* Fix CircleCI failures by adjusting the timing of problematic tests.
* Wait longer for Mongo to start.
* Increase lint wait time; run logs show we're close to the current timeout.
This restores the behavior of 8c70716954 by
default, with the option of disabling the prioritized file watching system
by setting METEOR_WATCH_PRIORITIZE_CHANGED explicitly to "false".
The self-tests where the environment variable is explicitly set form a
nice to-do list of tests that should be improved to be more robust to cope
with differences in file watcher timing.
Helps with #8648 and similar issues.
According to the README, this implementation is approximately 2.7 times
slower than native: https://www.npmjs.com/package/bcryptjs
Apps that wish to continue using the native bcrypt package should run
`meteor npm install --save bcrypt` in the root application directory, and
the npm-bcrypt package will prefer that implementation.
It's very useful to be able to include code only in production. It's
useful for React integration, but presumably also for many apps.
* Add prodOnly boolean flag alongside debugOnly in Package.describe
* Packages the set prodOnly to true auto-depend on isobuild:prod-only
(making them error in the old tool where prodOnly isn't supported)
* The `includeDebug` boolean build option is replaced by a string
named buildMode, which can be 'development' or 'production', just
like minifyMode.
Tested by self-test.
It used to create a directory with an underscore instead of a colon
Now, it just removes the prefix.
In cases where the name of the package has more than one colon or starts or ends
witha colon, we report an error.
It's been auto-failing for many months. If somebody wants to fix it,
feel free to revert this and fix it, but there's no reason to just
entirely stop self-test --slow from having any hope of passing forever.
It's been auto-failing for many months. If somebody wants to fix it,
feel free to revert this and fix it, but there's no reason to just
entirely stop self-test --slow from having any hope of passing forever.
This reverts commit 4a6dd52bca.
This made some tests flaky because notSpaceSensitive sometimes (but not
always!) ate the newline after the regexp.
We should just disable word wrapping in processes run by self-test
instead.
This reverts commit 4a6dd52bca.
This made some tests flaky because notSpaceSensitive sometimes (but not
always!) ate the newline after the regexp.
We should just disable word wrapping in processes run by self-test
instead.
Specifically, a "no pluginProviderPackageMap on isopack?" error would be
thrown when running publish in the following circumstances:
- You are inside an app (so it uses the app's .meteor/local/isopacks as
an IsopackCache instead of a temporary directory)
- Your package does not need to be rebuilt (so it gets read in "up to
date" mode by the IsopackCache)
- Your app has at least one cordova platform (so that
includeCordovaUnibuild was true on the cached isopack; it is always
true when building for publish)
In this case, we read the Isopack from disk but didn't keep the
pluginProviderPackageMap that we read from its isopack-buildinfo.json,
which later lead to a crash.
Fixes#3676.
