Bug fixed: we should never be calling meteorDebugFuture.wait() in
production, so we now use process.env.METEOR_INSPECT_BRK in developement
to enable the waiting.
Lesson learned: if you call Fiber.yield() in the only running Fiber
without any other events scheduled on the event loop, the program will
immediately exit with code 0, as it should.
Closes#8817.
This is the feature that excites me most about Meteor 1.6, hands down.
Benefits include:
* Works with `meteor test[-packages] --debug-port 9229` (for tests), as
well as just `meteor debug` (for apps).
* The application process waits patiently for the debugger to attach, so
you don't have to race to open the debugger.
* The application process pauses at a location just after all server code
has been evaluated, but before any code starts executing, giving you a
chance to set reliable breakpoints anywhere in server code. This is much
better than using the `node --inspect-brk` flag, since that stops too
soon to set any useful breakpoints.
* The application server runs at full speed, so you don't have to wait
forever to hit that all-important breakpoint, and you don't lose nearly
as much time if you accidentally continue past the line of code where
the trouble is occurring.
* Even if your application is stuck in an infinite loop, you can still
attach the debugger, pause execution, and debug the loop.
* No more `node-inspector`! Instead, you can now debug your server code in
native Chrome DevTools, or several other high-quality inspector clients,
such as VS Code or WebStorm (seriously, check out the documentation:
https://nodejs.org/en/docs/inspector/#inspector-tools-clients). The list
of debuggable processes can be found at the URL chrome://inspect.
* Realistic performance and memory profiling is now possible via the
familiar DevTools interface.
* I highly recommend this Chrome extension that automatically (re)connects
to any open inspector sockets, so you don't have to keep manually
(re)attaching the debugger: http://june07.com/nim
* The implementation of `meteor debug` no longer has to proxy multiple
private/public debugger ports. Look at all that deleted code!
This new inspector is so much better than the old `node-inspector` that
I've been using the release-1.6 branch to debug problems in Meteor 1.5,
despite the risks of using Node 8, because those risks are so far
outweighed by the quality of the new debugging experience.
That said, the experience isn't perfect (yet). I welcome your feedback on
the Meteor 1.6 PR: https://github.com/meteor/meteor/pull/8728
This actually changed ages ago, in Node.js 0.11 via
fd3657610e
however the descriptive error message, which was previously
enabled with `true` as the third argument, was silently lost.
This reimplements the descriptive error message as mentioned in
https://github.com/meteor/meteor/issues/3200#issuecomment-289685677 by
@d-schiffner.
With Meteor 1.6 / Node 8, I noticed _buildLocalPackages taking multiple
seconds on initial server startup and restart, and the problem seems to be
that we call the global.gc function too often. This wasn't a problem in
previous versions of Node, as far as I know, but it makes sense to heed
the comment in tools/utils/gc.js, now that it matters.
The root of the problem was that the es5-ext npm package contains
directories called '#', e.g.
https://github.com/medikoo/es5-ext/tree/master/array/%23
These directory names were being sanitized to '' and thus ignored when
reserving paths in the Builder, which led to reservation conflicts later.
This commit fixes the problem in three different and independently
sufficient ways:
* Use files.mkdir_p instead of files.mkdir when creating parent
directories of written files.
* Replace illegal characters in sanitized paths with '_' instead of ''.
* Allow '#' in sanitized paths (only needs to be escaped in the shell, not
actually forbidden in paths).
