I have a Sidekiq customer complaining of a lot of "stuck" threads. He sends a lot of email so I suspected misbehaving SMTP servers.

The mail gem has support for timeouts here but they default to nil:

https://github.com/mikel/mail/blob/7b3e100f42f2d7738c3af7bf1909777568270b67/lib/mail/network/delivery_methods/smtp.rb#L91

Rails does not appear to set them by default:

https://github.c

Getting an error here when using the remove_promotion_line_items method. Seems like it should be using variant rather than item
https://github.com/spree/spree/blob/8ca9ec736a471b66f66d718b157be69946f8c5fc/core/app/models/spree/promotion_handler/coupon.rb#L78

• Way to disable/customise the email collect message "Give them a way to reach"

I'm sorry to post an issue that is missing crucial evidence, but some debugging of a recent config loading bug makes me suspect that the react-rails railtie.rb may be modifying the load order of initializers, and preventing the values in Rails.application.config being as expected.

Whether this happens or not appears to be partly dependent on where in the Gemfile the react-rails gem is l

Steps to Reproduce the Problem

If buffalo fails midway through generating a resource it does not undo what it has done in app.go. This is a very minor issue and was easy to resolve by deleting the duplicated resources from app.go. If I wasn't so new to go I'd be attempting to see if I could make a PR to fix this issue.

![screen shot 2018-04-18 at 10 09 49 am](https://user-images.git

I'm not sure if this is the right approach and maybe I've not understood this new part of the pipeline, yet, but:

Suppose, the major part of the app code lives inside a rails engine including all javascripts. And there are several main apps using that engine. The main apps only contain some layout changes and some minor patches.

Therefore, in order to integrate the webpacker gem into this se

We have a problem: if files aren't loaded/required we don't have branch data. If they are added through track_files we give it 0/0 branches which we show as 100% coverage (all possible branches are covered).

That math is "wrong" though here because there are branches but we don't know what they are. We should probably count total branches here as "unkown" and establish that in our "math unive

A while ago, we converted API controller specs into requests specs, see #2052. But those specs still have references from their old "life" that we can clean a little bit with a small effort.

In fact, their filenames still have the _controller_ part, which makes no sense now. Eg. solidus/api/spec/requests/spree/api/addresses_controller_spec.rb

Also, they contain references to the previo