| Age | Commit message (Collapse) | Author |
|---|
|
For now, the way to select a gateway is to add a section in
bitmaskd.cfg:
[vpn_prefs]
locations = ["frankfurt", "seattle__wa"]
countries = ["DE", "US"]
Note that the location indication has priority over country code.
This will be exposed by the UI in release 0.11
- Resolves: #8855
|
|
Old versions of the webapp let agent download an invalid.json, instead
of returning a 404. We try to parse the json, and if no valid json is
found, we use the workaround for manually downloading the service files.
- Resolves: #9004
|
|
|
|
This is a first approach to automatic gateways selection.
More things are missing:
- allow manual selection, by location or country code.
- take the hemisphere into account.
- expose the selected gw to the api/cli
but overall seems this is a good approach to make 0.10 release usable in
terms of vpn.
- Resolves: #8804
|
|
If chromium is installed in the system, there's no need to depend on
qt5.
|
|
wrong rename in the debian files
|
|
|
|
this allows for better modularization in, for instance, the debian
package. we should be able to install just the submodules for the vpn
service, or the mail service.
a better modularization needs the mail_services under core to be moved
to bitmask.mail, and do the switch in the main service module for
bitmask daemon.
|
|
with this organization, there's no need to install too many dependencies
if user only needs either vpn or mail.
|
|
|
|
|
|
This refactors the confirmation component to use sinon instead of
manually checking the the result of calling a function.
|
|
Newer versions of npm don't autoinstall peer dependencies, so they
have to be explicitly installed in package.json. This adds the few
that npm was complaining about.
|
|
Defaults to True. Without it, the confirmation modal doesn't show and
you aren't able to remove a provider from the list.
|
|
|
|
|
|
|
|
- Resolves: #8986
|
|
|
|
|
|
|
|
- Resolves: #8993
|
|
|
|
|
|
from here on, we're depending on leap's fork.
|
|
|
|
|
|
|
|
In case of the helpers not installed it was crashing the VPNService.
|
|
|
|
Just in chance the firewall was not properly cleaned up in a previeous
run (bitmask crashed, some bugs in the code, ...). Let's clean up the
firewall when it starts.
- Resolves: #8979
|
|
|
|
We make an ugly step for that moving the VPNStatus into VPNManagement.
Let's iterate on it a second time to clean up that.
- Related: #8942
|
|
So it's coherent with FirewallManager.
- Related: #8942
|
|
VPNService has knowledge of the tunnel and the firewall and controls
them separately. Also exceptions from VPNControl are handled locally
instead of being propagated.
- Resolves: #8976
- Related: #8942
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Currently, our only OSX gitlab runner is unavailable too often
and the tests fail in CI anyway so we disable it for now until
we fix both.
|
|
|
|
|
|
* Move todo list to https://0xacab.org/leap/bitmask-dev/issues/8929
* use bundled pysqlcipher - debian package has not been fixed yet.
* reset bitmaskd for each scenario so they are isolated
* run functional tests on CI
* moved e2e tests before the bundle
* add test_functional_graphical Make target
* Install chromedriver in docker image
* add screenshots as artifacts on failure
* run chrome without sandbox for docker
Tests were failing on CI with chrome sandbox:
https://0xacab.org/leap/bitmask-dev/-/jobs/15196
Used this workaround:
https://stackoverflow.com/questions/28364012/webdriver-exception-chrome-not-reachable/28949227#28949227
- Resolves: #8929
|
|
Signed-off-by: Ruben Pollan <meskio@sindominio.net>
|
|
|
|
if there is no ca_cert passed and the provider is configured, use the
pinned ca cert when the autoconf parameter is true.
|
|
|
|
This can be produced by errors fetching keys from the server.
- Resolves: #8971
|
