
* Count retries properly.
* Now soledad sync only returns SoledadError (#6981).
|
|
Before this commit, the backend used plain pyzmq bindings for communicating
with the frontend. This implements the txzmq twisted-powered bindings instead.
Closes: #6360
|
|
Change the 'ip' command location to support Fedora/RHEL distros.
/bin/ip is pressent on Debian/Ubuntu but not on Fedora.
/sbin/ip is a symlink to /bin/ip on Debian/Ubuntu and a binary on Fedora.
- Resolves: #6894
|
|
- Related: #6876
|
|
On 64bit Fedora/RHEL systems the python path is /usr/lib64 instead of
/usr/lib like on Ubuntu systems. This fix allows those users to use the
bootstrap_develop.sh script.
|
|
If the quit() call is triggered by the system logout we should not
disable the autostart. Otherwise bitmask won't autostart on the next
session start.
- Resolves: #6424
|
|
- Resolves: #6426
- Resolves: #6681
|
|
Catch the keyring.errors.InitError exception.
The automatic login sequence now stops correctly instead of freezing if
the user cancel the keyring open operation.
- Resolves: #6682
|
|
|
|
|
|
Decouple SRPRegister from QT signaler and create SRPRegisterImpl
|
|
Decouple SRPAuth from QT signaler and move it to SRPAuthImpl
|
|
On SRP#authenticate, no need to use so many threads.
|
|
Replace the hardcoded port '1194' for the port specified in
eip-service.json.
Choose the best port to use according which one is enabled in the
eip-service.json file
Resolves: #6541
|
|
After a provider fails to pass the checks, the providers combo box is
not enabled when it should, we just took care of the provider line edit.
- Resolves: #6418
|
|
Also set text to 'bold' so it's easier to find at a first sight.
- Resolves: #5526
|
|
This concern is addressed by the following mail pullreq, that will be
into 0.4.0 release:
https://github.com/leapcode/leap_mail/pull/175
By the way, it is really funny that I cannot tell if this comment was
written by meskio, as the blame says, or if I was the original author of
the remark. I should stop drinking so much coffee.
|
|
The code was using 'fail' which is not a Deferred method.
- Resolves: #6772
|
|
Hopefully this will make our git log more informative.
- Releases: 0.9.0
|
|
Resolves: #6713
|
|
|
|
|
|
|
|
|
|
|
|
... on them, to user-writeable only.
I think we could consider ipc for the curve-case too, at least for the
platforms that support them.
|
|
FIX: Login attempt is made against previously selected provider.
Closes #6654.
Cherry-picked from release/0.8.x since it's a very annoying bug for the
testing cycles for the next release (0.9.0) -- kali.
|
|
FIX: Login attempt is made against previously selected provider.
Closes #6654.
|
|
Use global flag for ZMQ_HAS_CURVE.
Closes #6646
|
|
|
|
|
|
the set_proxies_cb function in backend/components expects to have access
to these public properties.
|
|
This closes: #6658 and #6691
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Closes #6717.
|
|
|
|
factory.do_Stop will handle this now.
|
|
we should deal with this with pure deferreds
|
|
|
|
|
|
From kali: add some notes about the improper handling of the mailbox
required to initialize the account, and draft some notes about how
to improve this in next iterations.
|
|
|
|
Beware, the markup has to remember that the changelog is appended
to the long description, and all together is displayed as the package
info in the pypi page.
So, we cannot reuse section markups so happily: they have to be
consistent.
|
