DEV Community: Sean Walker

The future of user interfaces

Sean Walker — Wed, 18 Mar 2020 18:45:32 +0000

How do we really want to interact with computers? I don’t actually want to tap on a screen, I actually want to...

write a blog post
record music
write some code
watch a video
listen to a podcast

So how do I do these things today? I figure out an app to do the thing and then I remember the icon and the name and tap it. Then I navigate to the function I want in that app across a multitude of different user interfaces.

Is this what I want? Not really. I want to tell the computer what information I need or what I need to accomplish and the move on with my life. I want something like a command line but with bits of UI mixed in along with a myriad of APIs. I don’t think the diversity of applications should happen at the UI layer, I think it should happen at the data layer. Of course, who would willingly give up the ability to control what a user sees when they interact with their data? No one. So we live with proliferation of complex, confusing interfaces.

There may be a way to have this command line GUI but it’s going to have to be open source and cross platform, similar to http/web browsers themselves or google search but better and by cross platform I mean voice activated as well.

Here’s a very simple concrete example that google already does well:

Just typing “di” probably brings up dictionary or “def” -> definition or you can type in any word and get the definition right there. Simple examples like that don’t really cost money and no one cares about multiple choices of dictionary. The vast majority of google searches are limited to information retrieval, what I'm talking about is something like command execution.

Take sending an email newsletter for example:

Given our magic command line GUI, you type "send email" and a few things may happen next

You may see a form with common email fields (to, from, subject, body) and possibly be able to upload or type in a list of people you’d like to send to
You’d get a list of choices of email newsletter providers with prices per # of emails to send
Finally you can opt to save your preference for next time

This future isn't impossible, and I can easily imagine the a "SaaS google" looking something like this and pushing the innovation to the data/API layer where quite a few data people will offer API integration at lower and lower price points as competition heats up.

The command line GUI is definitely something that seems way more plausible than it did even 5 years ago with the rise of native apps and js-heavy frontend applications in the browser backed by REST and graphql APIs.

Inspired by:

https://blog.repl.it/clui

Do Good Daily: The End

Sean Walker — Mon, 14 Oct 2019 21:58:59 +0000

📅 10/14/2019
🔥 0 day streak
📱 Do Good Daily
💰 $0.99 price
🤑 0 sales
⌚️ ?? hours spent
💻 58 changed files 1,405 additions ➕ and 58 deletions ➖.
🙅‍♀️ Today's goal: Put this project out to pasture

I did want to say something here about this sad, sad project that really, was doomed to fail. I actually did finish it, it's currently sitting in test flight, and it works just fine, looks just fine.

But that's really the problem, it's just fine.

It doesn't move the needle.

I poured quite a few hours into this thing and again, nothing.

It's not all bad and depressing though. I did learn SwiftUI 🤷‍♂️

But yeah, consider this the conclusion of this dreadfully short series 🗿

Do Good Daily Day 7

Sean Walker — Thu, 19 Sep 2019 03:56:56 +0000

📅 09/18/2019
🔥 7 day streak
📱 Do Good Daily
💰 $0.99 price
🤑 0 sales
⌚️ 6 hours spent
💻 7 files changed 42 additions(+) 2 deletions(-)
🙅‍♀️ Today's goal: Get the text editor working

SwiftUI TextField placeholder

Look, I'm not going to sit here and wax poetic about my 30 minutes of side project work today, I'll just go ahead and say, I didn't do what I set out to do. Having a full time job and a family is hard effin work.

Let's talk about textfields and placeholders, the only thing I did today:

TextField("My good for today", text: self.$input)
    .font(.custom(font, size: 18))
    .padding()

and that's it! Oh wait, I also figured out how to format today's date:

let fmt = DateFormatter()
fmt.dateFormat = "EEEE, MMMM dd"
self.date = fmt.string(from: Date())

and THAT's it!

Do Good Daily Day 6

Sean Walker — Wed, 18 Sep 2019 03:14:22 +0000

📅 09/17/2019
🔥 6 day streak
📱 Do Good Daily
💰 $0.99 price
🤑 0 sales
⌚️ 5.5 hours spent
💻 37 files changed 554 additions(+) 81 deletions(-)
✅ Today's goal: Finish up rough versions of intro views

SwiftUI is better than web development

When it works it really works. Today the preview feature was working well and I managed to breeze through a few more static screens with simple buttons. It's insane that all it takes to make something very custom looking is ~70 lines of code per screen, if you can even call it code. It definitely makes me wonder what the heck we're doing out there in web development land. Anyway, I did learn a few SwiftUI tricks that I thought I'd share

SwiftUI full screen background image

var body: some View {
        VStack {
        }
        .background(
            Image("bg")
            .resizable()
            .scaledToFill()
            .edgesIgnoringSafeArea(.all)
        )
}

That's all it takes. Either that or a ZStack and put the image as the first child.

SwiftUI center an image

Or anything really. It's funny I was messing around with this problem for a while but I came up with something pretty decent, assuming you only have one thing to center

var body: some View {
        VStack {
             Spacer()

             Image("bg")

             Spacer()
        }

}

There's some more code with some other goodies but there's always tomorrow.

Do Good Daily Day 5

Sean Walker — Tue, 17 Sep 2019 05:03:09 +0000

📅 09/16/2019
🔥 5 day streak
📱 Do Good Daily
💰 $0.99 price
🤑 0 sales
⌚️ 5 hours spent
💻 8 files changed, 53 additions(+), 9 deletions(-)
✅ Today's goal: Finish two intro views

SwiftUI is fast

Xcode swiftUI previews weren't working an hour ago, of course. Then I decided to do something crazy and restart my mac and POW 💥 they started working again. From this point it was like I reached the ballmer peak and the code flowed from my fingertips like lava flows down a raging volcano 🌋.

I'm tired and delirious, don't judge me. But feel free to go ahead and judge these latest views I coded up:

Do Good Daily Day 4

Sean Walker — Mon, 16 Sep 2019 04:00:03 +0000

📅 09/15/2019
🔥 4 day streak
📱 Do Good Daily
💰 $0.99 price
🤑 0 sales
⌚️ 3.5 hours spent
💻 8 files changed, 53 additions(+), 9 deletions(-)
✅ Today's goal: Write this blog post

My PIC made some epic designs today

The best thing about being on a side project team with a designer is that even when you're shoveling rocks from your driveway into a wheel barrow and then dumping them in your backyard to make your wife happy, he still gets stuff done on the app 👍

That's what happened today. I would share the designs with you because they're really good, but I'll wait until I have them implemented in swiftUI because this is dev.to not figma.to.

Alright I'll show one here because this post is kind of light

The design is unlike most apps on the app store, it's actually got me kind of excited to get this out there and see if it helps anyone besides the two of us.

This next week is do or die time, either a lot of progress will be made or this thing is not launching on Sept. 30th. Hopefully I'll be at it bright and early tomorrow morning and knock some of these great designs out of the park in an hour 😅

Stay tuned for tomorrow's post.

Do Good Daily Day 3

Sean Walker — Sun, 15 Sep 2019 04:15:52 +0000

📅 09/14/2019
🔥 3 day streak
📱 Do Good Daily
💰 $0.99 price
🤑 0 sales
⌚️ 3 hours spent
💻 8 files changed, 53 additions(+), 9 deletions(-)
✅ Today's goal: Recap yesterday

I have a life

It might surprise you but I actually have a life outside of making apps. I don’t want to bore you with the details but I didn’t get much done today.

Let’s dive right in. How do you set the initial state of a swiftUI view? Here’s how I did it

Setting SwiftUI initial state

import SwiftUI

struct MorningReminderView: View {
    @State var hour = 5
    @State var minute = 0

    var body: some View {
        VStack {
        }.onAppear {
            let app = Database().load()
            self.hour = app.morningHour
            self.minute = app.morningMinute
        }
    }
}

Custom fonts in swiftUI

There are three steps

Add the font to the project and make sure it’s added to the target
Add it to info.plist
Call it in your code

let font = "IM_FELL_English_Roman"
Text("Set Your Morning Reminder Time")
            .font(.custom(font, size: 16))

Serialize app state to a json file

Last thing about swiftUI today, serialize app state to a json file the ghetto way. Two parts, the first part is a class that handles loading and saving to the file along with json serialization, hey it does it all!

import Foundation

public struct App : Codable {
    public var morningHour:Int
    public var morningMinute:Int
    public var eveningHour:Int
    public var eveningMinute:Int
    public var entries:[Entry]
}

public struct Entry : Codable {
    public var goal:String
    public var accomplished:String
    public var timestamp:Int
}

public class Database {
    let paths = NSSearchPathForDirectoriesInDomains(.documentDirectory, .userDomainMask, true)[0] as String
    let db = "db.json"

    func getDocumentsDirectory() -> URL {
        let paths = FileManager.default.urls(for: .documentDirectory, in: .userDomainMask)
        return paths[0]
    }

    public func save(app: App) {
        let filename: URL = getDocumentsDirectory().appendingPathComponent(db)
        let data: Data = try! JSONEncoder().encode(app)

        try! data.write(to: filename)
    }

    public func load() -> App {
        let filename: URL = getDocumentsDirectory().appendingPathComponent(db)
        do {
            let data = try Data(contentsOf: filename, options: .mappedIfSafe)
            let app = try JSONDecoder().decode(App.self, from: data)

            return app
        } catch {
            return App(morningHour: 7, morningMinute: 30, eveningHour: 21, eveningMinute: 0, entries: [])
        }
    }
}

The second part is the SwiftUI view

import SwiftUI

struct MorningReminderView: View {
    @State var hour = 5
    @State var minute = 0

    var body: some View {
        VStack {
            Button(action: {
                self.saveTime()
            }) {
                Text("Save")
                .frame(width: 200)
                .font(.custom(self.font, size: 24))
                .padding()
                .foregroundColor(Color.black)
                .overlay(
                    RoundedRectangle(cornerRadius: 40)
                        .stroke(Color.black, lineWidth: 2)
                )
            }
        }
    }

    func saveTime() {
        var app = Database().load()
        app.morningHour = self.hour
        app.morningMinute = self.minute
        Database().save(app: app)
    }
}

And that's that for today, I learned quite a bit and I hope it helps you handle the basics of any swiftUI app. Let me know if you want to learn swift, I'm on twitter (probably too much) or just leave a comment and if I can answer any of your questions, I will!

Do Good Daily Day 2

Sean Walker — Sat, 14 Sep 2019 06:04:40 +0000

📅 09/13/2019
🔥 2 day streak
📱 Do Good Daily
💰 $0.99 price
🤑 0 sales
⌚️ 2 hours spent
💻 12 files changed, 693 insertions(+), 0 deletions(-)
✅ Today's goal: Add a swiftUI view that lets people set the first daily push notification time

Captain's Log: 🤦‍♂️

You know that feeling when you start a new side project and you're full of optimism and everything is going well?
That's a great feeling and you can hold on to that by never starting that side project.

Things went from great to good to bad for me really quickly. I don't know where start. Yes I do: xcode.

Ah xcode, the editor everybody loves to hate, but with good reason. It's horrible. The beta is especially bad this year with the introduction of swiftUI. Everything goes smoothly until you add any ObservableObject into your app. Then the preview starts failing and never recovers. Eventually I learned that the preview is only good for quick visual tweaks and leave it off now.

There's one more thing that really grinds my gears and that's a rogue swift process spinning up my CPU fan like it's an effin' apache helicopter. Closing xcode doesn't even put a stop to it, no this process needs Rey because I had to FORCE QUIT it.

Progress Still Happened

Learning any new thing is going to cause some grumpy cat moments, but swiftUI really takes the cake. Maybe I should focus on the positive. I did get a picker working and I serialized app state to a json file, I’ll get into the nitty gritty tech stuff tomorrow

Do Good Daily Day 1

Sean Walker — Fri, 13 Sep 2019 04:28:39 +0000

📅 09/12/2019
🔥 1 day streak
📱 Do Good Daily
💰 $0.99 price
🤑 $0 revenue
❌ 0 app store rejections
📈 0 sales
📉 0 upvotes
⌚️ 0 hours spent
😄 0 rewrites
🛬 0 landing pages
💻 12 files changed, 693 insertions(+), 0 deletions(-)
✅ Today's goal: Create the swift UI project

Intro

It's the middle of the month and fall is almost upon us. I'm not sure if the summer has driven me insane or if I'm just insane by default but I'm going to pick up my ridiculous plan from last year (and earlier this year) to make a new app every month, except this time I'm making actual ios and mac apps, not websites.

Which means the best bio ever© is getting a refresh:

~~Takes hikes 🏔 makes sites 👨‍💻~~

to this

Takes naps makes apps

No I'm kidding, although probably accurate given how senior I am now.

Alright this is it:

🚢 12 apps in 12 months… because the world needs more apps

September's App

Now that that's out of the way, let's get to the nitty gritty, what app are you dear reader going to look out for the apple app store on September 30th?

Do Good Daily, that's what.

Alright so this one is very simple mostly because I've got 15 days give or take to get it done and put in the app store, including today.

Features

Set a morning reminder time which will push notify you and ask "What good shall I do this day?"
Set an evening reminder time which will push notify you and ask "What good have I done today?"
Big text field to save your thoughts on if you accomplished your goal or "did your good" that day

Is this worth $0.99? I don't know. Let's find out.

Start your xcodes

That was easy. I'm going to pat myself on the back for today's progress.

How I made the world's worst clojurescript REPL

Sean Walker — Sun, 26 May 2019 17:26:49 +0000

Read the previous post here

TL;DR
I tried to run two clojure web servers and call a JSON endpoint and the JVM fell over on my $3.50/mo 512 MB RAM VPS server, so I switched to clojurescript and wrote my own prepl client for atom

Atom package development

If you've never done any atom package development, you're in for a wild, documentation reading ride! They made it really easy to get started. Of course at first I didn't even know where to start.

You start here btw.

What's funny is that post is 3 years old, so I thought, there's no way this will work with how many releases of atom there have been between then and now, but it does work.

I was initially confused because I kept searching "atom package development" but the blog post is title "… ATOM PLUGIN".

Classic half-baked intro coding tutorial (of which I have written many).

Alright so now I that I kind of understood what was going on and where the files go, it was time to head over to the documentation to learn how to get text out of an editor and put it into another editor and for that I consulted the all knowing, all seeing, all dancing atom documentation.

I initially used the getWordUnderCursor function which was great, but I needed to get the "outer most form" in a clojure form like this 👇

(defn say-hello [s] ; <—— outer most form is `defn`
  (println "hello" s))

...and what I was doing was not working.

So I did what every great artist does and I stole from the pre-eminent clojure repl in atom: proto-repl.

Clearly that code is much, much better than what I would write initially, so I decided to just stick that in my project and I was off to the races!

console.log('cljs-repl:send');

let editor;
if (editor = atom.workspace.getActiveTextEditor()) {
  let range = EditorUtils.getCursorInClojureTopBlockRange(editor);
  let s = editor.getTextInBufferRange(range);
  console.log(s);
}

That's the whole bit of code now using proto-repl's code that gets the outer most form in a clojurescript. After this I was a hop, skip and a jump away from a fully running clojurescript prepl client, since the prepl part does all the heavy lifting for you.

At this point all I had to do now was

Connect to the running clojurescript prepl socket
Send the code over
Parse the EDN that comes back
..and show it in a new pane!

Wait, parse the EDN, I'm writing this plugin/package/whatever in javascript, how do I parse EDN with javascript?

You know what? Put some fs in the chat because this js is going bye bye.

Clojurescript can parse EDN, no problem.

Switching from javascript to clojurescript

It sounds horrible, rewriting everything, (all 50 lines or something 😅) and starting over with a new thing I don't have any understanding of. I've seen a ton of blog posts with lines and lines of edn and complicated installation procedures to get clojurescript compiling with something called figwheel? It's intimidating. Luckily the clojurescript folks have made things really easy, but it's hard to tell if you're just starting out. Here's the secret sauce to making clojurescript running in node.js work for you.

Step 1. Make a file named build.clj

; build.clj

(ns build
  (:require [cljs.build.api :as b]))

(b/build "src"
  {:output-to "whatever/whatever.js"
   :output-dir "target"
   :optimizations :simple
   :target :nodejs
   :output-wrapper true
   :pretty-print false
   :hashbang false
   :main 'your-app.core})

Step 2. Run it

clj build.clj

Step 3. Profit

That's it, if you really want to get fancy and don't feel like re-running that every time you change something, you can make a watch file that is the same:

; watch.clj

(ns watch
  (:require [cljs.build.api]))

(cljs.build.api/watch "src"
  {:main 'your-app.core
   :output-dir "target"
   :target :nodejs
   :output-wrapper true
   :pretty-print false
   :hashbang false
   :optimizations :simple
   :output-to "whatever/whatever.js"})

… and run clj watch.clj instead. Now when you change any clojurescript file in the src directory, the js will be output to the :output-to directory.

That was easy.

With that out of the way I could finally finish my prepl client:

1. Connect to the running clojurescript prepl socket

(def element (.createElement js/document "div"))
(-> element .-classList (.add "cljs-repl"))


(def modalPanel (-> js/atom .-workspace (.addModalPanel #js {:item element :visible false})))

(def input (.createElement js/document "input"))
(.setAttribute input "style" "padding: 7px")
(set! (.-type input) "input")
(-> input .-classList (.add "input-text"))
(-> input .-classList (.add "native-key-bindings"))
(set! (.-value input) "localhost:5555")
(aset input "onkeydown" (fn [event] (condp = (.-key event)
                                      "Escape" (.hide modalPanel)
                                      "Enter" (connect input)
                                      "")))
(.appendChild element input)


(defn connection []
  (.show modalPanel)
  (.focus input))


(defn connect [input]
  (let [value (.-value input)
        [url port] (.split value ":")]

    (-> client (.connect port url (fn [])))))

2. Send the code over

(def client (net/Socket.))

(defn send []
  (let [editor (-> js/atom .-workspace .getActiveTextEditor)
        range (.getCursorInClojureTopBlockRange EditorUtils editor)
        s (str (.getTextInBufferRange editor range) "\n")]
    (helpers/log s)
    (-> client (.write s))))

3. Parse the EDN

(-> client (.on "data" (fn [s]
                         (helpers/log (pr-str s))
                         (let [data (cljs.reader/read-string (str s))]
                           (-> js/atom .-workspace (.observeTextEditors #(update-repl % data)))))))

4. Show it in a new pane!

(defn update-repl [editor m]
  (when (= "CLJS REPL" (.getTitle editor))
    (when (contains? m :form)
      (do
        (.insertText editor (str (:form m)))
        (.insertNewline editor)))
    (.insertText editor (str (:val m)))
    (.insertNewline editor)))

Everything is awesome

Saying this is the world's worst clojurescript REPL is pretty negative, so I wanted to say something positive at the end of this post and now I have. Everything is awesome.

If you want to make a much better repl than the one I've made, you can stand on the shoulders of regular sized me and get a head start: https://github.com/swlkr/cljs-repl

How I made the world's worst clojurescript REPL

Sean Walker — Thu, 09 May 2019 16:36:47 +0000

TL;DR
I tried to run two clojure web servers and call a JSON endpoint and the JVM fell over on my cheap 512 MB VPS server, so I switched to clojurescript and wrote my own prepl client for atom

My clojure deployment workflow in a nutshell

I use atom in vim mode for clojure development and I made something cool, something for all clojurists to enjoy! The idea is to scrape conference websites, and get json from the HN api, the clojureverse api and the reddit api for clojure related news and put it all on one website. I'm not really an expert at all things JVM, but I know enough to be dangerous with clojure:

build the classpath
compile to bytecode
make sure you have a -main function

…those sorts of things. Unfortunately, I did what I normally do with a twist, instead of building a whole uberjar, I simply compiled to bytecode with the help of this post. I then put it on my 512 MB VPS server and it won’t even do the thing where it downloads some json. It runs for a while and then unceremoniously reports back… Killed.

My Cheap VPS Dreams Were Killed

That’s it, that’s all the process says, no exceptions, nothing, just killed. I set the JVM to only use 100 MB of RAM but not even that is enough on a VPS running two other JVMs (in addition to two clojure REPL servers). I decided then and there that I had had enough, I decided to switch to clojurescript on the server. Surely v8 uses less resources than the JVM running clojure and my full stack framework coast 🙄

In my side projects I don't look for solutions, I look for excuses.

How does clojurescript in node.js even work?

Now having the perfect excuse to switch the project to clojurescript, the question is how do I get my current atom repl client chlorine working with clojurescript? Wait, back up a second… how do I even get clojurescript running on node to begin with?

DDG to the rescue!

Here's the breakdown of the clojurescript tooling landscape as I understand it. There are three major ways to get your clojurescript files compiled down to js so that node can understand them:

I'm not supposed to give my opinion but…

shadow-cljs looks like lein and after tools.deps I'm not crazy about the sheer amount of edn I have to understand to get projects working, especially since node is supposed to be as easy as node index.js
lumo looks interesting, but I'm not sure what I gain from using self hosted clojurescript vs the clojure bootstrapped version
clj is familiar to me and it looks so simple, just write a build.clj file like this:

(ns build
  (:require [cljs.build.api :as b]))

(b/build "src"
  {:output-to "main.js"
   :output-dir "target"
   :optimizations :simple
   :target :nodejs
   :main 'your-project.core})

and run it with clj build.clj. It doesn't get much easier than that, I didn't even have to install anything other than clojure!

So, now that I've settled on clj I can use chlorine and… oh wait, it only supports lumo and shadow-cljs. At this point, a sane person would say ok I value my free time, I'm going to just use one of those and I can continue on to my real goal of putting out this clojure website for aggregating clojure info in one place. If you've gotten this far you know I'm not a sane person.

Insane in the membrane

I want to run this in my terminal:

clj -J-Dclojure.server.node="{:port 5555 :accept cljs.server.node/prepl}" -m cljs.main --repl-env node

and I want atom to do the right thing: connect to it and allow me to send bits of clojurescript code to it. I don't know why this is so complicated, clojure is supposed to be SIMPLE. Eff.

…Stay tuned for my next post where I talk about all things atom package development and how I switched from plain old javascript to clojurescript grudgingly just to read the EDN from the prepl server.

How did my password manager do after launch?

Sean Walker — Fri, 01 Feb 2019 23:29:09 +0000

📅 02/01/2019
🔥 31 day streak
💰 $4.99 price
🤑 $18.60 revenue
📈 7 sales
🚀 168 upvotes
⌚️ 57 hours spent
😭 1 Rewrite
🛬 1 Lander with Source
💻 114 files changed, 250127 insertions(+), 559 deletions(-)
🏁 Today's goal: ~~Reflection~~

What went right

Honestly, I couldn't have asked for a better launch. I know some solo devs out there make a bigger splash, but that's just not me. I don't plan on quitting my job and I don't plan on moving to Bali. I'm a regular guy who has a house, a wife, a dog and a life outside of software development. I would love for this to be a much bigger source of income for me, but the truth is I'm not about to make any crazy sacrifices to switch bosses from a real human who is amazing to product hunt and twitter's algorithms. Anyway, the things that went right were 168 upvotes on product hunt, and 7 sales! That's six more sales than my last project I actually launched got! Amazing!

What went wrong

So many things, I started to spend a lot of time on it towards the end there instead of trying to keep it to max 1.5 hours a day for the whole month, which is why I spent 57 hours making it instead of the 30-40 I wanted to spend on it. That's the thing with life outside of school or even outside of a company as a full time employee. Effort is not rewarded, the final product is. So that's why companies skirt around things and make them cheap for scale and only a few companies make quality things for sky high prices, because the faster you can get things out into the world and charge money for them, the more money you can potentially make. Alright, the list of things that went wrong:

The lander never had the mini-game I wanted to put in there with the clouds and +1's and getting a discount.
The mac app store version wasn't approved until halfway through the launch day 😅
Lots of other little things

In the end

I could have done quite a few things better, but that's the point of this whole process of shipping apps quickly instead of perfectly.

One thing that did happen to me, was quite a few of you started following me on here. That's pretty cool. One thing I've learned recently is that consistency is the best thing you can do for anything in your life. When you let other people in on that consistency? It makes for at least a little success.

What's next

So, after the whole month of January, my most productive month, EVER. I thought I was going to take a break, I'm not.

I think I'll give the dev.to servers a break and not post every day, but I will be posting every day on my own blog and I'll be attempting to livestream for 30 minutes every day.

Making small changes every day instead of striving and making a lot of changes sporadically has worked out much better for me. Version 2 and the iOS app are on their way!