Adjusted acceptance_helper.rb so that the tests will pass on RBX #402
Conversation
| @@ -218,7 +218,7 @@ def _add_changes(type, changes, dst) | |||
| dst[type].sort! | |||
|
|
|||
| rescue RuntimeError => e | |||
| raise unless e.message == "can't modify frozen Hash" | |||
| raise unless e.message == "can't modify frozen Hash" || e.message == "can't modify frozen instance of Hash" | |||
|
I'd instead use: raise unless /can't modify frozen(?: instance of)? Hash/ =~ e.messageAlso, switch to RBX-3 in |
|
Just so you know: this likely won't get the tests to pass, but we'll know what the difference in lag time is. If it's too big, it means something is broken. If it's small, it could just be that Travis is doing too much work (e.g. from having too many Listen builds running simultaneously and running out of threads and/or disk access bottlenecks). Have you tried running the tests locally? |
|
The tests pass locally with this change. At least now... |
Then it was probably just Travis not being fast enough. The lag value is configurable: https://github.com/guard/listen/blob/master/.travis.yml#L28 Ideally, it's best to know what the lag on Travis is first, before increasing it. |
|
Tests now pass on RBX 3 so I don't think this PR is needed anymore? |
|
On current master branch's https://travis-ci.org/guard/listen/jobs/387772775 At least I think RBX has the different message that makes this issue. |
|
What we can is
|
|
My pull-request to RBX to fix the difference of the message was merged to master branch. |
| @@ -218,7 +218,7 @@ def _add_changes(type, changes, dst) | |||
| dst[type].sort! | |||
|
|
|||
| rescue RuntimeError => e | |||
| raise unless e.message == "can't modify frozen Hash" | |||
| raise unless /can't modify frozen(?: instance of)? Hash/ =~ e.message | |||
There was a problem hiding this comment.
I'm not sure why we need this in the first place? It seems we .dup each change in #_relative_path: unfrozen_copy = change.dup so that case shouldn't actually happen?
|
Please rebase and resubmit the PR if it's still relevant. |
I no longer use Ruby. |
No description provided.