I've not been able to test this locally but it compiles in Xcode 11.2 so might work for Xcode 9.2?

When adding ~/.yarn/bin to fish_user_paths the output message was inaccurate. This also means that "Please open another terminal where the yarn command will now be available." is not accurate (it will be available immediately), although I am unsure of the best way to structure the code to alter this output.

When copy/pasting this command in fish (untested in other shells) the shell will try to expand the URL due to the ? in the URL (No matches for wildcard 'https://sentry.io/api/0/projects/[...]/processingissues/fix?_=[...]'. See `help expand`.)

This was raised in support request 22916. Colleen was kind enough to point me to this line of code.

I am not totally familiar with JSX syntax so I hope that simply adding quotes works and does not need any escaping.

I have branched off of https://github.com/nerdishbynature/RequestKit/pull/63 to (hopefully) fix CI, and add support for Swift 5.1 on Linux (by importing FoundationNetworking, as per https://forums.swift.org/t/foundationnetworking/24769)

I have added Dependabot to one of my projects, and Danger Swift is failing because the PR was created by a bot user: https://github.com/JosephDuffy/Partial/pull/14, https://travis-ci.com/JosephDuffy/Partial/jobs/211151759.

I used the GitHub API to request the data for the above PR and then used the value of theuser key to create a test for decoding this type of user.

It looks like the documentation should be updated to reflect the new UserType case. I have run make docs, but it only updates half of the docs, eventually failing in ./Scripts/update_docs.rb with:

The operation couldn’t be completed. (SourceKittenFramework.Request.Error error 2.)
/System/Library/Frameworks/Ruby.framework/Versions/2.3/usr/lib/ruby/2.3.0/json/common.rb:156:in `initialize': A JSON text must at least contain two octets! (JSON::ParserError)
	from /System/Library/Frameworks/Ruby.framework/Versions/2.3/usr/lib/ruby/2.3.0/json/common.rb:156:in `new'
	from /System/Library/Frameworks/Ruby.framework/Versions/2.3/usr/lib/ruby/2.3.0/json/common.rb:156:in `parse'
	from ./Scripts/update_docs.rb:27:in `<main>'

It would appear that SourceKitten is throwing an error, but I am unable to figure out how to debug/fix this! :(

Previously the global path was being passed, but the configuration is created in the local scope of the data source

As discussed in #1179 this PR updates the release Fastlane lane to upload a zip archive containing prebuild versions of Kingfisher to GitHub

Currently only the mapping is done, still all the UI etc. to be done

Carthage expects a single .zip file containing all frameworks for all platforms. The current method means that only the FBSDKCoreKit framework will be downloaded (when not using --no-use-binaries), this should allow them to all be downloaded automatically by Carthage.

Thanks for proposing a pull request.

To help us review the request, please complete the following:

• sign contributor license agreement
• describe the change (for example, what happens before the change, and after the change)

Classing as a draft since there are likely a few issues and this hasn't really been used "in the wild" yet.

But anyway I think this is the basic required.

The general idea is that the EmbeddedDataSource implements a few more of the delegate methods around sizing, and has a custom sizing strategy. Overall the API is pretty much the same (it needs to be told the sizing strategy for the embedded data source but I think that's it).

One odd thing is that the data source you pass in to the embedding one will contain a sizing strategy, but it will be largely (maybe entirely?) ignored, but it still needs to provide UI because it's going to be providing the cells themselves.

I'm not 100% about the lack of support for optionals here.

On one hand I think it could be up to the caller to not create this with a nil element, but on the other hand I can see it might be useful to return 0 for numberOfSections and numberOfElements(in:) when the value is nil 🤔

This allows for an init(elements:) initialiser