Otherwise it fails with:
```
performing a single-user installation of Nix...
directory /nix does not exist; creating it by running 'mkdir -m 0755 /nix && chown jenkins /nix' using sudo
sudo: a terminal is required to read the password; either use the -S option to read from standard input or configure an askpass helper
/tmp/nix-binary-tarball-unpack.roliyEDk9X/unpack/nix-2.19.3-x86_64-linux/install: please manually run 'mkdir -m 0755 /nix && chown jenkins /nix' as root to create /nix
```
Related to:
https://github.com/status-im/status-go/pull/4388
Signed-off-by: Alexis Pentori <alexis@status.im>
Signed-off-by: Jakub Sokołowski <jakub@status.im>
Had to be changed to `PLATFORM` due to conflict of `TARGET` with our Nix
shell setup in `status-mobile` and `status-go`.
Signed-off-by: Jakub Sokołowski <jakub@status.im>
If we don't we regularly get weird Git failures on Windows like:
```
fatal: Unable to find current revision in submodule path 'vendor/qzxing/tests/resources'
fatal: Failed to recurse into submodule path 'vendor/qzxing'
```
Or
```
stderr: fatal: Unable to create '.../.git/index.lock': File exists.
```
Which are very annoying and require manual workspace purge by infra team.
Many of you will get annoyed with resulting slower checkouts
on subsequent builds, but that's a price I'm willing to pay.
Signed-off-by: Jakub Sokołowski <jakub@status.im>
This way it actually reflects the purpose accurately and allows us to
build without debug symbols in CI without pulling in release credentials.
https://github.com/status-im/status-desktop/issues/13079
The only utility for `RELEASE` parameter is inheriting it
from the parent meta-job(`ci/Jenkinsfile.combined`).
Signed-off-by: Jakub Sokołowski <jakub@status.im>
QA needs easier access to public URLs from DigitalOcean spaces, and when
this meta-job fails the URLs are buried within child jobs and hard to access.
This not only sets the description if build fails, it sets it after
every single child build finishes.
Signed-off-by: Jakub Sokołowski <jakub@status.im>
The method of triggering new E2E builds from the Linux build makes sense
for PRs, but doesn't make as much sense for nightly or release builds.
By triggering it this way we will see the result as part of the whole
build that includes all platforms and all tests.
Signed-off-by: Jakub Sokołowski <jakub@status.im>
Since getProjectName() is not available for all types returned from
utils.parentOrCurrentBuild(), which can result in a crash.
Signed-off-by: Jakub Sokołowski <jakub@status.im>
- Necessary env variables to build the app:
STATUS_BUILD_INFURA_TOKEN
STATUS_BUILD_INFURA_TOKEN_SECRET
STATUS_BUILD_POKT_TOKEN
STATUS_BUILD_OPENSEA_API_KEY
STATUS_BUILD_ALCHEMY_ETHEREUM_MAINNET_TOKEN
STATUS_BUILD_ALCHEMY_ETHEREUM_GOERLI_TOKEN
STATUS_BUILD_ALCHEMY_ARBITRUM_MAINNET_TOKEN
STATUS_BUILD_ALCHEMY_ARBITRUM_GOERLI_TOKEN
STATUS_BUILD_ALCHEMY_OPTIMISM_MAINNET_TOKEN
STATUS_BUILD_ALCHEMY_OPTIMISM_GOERLI_TOKEN
- The list of available env variables as well as CL arguments can be seen running
the app providing `--help` argument. All env vars are prefixed with `STATUS_RUNTIME_`.
Runs all tests from Storybook directory via ctest:
- tests for Storybook itself
- UI unit tests using the same stubs mechanism as Storybook pages
- PagesValidator checking if SB pages are compilable
Closes: #12213Closes: #12437
It will default to `true` for Linux builds except release ones.
This setting can always be overriden manually for any build.
Resolves:
https://github.com/status-im/status-desktop/issues/12412
Signed-off-by: Jakub Sokołowski <jakub@status.im>
Because the environment variable name is far too generic and barely
means anything. It needs to clearly indicate that it's supposed to be
used at runtime and it changes how the Status app behaves.
Signed-off-by: Jakub Sokołowski <jakub@status.im>
After a LOT of work from @jrainville bisecting the release branch he
identified the issue to be this commit:
https://github.com/status-im/status-desktop/commit/dad8e453
ci: bump client build retries to 20
Which actually was poorly squashed by @iurimatias and contains the
commit that also adds the retries in the first place:
https://github.com/status-im/status-desktop/commit/a5ba2537
ci: retry nim_status_client build 3 times
The reason why that's an issue is because normally it was caled by `make
tgz-linux` target which included this setting:
```
override RESOURCES_LAYOUT := $(PRODUCTION_PARAMETERS)
```
Which caused `NIM_PARAMS` to use `-d:development` instead of `-d:production`.
The result was the app trying to write config file changes to the read-only
filesystem of the AppImage.
Resolves: https://github.com/status-im/status-desktop/issues/11295
Signed-off-by: Jakub Sokołowski <jakub@status.im>
Anton Iakimov
Jakub Sokołowski
Anastasiya Semenkevich
Dario Gabriel Lipicar
vlado@status.im
Vladimir Druzhinin
Igor Sirotin
Sale Djenic
Michał Cieślak