[workflow] add markdown link checker

[jessesquires]

[workflow] add markdown link checker

[hal-codebot]

Proselint found issues

_posts/2014-06-25-apples-to-apples.md

Line	Message	Severity
6	Minimize your use of corporate catchphrases like this one.	warning
11	Months should be capitalized. 'June' is the preferred form.	warning
207	More than 30 ppm of exclamations. Keep them under control.	warning
234	'Hopefully' is a bit of a skunked term — impossible to use without issue. Find some other way to say it.	warning

_posts/2014-08-06-apples-to-apples-part-two.md

Line	Message	Severity
6	Minimize your use of corporate catchphrases like this one.	warning
16	Minimize your use of corporate catchphrases like this one.	warning
88	Days of the week should be capitalized. 'Friday' is the preferred form.	warning
188	More than 30 ppm of exclamations. Keep them under control.	warning
248	Minimize your use of corporate catchphrases like this one.	warning

_posts/2014-08-19-on-the-value-of-benchmarks.md

Line	Message	Severity
10	Minimize your use of corporate catchphrases like this one.	warning
14	No paragraph should start with a 'But'.	warning
19	'with regard to' is commercialese.	warning

_posts/2014-08-21-apples-to-apples-part-three.md

Line	Message	Severity
6	Minimize your use of corporate catchphrases like this one.	warning
14	Minimize your use of corporate catchphrases like this one.	warning
138	'with regard to' is commercialese.	warning

_posts/2014-10-22-swift-failable-initializers.md

Line	Message	Severity
81	'hopefully' is a bit of a skunked term — impossible to use without issue. Find some other way to say it.	warning

_posts/2014-12-08-introducing-jsqmessagesvc-6-0.md

Line	Message	Severity
26	Substitute 'damn' every time you're inclined to write 'very'; your editor will delete it and the writing will be just as it should be. Found 6 times elsewhere.	warning
48	'easier said than done,' is a cliché.	warning
48	'easier said than done,' is a cliché.	warning
48	Needless variant. 'extendable' is the preferred form.	warning
82	'Hopefully,' is a bit of a skunked term — impossible to use without issue. Find some other way to say it.	warning
82	Redundancy. Use 'same' instead of 'exact same'.	warning
94	'Needless to say,' is a cliché.	warning

_posts/2015-01-05-swift-coredata-and-testing.md

Line	Message	Severity
33	Substitute 'damn' every time you're inclined to write 'very'; your editor will delete it and the writing will be just as it should be. Found 3 times elsewhere.	warning
37	No hyphen is necessary in phrasal adjectives with an adverb ending in -ly, unless the -ly adverb is part of a longer phrase	warning
94	'Back to the drawing board.' is a cliché.	warning
94	Minimize your use of corporate catchphrases like this one.	warning

_posts/2015-02-17-better-coredata-models-in-swift.md

Line	Message	Severity
71	'hopefully' is a bit of a skunked term — impossible to use without issue. Find some other way to say it.	warning
93	Substitute 'damn' every time you're inclined to write 'very'; your editor will delete it and the writing will be just as it should be.	warning
183	Redundancy. Use 'free' instead of 'for free'.	warning
183	'for free' is cliché.	warning

_posts/2015-03-31-functional-notifications.md

Line	Message	Severity
10	Substitute 'damn' every time you're inclined to write 'very'; your editor will delete it and the writing will be just as it should be.	warning
135	'deceptively' is a bit of a skunked term — impossible to use without issue. Find some other way to say it.	warning

_posts/2015-07-19-swift-namespaced-constants.md

Line	Message	Severity
11	Days of the week should be capitalized. 'Friday' is the preferred form.	warning
16	More than 30 ppm of exclamations. Keep them under control.	warning
31	No hyphen is necessary in phrasal adjectives with an adverb ending in -ly, unless the -ly adverb is part of a longer phrase	warning
55	Hedging. Just say it.	warning
83	'(hopefully)' is a bit of a skunked term — impossible to use without issue. Find some other way to say it.	warning
87	'with regard to' is commercialese.	warning

_posts/2015-07-26-swift-enumerations-and-equatable.md

Line	Message	Severity
15	Redundancy. Use 'free' instead of 'for free'.	warning
15	'for free' is cliché.	warning
56	No hyphen is necessary in phrasal adjectives with an adverb ending in -ly, unless the -ly adverb is part of a longer phrase	warning
70	Substitute 'damn' every time you're inclined to write 'very'; your editor will delete it and the writing will be just as it should be.	warning

_posts/2015-10-14-UIKit-changes-in-iOS-9.md

Line	Message	Severity
16	Days of the week should be capitalized. 'Friday' is the preferred form.	warning
22	'playing with fire.' is a cliché.	warning
24	'Playing with fire' is a cliché.	warning
44	Days of the week should be capitalized. 'Friday' is the preferred form.	warning

_posts/2015-10-25-building-data-sources-in-swift.md

Line	Message	Severity
16	'time and time again.' is a cliché.	warning
99	More than 30 ppm of exclamations. Keep them under control.	warning
107	'hopefully' is a bit of a skunked term — impossible to use without issue. Find some other way to say it.	warning

_posts/2015-12-06-swift-open-source.md

Line	Message	Severity
24	Substitute 'damn' every time you're inclined to write 'very'; your editor will delete it and the writing will be just as it should be. Found 4 times elsewhere.	warning
36	More than 30 ppm of exclamations. Keep them under control.	warning
40	With lowercase letters, the periods are standard.	warning
40	It's standard to put a space before 'a.m.' or 'p.m.'.	warning

_posts/2015-12-10-open-source-swift-weekly-1.md

Line	Message	Severity
10	More than 30 ppm of exclamations. Keep them under control.	warning
20	Substitute 'damn' every time you're inclined to write 'very'; your editor will delete it and the writing will be just as it should be.	warning
24	Annotation left in text.	warning

_posts/2015-12-17-open-source-swift-weekly-2.md

Line	Message	Severity
10	More than 30 ppm of exclamations. Keep them under control.	warning
54	Substitute 'damn' every time you're inclined to write 'very'; your editor will delete it and the writing will be just as it should be.	warning
54	Comparison of an uncomparable: 'least only ' is not comparable.	warning
54	Inconsistent spacing after period (1 vs. 2 spaces).	warning

_posts/2015-12-24-open-source-swift-weekly-3.md

Line	Message	Severity
10	More than 30 ppm of exclamations. Keep them under control.	warning
30	Needless variant. 'extendable' is the preferred form.	warning

_posts/2016-01-07-open-source-swift-weekly-4.md

Line	Message	Severity
10	More than 30 ppm of exclamations. Keep them under control.	warning

_posts/2016-02-06-jsqmessages-call-for-contributors.md

Line	Message	Severity
7	More than 30 ppm of exclamations. Keep them under control.	warning

_posts/2016-02-19-swifty-presenters.md

Line	Message	Severity
16	More than 30 ppm of exclamations. Keep them under control.	warning

_posts/2016-03-21-contributing-to-swift.md

Line	Message	Severity
7	More than 30 ppm of exclamations. Keep them under control.	warning

_posts/2016-05-20-swift-documentation.md

Line	Message	Severity
85	Redundancy. Use 'free' instead of 'for free.'.	warning
85	'for free.' is cliché.	warning
87	No paragraph should start with a 'But'.	warning
91	Redundancy. Use 'free' instead of 'for free,'.	warning
91	'for free,' is cliché.	warning

_posts/2016-05-22-open-source-everything.md

Line	Message	Severity
52	Annotation left in text.	warning
52	Hyphenate 'open source community', a phrasal adjective, as 'open-source community'.	warning
58	'Easier said than done' is a cliché.	warning
58	'Easier said than done' is a cliché.	warning
62	'easier said than done.' is a cliché.	warning
62	'easier said than done.' is a cliché.	warning

_posts/2016-06-04-avoiding-objc-in-swift.md

Line	Message	Severity
96	No paragraph should start with a 'But'.	warning

_posts/2016-07-31-enums-as-configs.md

Line	Message	Severity
10	Needless variant. 'extendable' is the preferred form.	warning
88	More than 30 ppm of exclamations. Keep them under control.	warning
216	Needless variant. 'extendable' is the preferred form.	warning

_posts/2016-10-01-shipping-swift-3.md

Line	Message	Severity
10	More than 30 ppm of exclamations. Keep them under control.	warning

_posts/2016-11-05-140-proposals-frenchkit-video.md

Line	Message	Severity
26	Hyphenate 'open source community,', a phrasal adjective, as 'open-source community'.	warning

_posts/2017-01-16-testing-without-ocmock.md

Line	Message	Severity
16	More than 30 ppm of exclamations. Keep them under control.	warning
63	No paragraph should start with a 'But'.	warning

_posts/2017-02-10-refactoring-singletons-in-swift.md

Line	Message	Severity
138	'If and when' is a cliché.	warning

_posts/2017-02-15-adapting-to-change-and-cutting-corners.md

Line	Message	Severity
18	'[start from scratch]' is a cliché.	warning

_posts/2017-06-04-writing-better-singletons-in-swift.md

Line	Message	Severity
26	Redundancy. Use 'add' instead of 'add an additional'.	warning

_posts/2017-07-16-officially-deprecating-jsqmessagesviewcontroller.md

Line	Message	Severity
45	Substitute 'damn' every time you're inclined to write 'very'; your editor will delete it and the writing will be just as it should be.	warning
47	'off' is the preferred form.	warning

_posts/2017-07-17-a-story-about-swift-source-compat.md

Line	Message	Severity
16	'few and far between,' is a cliché.	warning

_posts/2017-09-10-building-a-site-with-jekyll-on-nfsn.md

Line	Message	Severity
8	'off' is the preferred form.	warning
11	'off' is the preferred form.	warning
11	'hopefully' is a bit of a skunked term — impossible to use without issue. Find some other way to say it.	warning
34	Inconsistent spacing after period (1 vs. 2 spaces).	warning
182	'off' is the preferred form.	warning

_posts/2017-09-18-measuring-compile-times-xcode9.md

Line	Message	Severity
41	Substitute 'damn' every time you're inclined to write 'very'; your editor will delete it and the writing will be just as it should be.	warning

_posts/2017-09-24-when-your-app-is-used-in-unexpected-ways.md

Line	Message	Severity
35	No hyphen is necessary in phrasal adjectives with an adverb ending in -ly, unless the -ly adverb is part of a longer phrase	warning

_posts/2017-10-01-floating-point-swift-ulp-and-epsilon.md

Line	Message	Severity
303	Redundancy. Use 'skipped' instead of 'skipped over'.	warning
303	'Hopefully' is a bit of a skunked term — impossible to use without issue. Find some other way to say it.	warning

_posts/2017-12-28-swift-weekly-brief-hiatus.md

Line	Message	Severity
33	Substitute 'damn' every time you're inclined to write 'very'; your editor will delete it and the writing will be just as it should be. Found once elsewhere.	warning
41	'First and foremost,' is cliché.	warning
51	'off' is the preferred form.	warning
51	More than 30 ppm of exclamations. Keep them under control.	warning

_posts/2018-07-05-swifts-new-calling-convention.md

Line	Message	Severity
16	'if and when' is a cliché.	warning

_posts/2018-07-29-swift-tip-building-arrays-with-compactmap.md

Line	Message	Severity
32	No paragraph should start with a 'But'.	warning
84	Substitute 'damn' every time you're inclined to write 'very'; your editor will delete it and the writing will be just as it should be. Found once elsewhere.	warning

Spell Checker found issues

README.md

Line	Typo
1	# [jessesquires.com](https://www.jessesquires.com
11	](https://getbootstrap.com), **jQuery,** and [Fo
11	/fontawesome.com). Hosted at [NearlyFreeSpeech](https://nearlyfreespeech.net
15	- **Bundler**
22	- **jekyll** ([Late
23	- [jekyll-sitemap](https://github.com/jekyll/je
24	- [jekyll-paginate](https://github.com/jekyll/je
35	- **jQuery** ([pkg](h
46	- [GoatCounter Analytics](https://www.goatco
46	- GoatCounter **Analytics**
97	> Copyright © 2014-present Jesse Squires.

_posts/2014-06-25-apples-to-apples.md

Line	Typo
11	When Craig Federighi arrived at his presentation s
15	{% include updated_notice.html
16	update_message='
30	{% include blockquote.html
31	quote="And then Federighi said, let there be Swift; and
35	The third floor of Moscone West erupted with applause as
37	is still in beta (along with Xcode 6, iOS 8, and OS X 10.10), so
39	s a fun and interesting code **kata,** I deci
39	a.com), I decided to port my [objc-sorts](https://github.com/jessesqui
44	Mavericks 10.9.3 10.9.4, Xcode6-beta4 beta2 WWDC seed
44	.9.3~~ 10.9.4, Xcode6-beta4 **beta2** WWDC seed
45	* Hardware: 2008 unibody MacBook Pro, 2.4 Ghz Intel Co
45	* Hardware: 2008 unibody MacBook Pro, 2.4 Ghz Intel Core 2 Duo
45	2008 unibody MacBook Pro, 2.4 Ghz Intel Core 2 Duo, 8 GB 1067 M
51	wift's Array and Int. The APIs are language-specific too, fo
75	in seconds. The average case runtime complexity for each algorithm
199	on level for Swift and -Ofast, though faster, remove
199	tire app compiled with -Ofast. More on that below.</
214	no optimizations and -Ofast in Swift is stark. On
216	aster though. Std lib sort is 6.5x faster. Quick sort is 7.0x fa
216	is 6.5x faster. Quick sort is 7.0x faster. Heap sort is 10.4x fa
216	is 7.0x faster. Heap sort is 10.4x faster. Insertion sort is 16.
216	.4x faster. Insertion sort is 16.0x faster. Selection sort is 2.4
216	.0x faster. Selection sort is 2.47x faster.
218	3. Only with -Ofast do we begin to experie
218	ve-C is almost twice as fast (1.84x). However, when comparing Swi
220	e this (when not using -Ofast, see Table 1 and *Ta
220	dash; in debug Objective-C is 308.0x faster. I'm still puzzled by
222	algorithm to use based on the dataset. I suspect that we would see
222	orts perform best with larger datasets and/or with complex objects.
224	eeing different results here. Federighi noted that for complex object
224	ort, Objective-C performed at 2.8x and Swift performed at 3.9x,
224	t 2.8x and Swift performed at 3.9x, using Python as the baseline
224	using Python as the baseline (1.0x). It is not clear at this tim
230	es are regarding using -Ofast. They recommended the
230	mpile the module using -Ofast and link it to your ap
234	eleases updates for Swift and Xcode6-beta.**
238	### Futher reading
241	w.splasmata.com/?p=2798) from Splasm Software
242	-7/collections.html) by Peter Steinberger, objc.io issue #7
242	s.html) by Peter Steinberger, objc.io issue #7

_posts/2014-08-06-apples-to-apples-part-two.md

Line	Typo
14	ith the fifth beta release of Xcode 6, iOS 8, OS X Yosemite, and
36	in seconds. The average case runtime complexity for each algorithm
38	conds, then Swift is 3 times (3x) faster.
88	tantially better. But this is ok — since you will be shi
188	optimization level of -Ofast has been deprecated in
188	en deprecated in favor -Ounchecked. This renaming is a de
188	. Previously it read as, "Ooooh fast!", begging for misu
192	f not all, developers will be compliling their modules and apps at the

_posts/2014-08-19-on-the-value-of-benchmarks.md

Line	Typo
14	echanisms of the behavior of Caenorhaditis elegans (i.e. the nematode w
14	he behavior of Caenorhaditis elegans (i.e. the nematode worm) by
14	r of Caenorhaditis elegans (i.e. the nematode worm) by [embedd
14	by [embedding a model of its connectome in a simplified body and envi
17	inform our decisions on which toolsets to use. Further, every app wi

_posts/2014-08-21-apples-to-apples-part-three.md

Line	Typo
10	If you abandon your favorite runtime, forget about garbage collect
14	lcome this week's arrival of [Xcode 6 beta 6](https://developer.a
19	ware:* OS X Mavericks 10.9.4, Xcode 6 beta 6
20	* Hardware: 2008 unibody MacBook Pro, 2.4 Ghz Intel Co
20	* Hardware: 2008 unibody MacBook Pro, 2.4 Ghz Intel Core 2 Duo
20	2008 unibody MacBook Pro, 2.4 Ghz Intel Core 2 Duo, 8 GB 1067 M
22	tegers in the range [0, UINT32_MAX). Then, each sorting a
24	while C uses qsort() from [cstdlib](http://www.cplusplus.com/ref
28	onds, then Swift is 3 times (-3x) slower. Also note that the d
82	its slowest, Swift was still 6x faster than Objective-C, whic
138	e a choice with regard to our toolsets. Programming languages come i

_posts/2014-10-22-swift-failable-initializers.md

Line	Typo
10	new feature in Swift 1.1 in [Xcode 6.1](https://developer.apple.
10	om/xcode/downloads/) — failable initializers.
10	downloads/) — failable initializers.
14	#### Failable initializers
14	#### Failable initializers
21	nil, with the exception of [optionals](https://developer.apple.com/
21	he syntactic difference for a failable initializer — from `ini
21	tic difference for a failable initializer — from init() to `ini
23	e should not be a need for a failable initializer. However, this i
23	ot be a need for a failable initializer. However, this is a great fea
23	ng so great. The problem with failable initializers is, ironically,
23	at. The problem with failable initializers is, ironically, nil. In Obj
23	o an array or a dictionary at runtime. We all loathe that embarassm
23	t runtime. We all loathe that embarassment. But Swift comes to the rescu
23	erally speaking, we must [use optionals with care](http://owensd.io/2
25	The issue with failable initializers in Swift is the
25	The issue with failable initializers in Swift is the same issue wi
27	#### Be failable, not fallible
29	ons" is the language's latest tagline, which is probably as frighte
29	Swift's best practices. Using failable initializers in Swift should
29	est practices. Using failable initializers in Swift should be done spari
33	However, failable initializers might seduce you
33	However, failable initializers might seduce you into doing s
60	g a non-optional image to our initializer.
79	constructing a UIImage, the initializer `init(named name: String) ->
79	n to dirty up this class with optionals.
81	and just plain shameful with failable initializers if we aren't car
81	plain shameful with failable initializers if we aren't careful. Imagine
83	#### A fallible place to be failable
85	e at the core of the problem: failable initializers misplace respons
85	core of the problem: failable initializers misplace responsibilities in
85	lace the point of failure. An initializer should construct an instance.
87	One great solution would be a validator class, `MyBetterPostValidator
87	ass would also not have a failable intializer, but instead it wo
87	also not have a failable intializer, but instead it would have a
91	Failable initializers do address some
91	Failable initializers do address some of our proble
91	ey introduce into our design. Failable initializers feel a lot like
91	uce into our design. Failable initializers feel a lot like Objective-C,
91	tage of Swift's new features (enums, tuples, generics, etc.) and
93	In a way, before failable initializers, Swift helped en
93	In a way, before failable initializers, Swift helped enforce the [SO

_posts/2014-12-08-introducing-jsqmessagesvc-6-0.md

Line	Typo
24	It all began with **Hemoglobe,** an a
24	Hub](https://github.com) and [CocoaControls](https://www.cocoacontrols.co
26	**CocoaPods** was ju
26	/3) was opened that requested CocoaPod support and I received my [fi
26	wController/pull/4) to add a [podspec](http://guides.cocoapods.org/
34	tion view and has a number of customizable subviews and properties, whic
34	has a number of customizable subviews and properties, which are out
34	g diagram. The labels for the subviews in the diagram are the names
34	m are the names of the actual subview properties for a cell. See th
34	operties for a cell. See the [JSQMessagesCollectionViewCell](http://cocoadocs.org/docsets
36	{% include image.html
37	file="jsqmessages_cell_anatomy.jpg"
40	source_link=null
41	half_width=false
44	has three labels for message metadata — the sender, the date,
44	optional. Finally, there's a customizable margin (`messageBubbleLeftRig
44	layout object. Each of these subviews is customizable via the usual
44	ct. Each of these subviews is customizable via the usual [UIKit](https:/
44	s customizable via the usual [UIKit](https://developer.apple.com/
44	t/Reference/UIKit_Framework/) APIs. Now that we understand the b
48	one, and I've made plenty of [embarassing mistakes](https://github.com/
59	sage data (text or media) and metadata (sender, date, etc.)
70	ample, media data could be a [CLLocation](https://developer.apple.com/
70	tion_Class/index.html), or a [UIImage](https://developer.apple.com/
70	age_Class/index.html), or an [ABRecordRef](https://developer.apple.com/
72	{% include image.html
73	file="jsqmessages_media_type.jpg"
74	alt="JSQMessage Media Items"
76	source_link=null
77	half_width=false
82	message or media types as an [Enum](http://nshipster.com/ns_enum
82	rovided media types. Using an enum would litter the code base wi
82	of media. The only option for extensibilty would have been providing a "
82	ave been providing a "custom" enum case, such as `JSQMediaTypeCu
82	umer of the API. This special enum case further increases comple
90	y good state. It's stable and extensbile, but there's still plenty of
96	reat developers, I discovered CocoaPods, I've learned how to effectiv

_posts/2015-01-05-swift-coredata-and-testing.md

Line	Typo
22	Because Swift classes are namespaced, you must prefix the class na
22	your .xcdatamodeld file in Xcode, selecting an entity, and ope
33	nerate the classes, otherwise Xcode will not create the classes p
37	ts into Core Data and avoid [stringly-typed Objective-C](http://corner.s
78	unction from the [Objective-C runtime library](https://developer.ap
78	ve always avoided using such [runtime](http://nshipster.com/associa
102	ames, like "Person", to the initializer
106	iteria above is a convenience initializer:
121	class function. In Swift, an initializer is guaranteed to return a non
123	to implement this convenience initializer and provide their own value f
123	his by moving the convenience initializer to a new extension and replac
123	ill-not-work) due to Swift's [initializer delegation and two-phase init
125	eventually be automated via [mogenerator](https://github.com/rentzsch/

_posts/2015-02-17-better-coredata-models-in-swift.md

Line	Typo
16	ckily, we can harness Swift's optionals, enums, and other features to
16	an harness Swift's optionals, enums, and other features to make m
16	rawbacks and limitations with Xcode's current toolset.
16	itations with Xcode's current toolset.
20	### A great case for optionals
22	html) in the community about [optionals](https://developer.apple.com/
26	g the .xcdatamodeld file in Xcode, then selecting the entity, t
26	pector in the sidebar. Or, at runtime you find out that your `NSMan
69	muted">Note: Xcode does not generate Swift class
71	the rest of your app. As with optionals in general, you will need to
71	pleasant if you have a lot of optionals. If this is the case, hopeful
71	hem to be dismissed before. Optionals complicate your model, whic
73	equired to have an address. Ahh, this is looking much better
93	rtype. In the depths of the **codebase,** there may be operations on
115	As noted in **objc.io,** *we can
183	y. On the bright side, we get optionals and type aliases for free &md
185	ctive-C. Here's to hoping the toolset will improve — and when

_posts/2015-03-31-functional-notifications.md

Line	Typo
10	ation of this pattern is via [NSNotificationCenter](https://developer.apple.com/
10	r, the NSNotificationCenter APIs are kind of cumbersome to use
16	A while back, **objc.io** posted f
45	object (id) property and userInfo (NSDictionary) property. I
50	Say hello to [JSQNotificationObserverKit](https://github.com/jessesqui
50	it), there is only 1 class, 1 struct, and 1 function. This is all
52	eating a Notification. This struct has two type parameters: a va
87	r is set to nil (when it is deallocated), then it is removed from `NS
96	een resolved. Registering and unregistering for a notification is now as
100	oa notifications that send a userInfo dictionary? We can do it all
105	oes not send any data in the userInfo dictionary.
131	the unit tests included with [JSQNotificationObserverKit](https://github.com/jessesqui
135	so little — 1 class, 1 struct, and 1 function. An equivalen
135	l not type-safe. I think the [objc.io snippets](http://www.**objc.io**/
135	ro-libraries, check out Chris Eidhof's talk on [Tiny Networking](h

_posts/2015-04-06-swift-failable-initializers-revisited.md

Line	Typo
10	ost, I discussed how Swift's [failable initializers](https://develop
10	scussed how Swift's [failable initializers](https://developer.apple.com/
10	fail. We should aim to avoid optionals as much as possible to reduce
10	sence of values. Recently, @danielgomezrico asked a great [question](ht
10	out a possible use case for a failable initializer — parsing J
10	sible use case for a failable initializer — parsing JSON. [Given]
16	del object. Should we write a failable initializer for this model th
16	t. Should we write a failable initializer for this model that receives
46	XML? Then we would need a new initializer, init?(xml: XML), and the m
54	st step is creating a generic validator object. We'll use a [phantom
54	ets/13.html) to ensure that a validator can only validate the JSON fo
54	e of model. We initialize the validator with a closure that receives
54	is saved as a property on the validator.
81	e us to construct many unique validators, while maintaining a single g
81	reate many different concrete validators (or validator subclasses) for
81	erent concrete validators (or validator subclasses) for many differen
81	safety and readability to the validator. We know that this validator
81	validator. We know that this validator is for MyModel instances.
83	new (more proper) designated initializer to MyModel that uses [depen
110	tinue. With this solution, no failable initializers are required.
110	th this solution, no failable initializers are required.
132	ided the problem into smaller subproblems and addressed each one indivi
132	ponents above — JSON, a validator, and a parser — and ret
154	Suffering from [angle-brack-T blindness](http://inessential
154	specifies the model type. The validator must be a validator for T,
154	type. The validator must be a validator for T, and the function ret
160	easily testable code, and no failable initializers. Of course there
160	estable code, and no failable initializers. Of course there will be situ
160	t be instantiated and using a failable initializer is perfect. In th
160	antiated and using a failable initializer is perfect. In this situation
162	ughtful design that will push optionals and failure states further to

_posts/2015-07-19-swift-namespaced-constants.md

Line	Typo
11	stants-and-functions.html) on namespaced constants and functions in C.
11	xes. Although Swift types are namespaced by their module, we can still
15	{% include updated_notice.html
16	update_message='Inspired by this post, <a hr
16	a Swift source file based on Xcode assets. Checkout the repo <a
16	on Xcode assets. Checkout the repo <a href="https://github.com/l
16	utoAssets" class="alert-link">AutoAssets on GitHub. Thanks Lee!'
23	{% include image.html
24	file="tabbar.png"
27	source_link="https://developer.apple.com/
28	half_width=false
31	the past, we strove to avoid [stringly-typed](https://corner.squareup.com/
53	we are using the new [default enum naming](http://ericasadun.com
53	stuff-in-xcode-7-beta-3/) for enums with a string raw type. This
55	. But the goal here is proper namespacing, which makes our code easier
55	essentially in the UIImage namespace and I would argue that this f
55	a new type, we create our own namespace. We immediately know we are d
55	on UIImage — was it musicIcon and musicSelectedIcon, or
55	dash; was it musicIcon and musicSelectedIcon, or iconMusic and *iconMus
55	and musicSelectedIcon, or iconMusic and iconMusicSelected? Ins
55	tedIcon*, or iconMusic and iconMusicSelected? Instead, we can simply begi
59	be nice to be able to use an enum in this scenario, but an enum
59	enum in this scenario, but an enum raw value type must be a **va
59	r`. Alternatively, we can use structs and nested structs.
59	we can use structs and nested structs.
83	ars (hopefully). Using nested structs gives us our own namespace wi
83	sted structs gives us our own namespace with much richer semantics, a
87	ces in our apps: storyboards, xibs, or sound effects. You could
87	effects. You could even nest enums inside of structs, or enums i
87	uld even nest enums inside of structs, or enums inside of enums. So
87	t enums inside of structs, or enums inside of enums. So far, I've
87	f structs, or enums inside of enums. So far, I've found this patt
87	r, I've found this pattern of namespaced constants to be extremely use
91	what I have described, like [swiftrsrc](https://github.com/indragiek

_posts/2015-10-14-UIKit-changes-in-iOS-9.md

Line	Typo
10	2). (Well, actually what [Max von Webel](https://twitter.com/34
10	(Well, actually what [Max von Webel](https://twitter.com/343max/s
10	There's a hidden gem in the [UIKit diffs](https://developer.appl
18	when the pointed-to object is deallocated, the pointer is zeroed out or
26	matter? Because up to iOS 9, UIKit views that have a delegate
40	this extends well beyond UITableView and UICollectionView. UIBarBu
40	s well beyond UITableView and UICollectionView. UIBarButtonItem.target is as
40	bleView and UICollectionView. UIBarButtonItem.target is assign. UIGestureRecognize
40	rButtonItem.target is assign. UIGestureRecognizer.delegate is assign. UIActionSheet.dele
40	ecognizer.delegate is assign. UIActionSheet.delegate is assign. UIAccelerometer.de
40	tionSheet.delegate is assign. UIAccelerometer.delegate is assign. The list goes on,
40	e list goes on, everywhere in UIKit.
42	urceordelegate` object is deallocated before the view, and then t
46	>You're correct that __unsafe_unretained is the same as a regular assi
46	ers in a window controller's -dealloc. It's almost unheard of. And
50	of iOS 9, the aforementioned APIs have changed. In fact, all of
50	have changed. In fact, all of UIKit looks like it [has been audit
64	s. Swift was the impetus for [nullability annotations](https://develope

_posts/2015-10-25-building-data-sources-in-swift.md

Line	Typo
10	iewandUITableView`. These APIs make it simple to build inter
10	h which we can approach these APIs and reimagine how we use them
10	e can approach these APIs and reimagine how we use them to build apps
16	— re-implementing the [UITableViewDataSource](https://developer.apple.com/
16	/UITableViewDataSource) and [UICollectionViewDataSource](https://developer.apple.com/
20	d just recently updated) was [JSQDataSourcesKit](https://github.com/jessesqui
20	hese issues. Inspired by Andy Matuschak's [gist](https://gist.github.co
32	ctionandTableViewSection` structs. However, they conform to the
35	yandTableViewCellFactory**structs,** which conform to theCollec
37	ods from Objective-C classes (i.e., NSObject subclasses) to pu
103	ntegrates with Core Data and [NSFetchedResultsController](https://developer.apple.com/
103	/tree/develop/Example) in the repo that exercises all functional
107	n Swift, I encourage you try [JSQDataSourcesKit](https://github.com/jessesqui

_posts/2015-12-06-swift-open-source.md

Line	Typo
24	Swift going to be dumped on [opensource.apple.com](http://www.opensource.apple.
24	uld it be put on GitHub like [ResearchKit](https://github.com/ResearchK
24	thing you need to know is on **Swift.org.**
30	- Chris Lattner's [first commit](https://github
32	- The main [Swift repo](https://github.com/apple/swi
34	ll requests across all of the repos. Many of them accepted and me
36	tions and more — [Chris Lattner](https://twitter.com/clattner
36	tter.com/clattner_llvm), [Joe Groff](https://twitter.com/jckarter
40	- Chris Lattner is [merging pull requests](ht
40	.com/swiftlang/swift/pull/166) at 10pm on Saturday.
44	ift 3.0. And thanks to [Erica Sadun](https://twitter.com/ericasad
46	- Chris Lattner [commits](https://github.com/
48	wift-compiler-crashes) from [@practicalswift](https://twitter.com/practica
48	lswift) has been [part of the repo](https://github.com/apple/swi
52	- [Jacob Bandes-Storch](https://twitter.com/jtbandes
56	- Much of the [swift-corelibs-foundation](https://github.com/apple/swi
58	- The [initial checkin](https://github.com/apple/swi
58	imported from an internal SVN repo. "Swift SVN r4". You will not
58	internal SVN repo. "Swift SVN r4". You will notice the followi
58	is source file is part of the Swift.org open source project. Copyrigh
63	ed in only three days. As Lattner says, *the revolution will be

_posts/2015-12-10-open-source-swift-weekly-1.md

Line	Typo
14	### This week on Swift.org
16	- [Manav Gabhawala](https://twitter.co
16	- [Manav Gabhawala](https://twitter.com/ManavGab
16	tion/pull/37) to add implicit initializers to Swift. In particular, this
18	- [Alex Denisov](https://twitter.com/1101_deb
22	- Chris Lattner tends to [fix](https://github
24	- [Brian Gesiak](https://twitter.com/modocach
24	tml) after noticing that the [xctest](https://github.com/apple/swi
34	- Andrew **Naylor** am
34	rgon) ambitiously implements [NSJSONSerialization](https://github.com/apple/swi
36	- [Jacob Bandes-Storch](https://twitter.com/jtbandes
36	tly improve the bridging of C APIs.
38	that discourages or prevents subclassing is [fine by me](https://twitt
40	he Swift Programming Language iBook (ePub) is available for *dire
40	t Programming Language iBook (ePub) is available for *direct dow
40	able for direct download on Swift.org (instead of only the iBook St
40	wift.org (instead of only the iBook Store) and is now under a [Cr
42	674254974629502976) for Chris Lattner.

_posts/2015-12-17-open-source-swift-weekly-2.md

Line	Typo
10	The Swift.org community is finishing up its
16	Craig Federighi reflects on Swift's first wee
16	eek out in the open with John Gruber on [The Talk Show](http://dar
18	It looks like @**zhuowei**
24	[Slava Pestov](https://github.com/sl
24	[Slava Pestov](https://github.com/slavapest
26	Dominique **d'Argent** i
26	nubbel) introduced the first [unicode variable name](https://github
26	l happily buy a ☕ or 🍺 for anyone who can successfu
26	ge a pull request that uses 💩.
28	[Bill Abt](https://github.com/apple/swi
28	ft and the core libraries! As Federighi mentioned on The Talk Show, I
30	Chris Lattner [fixed](https://github.com/app
32	Daniel **Duan** su
34	[@practicalswift](https://twitter.com/practica
36	com/swiftlang/swift/pull/439) for ARMv7 hosts such as the Raspberry P
36	sts such as the Raspberry Pi, BeagleBone, and Nvidia Tegras.
36	Raspberry Pi, BeagleBone, and Nvidia Tegras.
36	ry Pi, BeagleBone, and Nvidia Tegras.
38	Brian Gesiak [continued to pursue](https:/
38	s-xctest/pull/14) testing the XCTest framework, and in terms of nu
38	ctest/graphs/contributors) on corelibs-xctest.
42	) and say thank you to [Erica Sadun](https://twitter.com/ericasad
44	s://github.com/ddunbar), and [Mattt Thompson](https://github.com/
46	Max Moiseev's [proposal](https://github.com
50	Erica Sadun also has a [great post](http:
54	c method dispatch. From Chris Lattner: *"TL;DR: What I’m really get
56	Fabian Ehrentraud started a [discussion](https:
56	ting Objective-C code without nullability attributes. Currently, un-ann
56	bility attributes. Currently, un-annotated Objective-C members are bridg
56	Swift as implicitly unwrapped optionals (e.g., view!). This proposa
56	plicitly unwrapped optionals (e.g., view!). This proposal sugg
56	ts importing these members as optionals (view?) instead, which woul
56	sure why implicitly unwrapped optionals exist to begin with, as they
58	Colin Cornaby [suggested](https://lists.swi

_posts/2015-12-24-open-source-swift-weekly-3.md

Line	Typo
10	221/000540.html) for a bit on Swift.org. I have been traveling for the
16	@**tienex** s
16	ple/swift/pull/608) for Linux/armv7 support.
18	[@practicalswift](https://github.com/practical
20	[@masters3d](https://github.com/masters3d
22	Doug Gregor [implemented SE-0001](https:/
22	Doug Gregor [implemented SE-0001](https://github.com/apple/swi
26	Oisin Kidney's [proposal (SE-0008)](https://github.com/apple/sw
26	or-optionals.md), Add a Lazy flatMap for Sequences of Optionals,
26	Lazy flatMap for Sequences of Optionals,* has been [accepted](https:/
28	Kevin Ballard's [proposal (SE-0015)](https://github.com/apple/sw
30	Joe Groff submitted [a proposal](https:
30	wift has some special-purpose hardcoded behaviors, for example,lazy
30	mple behaviors include: lazy, memoization, delayed initialization, rese
34	Andyy Hope started [a discussion](h
34	allValues` static variable to enums, which would return an array
34	an array of all cases in the enum. Looks like there is a lot of
34	t for the idea so far. [Jacob Bandes-Storch](https://lists.swift.org/pipe
36	e I should go ahead and setup swiftweekly.org?

_posts/2016-01-07-open-source-swift-weekly-4.md

Line	Typo
10	e started to pick up again on Swift.org. If you are following any of t
10	you are following any of the repos on GitHub, you have probably
10	this week I just discovered [SwiftExperimental.swift](https://github.com/apple/swi
10	it defines a bunch of custom unicode operators for Set. It's rea
10	ool. I would love to see more APIs like this in the standard lib
16	[Austin Zheng](https://github.com/austinzhe
18	Andrew **Naylor** me
18	) to speed up JSON parsing in corelibs-foundation. We all know how much the Swi
20	the Swift package manager to corelibs-xctest.
22	Chris Lattner completely [redesigned](https
26	initialization.md) to improve memberwise initializers is now [under re
26	ion.md) to improve memberwise initializers is now [under review](https:/
26	2016-January/000010.html). As Lattner [pointed out](https://lists.s
26	deficiencies with the current memeberwise initializer behavior in Swift
26	with the current memeberwise initializer behavior in Swift. I have a g
30	Doug Gregor has submitted a [proposal](ht
34	Doug Gregor [notes](https://lists.swift.o
34	ot exposed to the Objective-C runtime. Luckily, I haven't run into

_posts/2016-01-14-new-weekly-brief.md

Line	Typo
10	o be originally published at [swiftweekly.github.io](http://swiftweekly.github.io
16	p](http://getbootstrap.com), **jQuery,** and [Fon
20	e thing I always liked about **NSHipster** and [o
20	r](http://nshipster.com) and **objc.io**

_posts/2016-02-06-jsqmessages-call-for-contributors.md

Line	Typo
10	ontributing to a widely used, impactful project? Then I have a propos
10	ontributors to help maintain [JSQMessagesViewController](https://github.com/jessesqui
18	ve as fast as I would like. JSQMessages is not being abandoned. But
22	some way*. You might even use JSQMessages. Working on this library will
29	* There's a nice [roadmap](https://github.com/jessesqui

_posts/2016-02-19-swifty-presenters.md

Line	Typo
16	-language/events/227833264/)) meetup. A [video of the talk](https:
22	azing job with these posts! 🙌
24	res ideas for making it more swifty. It isn't too technical, but
24	sn't too technical, but these APIs are really interesting to me
24	evelopers could utilize these APIs, they instead opt for view co
26	) in a new framework — [PresenterKit](https://github.com/jessesqui
26	when it is I will push it to **CocoaPods.** Mean
28	/twitter.com/jesse_squires) 😀
32	> One major shortcoming of UIKit is that view controllers have
32	common operations with a more Swifty API that reduces boilerplate

_posts/2016-03-21-contributing-to-swift.md

Line	Typo
22	kers presented in Japanese! 😎 The speakers met with the tr
22	ge barrier. It was awesome. 🎉
24	/twitter.com/jesse_squires) 😄
26	) of the talk is excellent. 🙌

_posts/2016-05-20-swift-documentation.md

Line	Typo
21	{% include updated_notice.html
22	update_message='After reading this, be sure
33	these common attributes. See [Alamofire](https://github.com/Alamofire
33	ub.com/Alamofire/Alamofire), **CocoaPods,** and
33	(https://cocoapods.org), and [JSQMessagesViewController](https://github.com/jessesqui
39	parameter values? Search for APIs that are similar to the ones
41	Here's an example from [JSQMessagesViewController](http://cocoadocs.org/docsets
55	follows the precedent set by UIKit, not only in terms of the API
57	ally. It might even be fun! 😉
61	thub.com/onevcat)'s plug-in, [VVDocumenter-Xcode](https://github.com/onevcat/V
61	cat/VVDocumenter-Xcode). This Xcode plug-in will generate the doc
67	urce tools, maintained by JP **Simard** an
67	github.com/jpsim) and [Samuel Giddins](https://github.com/segiddins
69	, but here's an example from [PresenterKit](https://github.com/jessesqui
85	istributing your library with CocoaPods, then you'll get [CocoaDocs](
85	h CocoaPods, then you'll get **CocoaDocs** for fr
85	oadocs.org) for free. Thanks **Orta!** N
85	/twitter.com/orta)! Note that CocoaDocs uses jazzy for Swift pods.
87	) or projects that do not use CocoaPods you'll need another solution.
87	t named gh-pages, push some html or markdown using [Jekyll](ht
91	ted">Note: Even though CocoaDocs provides docs for free, I sti
91	his. One small shortcoming of CocoaDocs is the inability to update do
91	se. Suppose you release v1.0 of your library and th
91	your documentation. To update CocoaDocs, you'll need to push another
91	sh another version, say v1.0.1, to CocoaPods. In any
91	, say v1.0.1, to CocoaPods. In any case, it's nice to ha
91	rs your library via GitHub or CocoaPods, the docs are right there.</p
95	l of my libraries. We'll use [JSQCoreDataKit](https://github.com/jessesqui
99	he library, an example app, a readme, and other supporting files l
100	- The [gh-pages](https://github.com/jessesqui
104	On my machine I checkout the repo twice. One checkout for bot
104	t for both the develop and gh-pages branches.
107	DataKit` is the checkout for gh-pages
109	se a lot of churn, since the gh-pages branch doesn't contain the l
111	#### Submodules
113	t all comes together. On the gh-pages branch, I add the develop
113	anch of the library as a [git submodule](https://git-scm.com/book/en/
113	git submodule update, the gh-pages branch can pull in the libra
113	om the root directory of the gh-pages checkout and provide the sub
113	ges* checkout and provide the submodule directory as the `--source-di
117	On the [gh-pages](https://github.com/jessesqui
117	Kit/tree/gh-pages) branch of [JSQCoreDataKit](https://github.com/jessesqui
119	The first is [build_docs.sh](https://github.com/jessesqui
119	s.zsh), which will update the submodule and build the docs. These can
133	The second is [publish_docs.sh](https://github.com/jessesqui
145	I switch directories to the gh-pages checkout and run `./publish_
159	ithub.io/JSQCoreDataKit/) for JSQCoreDataKit.

_posts/2016-05-22-open-source-everything.md

Line	Typo
22	the fourth major version of [JSQCoreDataKit](https://github.com/jessesqui
22	this release. That's right. [Sébastien Duperron](http://code-craftsm
22	ase. That's right. [Sébastien Duperron](http://code-craftsman.fr/abo
22	ode-craftsman.fr/about/) ([@Liquidsoul](https://github.com/Liquids
40	complete picture for how your APIs should behave. Continuous int
40	lures or test coverage. Once [travis-ci](https://travis-ci.org/jesses
40	esquires/JSQCoreDataKit) and [codecov.io](https://codecov.io/gh/jesses
42	{% include image.html
43	file="tests_pass.jpg"
46	source_link=null
47	half_width=false
52	ing** — bugs, features, todo items, refactors, future work
54	o makes it easier to write a [changelog](https://github.com/jessesqui
56	in their changes because of [travis-ci](https://travis-ci.org/jesses
56	esquires/JSQCoreDataKit) and [codecov.io](https://codecov.io/gh/jesses
60	So that's what happened. Sébastien was able to synthesize and na
60	ue for the 4.0.0 release of JSQCoreDataKit. We had some discussion and I

_posts/2016-06-04-avoiding-objc-in-swift.md

Line	Typo
10	yntax introduced by proposal [SE-0022](https://github.com/apple/swi
16	nfigurations, especially with UIKit.
92	- We're using structs (the ViewModel) in the prot
137	se this is the best and "most Swifty" answer to the problem. Howev
137	always immediately clear when Xcode suddenly starts yelling at yo
137	else* — especially when Xcode's fix-its are usually what you

_posts/2016-07-25-migrating-to-swift-3.md

Line	Typo
16	l the changes and errors from Xcode's migration tool at once. The b
27	branch from develop and run Xcode's migrator for Swift 2.3
28	e swift2.3 branch** and run Xcode's migrator for Swift 3.0
40	op` when the final release of Xcode 8 is out.
51	brary. This ensures semantic v**ersioning** and allow
55	directly to Swift 3.0. I use C**ocoaPods,** so I
57	escribed above. For example, [AlamoFire](https://github.com/Alamofire
59	Until Xcode 8 is final, you'll need to po
65	This tells CocoaPods to fetch the latest on the `s
67	rate your main app. I suggest commiting all of this migration in a si
73	** and having a list of green checkmarks once the migration is over is
74	grate, commit, then make Swifty API changes in a follow-up c
75	as, so you have to apply each Xcode fix-it individually.
81	Xcode's migration tool is not perfect
86	in contexts often resulted in derpy things like `(indexPath as! N
87	- Enums with associated NSDate valu

_posts/2016-07-31-enums-as-configs.md

Line	Typo
20	he same basic cell layout and subviews, but different visual treatme
128	ation to a class's designated initializer.
136	provide default values in the initializer:
198	cl/NSURLSessionConfiguration) APIs are designed. Among others, t
206	ifty-presenters/) about these APIs.
208	t's look at the evolution of [JSQMessagesViewController](https://jessesquires.github.
208	8) an enum to determine how timestamps were displayed in the message
208	y do not even have to contain timestamps! The API just asks for any ar
208	ide more powerful and simpler APIs for clients for custom config

_posts/2016-09-30-speaking-at-frenchkit.md

Line	Typo
16	-proposals-in-30-minutes) at **FrenchKit** in Pari
16	considering it was the first FrenchKit ever. I think the organizers
16	rs are already thinking about FrenchKit 2017, so keep an eye out and
28	and mentally paste it here. 😄 Always, these indie conferen
28	ple — and the people at FrenchKit were awesome. We do not often
30	fr/#programme), I got to meet Bas Broek ([@BasThomas](https
30	programme), I got to meet Bas Broek ([@BasThomas](https://twi
30	I got to meet Bas Broek ([@BasThomas](https://twitter.com/BasTho
30	weekly.github.io), as well as Sébastien Duperron ([@liquidseb](ht
30	hub.io), as well as Sébastien Duperron ([@liquidseb](https://twi
30	ll as Sébastien Duperron ([@liquidseb](https://twitter.com/liquid
30	contributed significantly to [JSQCoreDataKit](https://github.com/jessesqui
30	ks (including a few users of [JSQMessages](https://github.com/jessesqui
32	organizers opened with a mock Apple-esque keynote. It was hilarious and
32	The conference closed with 🧀 and 🍷 (of course) an
32	e closed with 🧀 and 🍷 (of course) and a fun pub qu
34	{% include image.html
35	file="frenchkit-michel.jpg"
35	file="frenchkit-michel.jpg"
36	alt="FrenchKit pub quiz"
38	source_link=null
39	half_width=false
44	tps://twitter.com/nlauquin), [Mathieu](https://twitter.com/Monsieur
44	nsieurDart), and everyone at **Xebia**

_posts/2016-10-01-shipping-swift-3.md

Line	Typo
16	ans if you were using version 3.x of a library, then the 4.0 re
16	ase is for Swift 3. Semantic **versioning** is key he
20	- [JSQDataSourcesKit](https://github.com/jessesqui
21	- [JSQCoreDataKit](https://github.com/jessesqui
22	- [PresenterKit](https://github.com/jessesqui
23	- [JSQWebViewController](https://github.com/jessesqui
24	- [DefaultStringConvertible](https://github.com/jessesqui
32	5. Rebase the swift3.0 branch onto `d
53	- [GrandSugarDispatch](https://github.com/jessesqui
54	- [JSQNotificationObserverKit](https://github.com/jessesqui
55	- [JSQActivityKit](https://github.com/jessesqui

_posts/2016-10-03-understanding-swift-evolution.md

Line	Typo
10	I recently [spoke at the FrenchKit conference]({{ site.url }}{%
24	ub.com/apple/swift-evolution) repo
29	wn. I've filed a few radars: [rdar://28589341](https://openradar
29	/radar?id=6066152501411840), [rdar://28589062](https://openradar
29	/radar?id=6673199689367552), [rdar://28589211](https://openradar
31	contain most of the proposal metadata, as well as the raw file cont
50	ts (1 or more), a Status, a filename, the file contents, and the t
111	rough metric for complexity, i**.e.** the longer a proposal is, the
111	complex change. For example, [SE-0107: UnsafeRawPointer API](http
111	ange. For example, [SE-0107: UnsafeRawPointer API](https://github.com/appl
111	at around 7,300 words, while [SE-0114: *Updating Buffer "Value" Nam
111	e lower-level UnsafePointer APIs. How does the complexity of p
127	- Erica Sad**un** has a
131	1. Erica Sadun, 25
132	2. Doug Gregor, 15
133	3. Joe Groff, 12
134	4. Dave Abrahams, 10
135	5. Chris Lattner, 10
136	6. Dmitri Gribenko, 7
137	7. Jacob Bandes-Storch, 6
139	9. Austin Zheng, 5
146	s still young — and our dataset of proposals is quite small.

_posts/2016-11-05-140-proposals-frenchkit-video.md

Line	Typo
16	As you may know, I [spoke at FrenchKit]({{ site.url }}{% post_url 20
26	als — exploring statistics on metadata and more, and examining why a

_posts/2017-01-08-swift-documentation-part-2.md

Line	Typo
16	e docs script that I use for [PresenterKit](https://github.com/jessesqui
34	nal, custom markdown docs. In PresenterKit, this is used to generate the

_posts/2017-01-16-testing-without-ocmock.md

Line	Typo
10	**OCMock** is a powe
10	ss to the dynamic Objective-C runtime? Or, what if you don't want y
18	ible, especially in a [legacy codebase](https://www.amazon.com/dp/01
29	ng are for networking events, analytics logging, and user interaction
54	ock ControllerProtocol with OCMock in an Objective-C test:

_posts/2017-02-10-refactoring-singletons-in-swift.md

Line	Typo
10	test, and they entangle your codebase when used implicitly in other
100	d liberate your debt-stricken codebase. While dependency injection i
102	there is a real problem. The initializer for MyViewController used t
102	er to view controller. Legacy codebases in particular will struggle t
104	The initializer for most classes (and especia
106	e in a complex, large, legacy codebase — at least not at once,
112	g too much complexity in your codebase. Maybe your app really will o
128	Now, the initializer has not changed from the pers
136	can continue working in your codebase as if nothing ever changed.
152	adding a new property and an initializer parameter with a default valu
154	xamples here — classes, structs, enums, functions. Every func
154	ere — classes, structs, enums, functions. Every function in

_posts/2017-06-05-protocol-composition-in-swift-and-objc.md

Line	Typo
16	l the time while working with UIKit, for example. But despite the
16	ional method, you'll crash at runtime with a 'does not respond to s
44	but you incur the additional runtime cost of checking `-respondsTo
60	ou do not have to worry about runtime crashes in Swift, but there i
60	ure relies on the Objective-C runtime and **it only exists for inte
60	ch means you cannot use Swift structs, enums, or other nice feature
60	you cannot use Swift structs, enums, or other nice features. This
68	exactly that. Thus, we could reimagine these APIs in the following w
68	hus, we could reimagine these APIs in the following way:
111	and delegate properties is unintuitive and cumbersome. How can we im
185	UPDATE: [@iankay](https://twitter.com/IanKay/s

_posts/2017-07-16-officially-deprecating-jsqmessagesviewcontroller.md

Line	Typo
11	Beginning immediately, [JSQMessagesViewController](https://github.com/jessesqui
15	{% include updated_notice.html
16	update_message='The community has started a
25	or so libraries available via CocoaPods. [CocoaPods](https://cocoapod
25	ies available via CocoaPods. **CocoaPods** now h
25	otes, or even proper semantic versioning. I had no idea what I was doi
41	the library as deprecated in CocoaPods today.
43	here are similar libraries in CocoaPods. I cannot vouch for any of th
51	g 11,000 stars. According to [CocoaPods stats](https://cocoapods.org/

_posts/2017-07-17-a-story-about-swift-source-compat.md

Line	Typo
10	erent story when your project seg faults the compiler. When tha
10	is released. This is why the [swift-source-compat-suite](https://github.com/apple/swi
14	### About Swift source-compat
18	The scripts read the project metadata from the manifest, do a `git
22	I maintain a unique project. [JSQDataSourcesKit](https://github.com/jessesqui
22	urcesKit) originally broke in Xcode 8.3 beta 3 (8W132p). [It cr
22	error. This was before the swift-source-compat-suite existed. I'm no compiler expe
22	-4088). Within a week or so, [Slava came to the rescue with a fix
22	887). In the final release of Xcode 8.3, my project compiled.
24	comment-255288) reappeared in Xcode 9 beta 1 (9M136h) — a
24	mdash; adding this project to swift-source-compat-suite had been on my Todo list fo
24	-compat-suite had been on my Todo list for weeks. It would hav
24	ity regression* in a previous Xcode beta. [Fool me once](http:/
26	were hitting this bug in the source-compat suite. This seemed to get eve
26	. By midnight that same day, [Slava had a fix ready](https://gith
28	In Xcode 9 beta 3 (9M174d), not only
69	First, you define all of the metadata about your project and find a
75	It is clear to me that swift-source-compat-suite is extremely valuable. Withou
75	munity. Also, huge thanks to [Slava](https://twitter.com/slava_pe

_posts/2017-08-08-customizing-git-log.md

Line	Typo
20	As much as Greg **Heo** would [
20	miss* about hg — the [smartlog](https://www.mercurial-scm.or
24	ll commit message, along with metadata like date and author. If you
24	Run git log in one of your repos.
38	you can set an alias called "smartlog". You can also find a [gist h
44	Here's a comparison (using [IGListKit](https://github.com/Instagram
48	H). If available, ref names (e**.g.** branches, tags, etc.) (%d`)
48	followed by a relative date (e**.g.** "2 days ago") (%cr). Next i

_posts/2017-09-03-supporting-json-feed.md

Line	Typo
10	upport for Brent Simmons' and Manton Reece's [JSON Feed](https://j
10	for Brent Simmons' and Manton Reece's [JSON Feed](https://jsonfeed.
14	{% include blockquote.html
24	s that — like Brent and Manton so clearly articulate —
28	.json). You fill-in your site metadata, and iterate through each pos
28	ice that it's similar to [the feed.xml](https://github.com/jessesqui

_posts/2017-09-10-building-a-site-with-jekyll-on-nfsn.md

Line	Typo
11	inexpensive bare-bones host, [NearlyFreeSpeech.net](https://www.nearlyfreespeech
15	{% include updated_notice.html
16	update_message="This post has been updated t
16	o reflect recent changes with NearlyFreeSpeech, as well as changes in my set
21	* This site hosted on [NearlyFreeSpeech.net](https://www.nearlyfreespeech
22	ttps://jekyllrb.com) via the [github-pages gem](https://github.com/githu
23	HTML, CSS, and JS. (And thus **jQuery.)** I've
23	themes.getbootstrap.com), or **JekyllThemes,** or
23	http://jekyllthemes.org), or [Octopress](http://octopress.org/docs/th
25	umps the output in the public www directory.
30	te in seconds by deleting the repo on GitHub if you decide it's
32	ave to view the errors on the repo settings page, and if the err
32	rror is something generic and unactionable, you're stuck navigating thro
34	e with unsupported "plugins" (i.e., gems) or certain configurati
36	b.io/`, for example [jessesquires.github.io/JSQDataSourcesKit](https://je
36	ample [jessesquires.github.io/JSQDataSourcesKit](https://jessesquires.github.
36	t>. For example, if my pages **repo** is jessesquires.github.io` a
38	main. For non-custom domains (i.e., the default `.gith
38	g/enforcing HTTPS is a simple checkbox in your repo settings. Appare
38	is a simple checkbox in your repo settings. Apparently there ar
38	om-domain-website-https/) via Cloudflare, but that seems cumbersome an
40	registered my domain through [NearlyFreeSpeech](https://www.nearlyfreespeech
44	### About NearlyFreeSpeech.net and why
46	ly you've noticed the amazing double-entendre — "nearly free" as in [
46	s that it's not Amazon or GoDaddy. You are free from the shackl
48	{% include updated_notice.html
49	update_message="My costs are now around $
54	n NFSN, setting up a bare Git repo, and configuring everything s
56	o create a Gemfile and use **Bundler.** For any
66	would recommend. My [current Gemfile looks like this](https://gith
84	For any NearlyFreeSpeech site, [there are several subd
87	r static HTML files, images, .htaccess files, and most scripts in th
90	essible via the web, like Git repos, or privately used gems for a
99	e setup for Ruby and install **Bundler.** After ss
99	er](http://bundler.io). After ssh'ing into NFSN, you'll need to add
116	r remote when you clone the repo to your machine.
125	You can clone this repo locally.
143	you git push to the remote repo on your NFSN server. In the b
143	your NFSN server. In the bare repo you created (site.git), the
166	latest changes from the bare repo. Because this runs from the b
166	cause this runs from the bare repo hooks/ directory, we need t
166	is takes care of that.) Using Bundler allows you to easily keep gem
180	s. In your local clone of the repo, git commit your changes an
188	en it comes to your local git repo. With GitHub Pages, you'll al
188	Pages, you'll already have a repo setup on GitHub and cloned lo
188	.origin` to NFSN (to the bare repo you created there), and add a
196	en source, then don't add the github remote. Push to NFSN and your
202	ve the CNAME file from your repo. Push to NFSN (origin) and
203	e your <username>.github.io repo. I renamed mine to `jessesqui
203	remote.github to your local repo, don't forget to reset the ur
203	po, don't forget to reset the url to the new repo name.
203	t to reset the url to the new repo name.
213	FSN, I decided to make a new [jessesquires.github.io](https://github.com/jessesqui
213	.github.io) GitHub Pages site repo. This is simply a landing pag
217	ee*, it is nearly free. &#x1F609;

_posts/2017-09-18-measuring-compile-times-xcode9.md

Line	Typo
16	0ccf0e1fb6729eea75bc09cbd5) a frontend Swift compiler flag that woul
16	g that would emit warnings in Xcode for functions that took too l
16	ok too long to type-check. In Xcode 9, there's a new, similar fla
22	[Bryan Irace](http://irace.me/swift-profil
22	race.me/swift-profiling) and [Soroush Khanlou](http://khanlou.com/2
22	swift-profiling) and [Soroush Khanlou](http://khanlou.com/2016/12/g
22	0` would trigger a warning in Xcode for any function that took lo
22	unction that took longer than 100ms to type-check. This was alway
22	e1fb6729eea75bc09cbd5): *As a frontend option, this is UNSUPPORTED a
22	ver, this flag still works in Xcode 9 and I haven't seen any disc
26	In Xcode 9, there is a new, similar fl
26	appearance in the official Xcode 9 GM [release notes](https://
39	shold is too low, for example 10ms, then you will get a ton
39	pressions to compile in under 200ms. Also, I recommend setting th
41	re two common scenarios where Xcode will start emitting warnings
47	usly a temporary (and kind of hacky) solution to improving compil
51	e [this doc in the main Swift repo](https://github.com/apple/swi

_posts/2017-09-24-when-your-app-is-used-in-unexpected-ways.md

Line	Typo
16	**PlanGrid** is
16	n simply did not exist before PlanGrid, and they still have a lot of
20	At PlanGrid, we recently discovered that
26	{% include blockquote.html
27	quote="Adapting PlanGrid to Archaeology
29	n of the construction program PlanGrid as a digital field registrati
29	egistration system at the Tel Burna Archaeological Project (Israe
29	ng: (1) the benefits of using PlanGrid with tablets and smartphones;
29	ur specific adaptation at Tel Burna; (3) this past season’s innov
29	other projects may implement PlanGrid as a digital archaeological t
35	the last few seasons, the Tel Burna Archaeological Project has ta
35	vor through our adaptation of PlanGrid as a replacement for our hand
35	ritten paper forms and plans. PlanGrid is a widely-used construction
37	the main features that allow PlanGrid to streamline workflow within
37	rchitectural plan management; searchable and adaptable annotations; ar
37	easurements; well-defined and useable administrator tools; simple r
45	{% include image.html
46	file="plangrid-archaeology-2.jpg"
46	file="plangrid-archaeology-2.jpg"
47	alt="PlanGrid for Archaeology"
48	caption="PlanGrid used for Archaeology"
49	source_link="http://popular-archaeology.c
50	half_width=false

_posts/2017-10-01-floating-point-swift-ulp-and-epsilon.md

Line	Typo
17	{% include updated_notice.html
18	update_message='I gave a talk on the topics
21	{% include image.html
22	external_url="https://media.giphy.com/medi
25	source_link=null
26	half_width=false
35	- [SE-0104](https://github.com/apple/swi
36	- [SE-0113](https://github.com/apple/swi
36	ntegral rounding functions to FloatingPoint* (Swift 3)
37	- [SE-0067](https://github.com/apple/swi
47	ificant digit in this value’s significand. For most numbers x, this is
53	As [Olivier Halligon](http://alisoftware.github.io
55	> NOTE: Rationale for "ulp" instead of "epsilon":
58	or of two or so. By contrast "ulp" is a term with a specific un
64	*By contrast “ulp” is a term with a specific un
64	ace) for Unit in the Last Place, or Unit of **L
64	or Unit of Least Precision. It's a unit, but what does i
68	We do not need any notion of "ulp" for integers. They are simpl
70	e set of all real numbers is [uncountably infinite](https://en.wikipedi
72	resentable value* — or, ulp**.** Thus, aside from the obvious
72	of silicon chips, this is why ulp needs to exist.
74	{% include image.html
75	file="numbers.jpg"
78	source_link=null
79	half_width=false
84	Before we define ulp**,** let's briefly review the mem
86	g/wiki/Exponent_bias)), and a significand. For single precision, there'
86	exponent, and 23 bits for the significand — 32 bits total. Double
86	exponent and 52 bits for the significand, which totals 64 bits. I won'
88	{% include image.html
89	file="float32memory.png"
90	alt="Floating-point format, binary32"
92	source_link="http://blog.reverberate.org/
93	half_width=false
98	Let's look at an example in base-10. We can represent 123.45 with
98	e can represent 123.45 with a significand of 12345 and exponent of -2:
116	Additionally, Swift's APIs allow us to explore the memor
116	verify them with this [handy IEEE-754 floating-point converter](htt
136	0.pdf) — 1.19e-07 for 32-bit values (Float) and `2.22e-1
136	(Float) and 2.22e-16 for 64-bit values (Double). As expecte
194	oilerplate') files, read [Ole Begemann's post](https://oleb.net/blog/2
221	dash; the sign, exponent, and significand. This code eventually calls i
221	ntually calls into the public initializer: `init(sign: exponentBitPatte
223	ew the implementation of this initializer. We see a lot of [bit manipul
223	ki/Bit_manipulation), namely [bitwise AND](https://developer.apple.
242	This initializer eventually calls into `init(b
242	y constructs a primitive LLVM 32-bit float (FPIEEE32) from the raw
250	Swift playground. For private APIs like `Float._infinityExponent
303	So that's ulp**.** There's still a lot to unpac
303	helpful to better understand ulp and a little about floating-p
310	programmer-should.html), Josh Haberman
312	g_point_visually_explained/), Fabien Sanglard
312	_visually_explained/), Fabien Sanglard
313	status/IEEE754.PDF), Prof. W. Kahan, UC Berkeley
314	- [IEEE**-754** Floating Point Converter](htt

_posts/2017-12-28-swift-weekly-brief-hiatus.md

Line	Typo
21	{% include updated_notice.html
22	update_message='The <a href="https://swiftwe
22	BasThomas" class="alert-link">Bas Broek and other contribut
22	homas" class="alert-link">Bas Broek and other contributors. T
22	ribe via RSS/Atom, JSON feed, AppleNews, or Twitter.'
33	d in the newsletter, the new [Swift.org forums](https://forums.swift.
41	I am more than happy to help onboard any newcomers — whateve
49	ted to take the model of [iOS Dev Weekly](http://iosd**ev**weekly.c
53	del and foundation for [Colin Eberhardt](https://twitter.com/ColinEbe
53	to start his own newsletter, [WebAssembly Weekly](http://wasmweekly.new
61	- Brian Gesiak, [@modocache](https://twitter
61	- Brian Gesiak, [@modocache](https://twitter.com/modocach
62	- JP Simard, [@simjp](https://twitter.com
62	- JP Simard, @si**mjp**
63	- Bas Broek, [@basThomas](https://t
63	- Bas Broek, [@BasThomas](https://twitter
63	- Bas Broek, [@BasThomas](https://twitter.com/BasThoma
64	- Greg Heo, [@gregheo](https://twitter.c
64	- Greg Heo, @gr**egheo**
65	- Ben Asher, [@benasher44](https://twitter.com/benasher
66	- Garric Nahapetian, [@garricn](https:
66	- Garric Nahapetian, [@garricn](https://twitter.c
66	- Garric Nahapetian, @ga**rricn**
67	- Roman Volkov, [@volkovre](https://twitter.
67	- Roman Volkov, [@volkovre](https://twitter.com/volkovre

_posts/2018-02-25-replacing-google-with-duckduckgo.md

Line	Typo
16	in general. I've been using **DuckDuckGo** for
26	oject.org) results, which are diplayed prominently in search, [threa
32	### Benefits of DuckDuckGo
34	Apple products and services, DuckDuckGo was established with [privacy
34	not mine your data to do so. DuckDuckGo also sometimes adds an affili
34	dds an affiliate code to some eCommerce sites, but this is done witho
34	business or legal purposes". DuckDuckGo provides a strikingly differe
36	{% include blockquote.html
38	source="Gabriel Weinberg, [DuckDuckGo Privacy](https://duckduckgo.c
41	-of-Google) of reasons to use DuckDuckGo. I won't bother reiterating e
41	an type !giphy oh yeah into DuckDuckGo and land directly on [Giphy's
41	kDuckGo and land directly on [Giphy's search results](https://giphy
45	After over a year of using DuckDuckGo, I think search quality is ex
45	l, I'm incredibly happy using DuckDuckGo. If you aren't satisfied with
45	If you aren't satisfied with DuckDuckGo's results, you can always `!goo
49	used here before switching to DuckDuckGo. You enter your site url, con
49	ckDuckGo. You enter your site url, configure some settings, and
51	Creating a [custom DuckDuckGo search](https://duckduckgo.co
51	our own search box with basic html forms — something Googl
73	search and replacing it with DuckDuckGo. It's much more elegant, ligh
75	class="text-muted">Thanks to [Stuard Colville](https://muffinresea
75	text-muted">Thanks to [Stuard Colville](https://muffinresearch.co.uk
75	t on doing this without using iframes.
79	time to switch. By supporting DuckDuckGo, you will only help improve t

_posts/2018-04-16-upgrading-to-bootstrap-4.md

Line	Typo
10	18/04/09/bootstrap-4-1/) from v3.3.7. Its a major version with lots
18	ation/#grid-system) to using [flexbox](https://developer.mozilla.or
34	tain. It's easy to get wrong. Flexbox simplifies this, allowing you
50	the links to GitHub, Twitter, LinkedIn, etc). Those are placed in a
68	highly recommend upgrading to v4. If you don't use Bootstrap,
76	de into new components using [jekyll includes](https://jekyllrb.co

_posts/2018-06-10-why-optional-swift-closures-are-escaping.md

Line	Typo
9	escaping by default. After [SE-103](https://github.com/apple/swi
13	make a closure escaping. Greg Heo provides [a great explanation
13	eat explanation over on Swift Unboxed](https://swiftunboxed.com/lan
15	scaping in Objective-C. [Doug Gregor explained the problem](https:
23	res were treated differently. Optionals are [just a 2-case enum](http
23	Optionals are [just a 2-case enum](https://github.com/apple/swi
23	e isn't much different from a struct that has a closure property:

_posts/2018-07-05-swifts-new-calling-convention.md

Line	Typo
14	on specifies how subroutines (e.g., functions) should receive th
14	ct between the caller and the callee, so that each knows which ins
16	, and when? The caller or the callee? This is what changed in Sw
26	tion prior to Swift 4.2, "callee-owned".
82	It is no longer the callee's responsibility to release the
82	e size reduction as well as a runtime performance improvement, beca
84	### Calling non-inlinable functions in other modules
86	change is the case of calling non-inlinable functions across module bound
86	s module boundaries. [Michael Gottesman](https://twitter.com/gottesma
172	ht. In the old convention ("callee-owned"), it was not possible to per
172	bility of the caller, not the callee, the optimizer can eliminate

_posts/2018-07-29-swift-tip-building-arrays-with-compactmap.md

Line	Typo
13	imilar scenario recently. In **PlanGrid,** we use
13	om/plangrid/reactivelists) to declaratively build our user interfaces for
59	esulting array is an array of optionals. We don't want that, because
59	he table. It's better to keep optionals as far to the edge of our obj
84	mprovement. We've removed the optionals. It works, but it has a few p
84	has a few problems. It's very stateful and imperative. It's not very
84	compactMap() to remove the optionals.
99	of these functions to return optionals or non-optionals and this cod
99	ctions to return optionals or non-optionals and this code never has to ch

_posts/2020-01-06-selecting-an-xcode-version-on-github-ci.md

Line	Typo
9	into is selecting a specific Xcode version, which is a bit cumbe
13	not find anything mentioning Xcode versions. I discovered the do
13	ll the installed software and SDKs, which is convenient and well
13	early specifies the available Xcode versions that are installed,
17	are doing it. I noticed that [Alamofire](https://github.com/Alamofire
17	ci.yml#L18) in their workflow yaml file. Apparently, you can set
29	who is a [top contributor to Alamofire](https://github.com/Alamofire
31	sts. It would be great to see Xcode version selection improved, b

Generated by 🚫 Danger