Re: Proposal to replace the NS_ASSERT function in JS with console.assert

2018-03-13 Thread Brian Grinstead
This has now landed in Bug 1431050 - remaining consumers were either migrated to console.assert or changed to throw an exception. You can continue to use console.assert instead of NS_ASSERT in JS. If you find a case where console logging from chrome code doesn't do what you expect please let me kn

Re: Proposal to replace the NS_ASSERT function in JS with console.assert

2018-02-26 Thread Marco Bonardo
On Mon, Feb 26, 2018 at 7:02 PM, Brian Grinstead wrote: > console.assert doesn't throw an exception, and neither does NS_ASSERT. So I > don’t think replacing consumers with exceptions is correct if we want to keep > the current behavior. But I guess the intent for places code in bug 1431050 > i

Re: Proposal to replace the NS_ASSERT function in JS with console.assert

2018-02-26 Thread Kris Maglione
+1 Please make NS_ASSERT die. On Mon, Feb 26, 2018 at 09:37:09AM -0800, Brian Grinstead wrote: There have been some improvements to the console API for chrome callers recently: * The WebIDL implementation is available as a global in JSMs - no need to import Console.jsm (bug 1425574) * Chrome

Re: Proposal to replace the NS_ASSERT function in JS with console.assert

2018-02-26 Thread Brian Grinstead
console.assert doesn't throw an exception, and neither does NS_ASSERT. So I don’t think replacing consumers with exceptions is correct if we want to keep the current behavior. But I guess the intent for places code in bug 1431050 is to change the behavior and make those assertions throw? If that

Re: Proposal to replace the NS_ASSERT function in JS with console.assert

2018-02-26 Thread Marco Bonardo
We were already working with a contributor to remove NS_ASSERT and debug.js in https://bugzilla.mozilla.org/show_bug.cgi?id=1431050 and we were just replacing those few calls with simple exceptions. The usage is not particularly spread in the codebase and the benefits in ruining Nightly testers lif