Our style guide includes formatting conventions. Instead of wasting time in reviewing PRs discussing things like indenting, and to avoid later deltas to fix bad formatting in earlier commits, we want to enforce these in the build.
The intent in this change is to fail the build as quickly as possible in travis, so those sending a PR immediately know they should run clang-format and update their commit. When running locally, we want users to know about formatting, but they may not want to act on it immediately, until they are done working. For this reason, it is only a warning outside of the continuous build.
This is done by having a check-format task which should run on most local builds, and an enforce-format task only run by travis.
Now, running protractor configs by default only runs e2e tests. If
the --benchmark flag is added, it runs only the perf tests, and always
restarts the browser in between tests. If the --dryrun test is added,
the perf tests are run only once.
This should make it easier to run perf tests versus example e2e tests,
and help stabilize the travis build because perf tests always
run with a clean browser.
* `npm install` now does a full install; auxiliary installation steps
have been integrated into the `postinstall` script.
* Updated developer docs `DEVELOPER.md` accordingly; also added
instructions to dev docs for performing full tests (via `npm test`) --
same as those run on Travis.
* Reorg in tests so that JS tests can run without a Dart env.
Partly fixes#945 **under the assumption that when running JS tests
locally, `ChromeCanary` is the desired browser to use**. Note that CI
tests (Travis) still uses `DartiumWithWebPlatform` across the board
(Maybe because ChromeCanary isn't being installed?)
Fixes#1012.
Closes#1010
Performed a slight refactoring of CI scripts to make it easier for
developers to run the **same** tests as those run on Travis. Defined
`npm` scripts `test-js` and `test-dart`. `npm test` now runs the whole
lot.
Closes#966
- use performance log of chromedriver / appium to get timeline data
for calculating metrics for benchmarks
- change all benchmarks to be made of a standalone application
and a protractor test that collectes timeline data
- fix and simplify benchmarks
- add dart2js to build
- remove benchpress
Closes#330
simplify:
- use same html file for dart and JS
- build benchmarks automatically when doing `gulp build`
- centralize configuration
modularize:
- move all build tasks into separate node.js modules under
`tools/build`.
changes:
- the `build` folder is now the `dist` folder
Closes#284
I don’t know what the issue really is. There is some weird race condition,
where the task copies the file (`pubspec.yml`) and then start `pub get`
which reads that file and the file is not there or is empty.
I have no idea why moving this task into a separate process fixes the
issue but I don’t wanna waste more time on it.
Yegor Jbanov
Alex Eagle
Victor Berchet
Julie Ralph
Patrice Chalin
Marc Laval
Tobias Bosch
Kevin Moore
Misko Hevery
vsavkin
Vojta Jina