Category Archives: Development

iPod Touch

Why doesn’t Apple improve on the iPod Touch every couple years? Why not take older chip sets — maybe a model from last years iPhone — and leave the form factor alone? Just keep stuffing updated tech into the existing design.

There are folks that can’t afford to buy their kids an iPhone but the iPod Touch may fit nicely into their budget.

Yep, it's a rooster.As a developer this is also a nice machine for testing. The only issue I see is it hasn’t been revved in a long time. It’s still using an A8 processor but it is still slightly ahead of the iPhone 5s which uses an A7. that’s important because the iPhone 5s is the lowest end iPhone that supports iOS 12. That means we could get another couple years use out of the current generation iPod Touch.

The iPod Touch with 128GB of storage sells for $299US.

Fingers crossed Apple updates it soon.

Just Code

Becky Hansmeyer: “For instance, I can take just about any Objective-C code and do a word-by-word translation to Swift that will compile and run, but it won’t be very “Swifty.””

The notion of being Swifty kind of bugs me. If you work in a language long enough you’ll naturally gravitate to language conventions or you won’t. If you’re an Indie, like Becky, I don’t think it really matters how Swifty your code really is. What matters is you’re willing to learn and adapt with the times and you can ship. If you don’t ship you don’t get paid.

I’m probably not very Swifty either, but I don’t really care. I can write code in Swift and it works just fine. During code review with my peers they’ll let me know of a more Swifty way of doing things so I can either make a change or get it the next time. I’m willing to do the new thing and I think that’s what we need to be willing to do if we’re going to survive as developers.

Be ready for change and adapt. That doesn’t mean you have to do it all on day one. Just be willing to make the change.

Higher

Quora: “Total cost for creating Tinder-like app for one platform would be about $5k – $10K.”

Charging $5-10$K for a Tinder-like application is how you go out of business. I’m speaking from experience here.

Watch out for the blowfly.Back in 2014 I made my second run at freelancing. This time around I failed. I learned a lot about who I am and how not to  run a business. Two of the primary reasons were charging too little for my work and not being a blazing fast coder.

I’m going to step out on a limb here and park a Tinder-like app — a service really — at around $100,000.00, not 10.  Call me crazy, it’s fine. Let’s take a look at some estimates given by folks that have actually built some great software.

First up, Craig Hockenberry: “With such a short schedule, we worked some pretty long hours. Let’s be conservative and say it’s 10 hours per day for 6 days a week. That 60 hours for 9 weeks gives us 540 hours. With two developers, that’s pretty close to 1,100 hours. Our rate for clients is $150 per hour giving $165,000 just for new code. Remember also that we were reusing a bunch existing code: I’m going to lowball the value of that code at $35,000 giving a total development cost of $200,000.

Emphasis is mine.

In the same Stack Overflow post we find Jonathan Wight weighing in on the cost for the Barack Obama App: “The Barack Obama app took 22 days to develop from first code to release. Three developers (although not all of them were full time). 10 people total. Figure 500-1000 man hours. Contracting rates are $100-150/hr. Figure $50,000-$150,000. Compare your app to Obama.app and scale accordingly.”

Both of these guys are world class software engineers. I have no reason to doubt their estimates because they have the experience needed to build any application.

Here’s Kyle Richter of the excellent Martian Craft: “In February 2013, the average cost of a house in the US was $152,000. By our estimates inside of MartianCraft, the average cost of an app is approximately $120,000.

If you want a high quality application be prepared to pay for it. Sure, you might be able to get your Tinder-like experience for $5-10K, but it may behave and scale like a $5-10K application and service.

Know what you want and what you’re getting into before you talk to developers about building your dream application or service.

Far from perfect

Free Code Camp: “VS Code’s success story is interesting because it’s far from perfect: its UI has that Microsoft-y clunkiness that seems to infect all their products, it’s a big resource hog, and it can be kinda slow to initialize.”

If you go read the article and don’t see the irony in the above statement I’ll point it out for you. Visual Studio Code uses Electron which allows developers to write desktop-like applications in JavaScript. Is the author saying Electron is slow, JavaScript is slow, a combination of the two, or that maybe Microsoft developers made it a fat — slow — pig?

Making your iOS App embrace the iPhone X notch

After yesterdays announcement I decided I’d rebuild RxCalc with the iOS 11 SDK. I figured things would rebuild and my app would fully embrace the new iPhone X without change, but it didn’t work. When I built the app it ran in what I’d call letter boxed mode. The top and bottom were cropped. It was like running an older app on iOS 7, if you’ve ever seen that.

UGH! Why you letterboxed bro?
UGH! Why you letterboxed bro?

I was puzzled. I tried all kinds of jiggery-pokery and nothing worked. So I finally did what any self respecting lazy developer would do. I asked Twitter for help, and Jeff Johnson saved the day.

I was — quite honestly — surprised something so little could make it work, but darn it, you gotta trust the experts. I added a new launch screen at the recommended 1125×2436 size and it worked! Yippee!

Hey, that worked!
Hey, that worked!

So, thanks Jeff! Also, if you’re looking for a bonafide Mac and iOS Developer you should consider reaching out to Jeff — he’s available for work — and has a history of shipping great Mac and iOS Software.

And remember — EMBRACE THE NOTCH!

Work Note: Loading View Controllers from a Storyboard

When I need to load a view controller from a Storyboard I like to create an extension to the view controller’s class and add a class function to it that does the work. It keeps things looking clean in the code where you use it.

I’m not sure if this is smart or dumb. I’m sure very smart people will let me know. Here’s an example.

https://gist.github.com/Fahrni/537fd42c38882a007a1c506369f3f2c6

Here’s how you’d use it in your code.

https://gist.github.com/Fahrni/fb03913774f026787892cd2d42ab741b

Old Apps

A cute little monkey.I got an email from Apple a week or so back letting me know I needed to upgrade one of my apps to 64-bit. I knew right away it was Arrgly but wasn’t really sure if I wanted to update it.

On Sunday morning I received an email from an Arrgly user asking if I was going to update it because he likes the app. It felt good to know someone else found my goofy app useful. I decided at a minimum I’d publish a little framework someone else could use to write their own version of the app if I couldn’t get to it, or couldn’t finish it on time.

It took less than an hour to create the project and get it published. If you’re doing Mac or iOS development and you need to shorten or expand URL’s for a YOURLS based service you’re welcome to use YOURLSKit. It’s all Swift 3 but should be fine with Swift 4 projects and iOS 11 or macOS High Sierra. It’s a tiny bit of code, but it does the job.

Scripting iOS

Last week Apple acquired automation workflow application Workflow. Of course there was a nice buzz around it and it was a big topic of conversation on various podcasts and websites.

This, of course, got me thinking about automation. I’ve always been a fan of open API’s and the ability to automate applications. We’ve also seen recently that Omni Group is opening up OmniGraffle to automation via JavaScript.

Back in 2010 x-callback-url was created as a way to allow applications to call into each other and return results so you could chain together calls to build custom workflows. Apps like Launch Center Pro and Workflow took advantage of x-callback-url to let you build those workflows and execute them. Now we have a bonafide standard, without a standard. The app ecosystem found a way to support automation without Apple’s help.

I’ve used Launch Center Pro but until recently I’d never used Workflow, and it’s pretty amazing. The Workflow guys did an amazing job creating a drag and drop UI for building what amounts to a program. Well worth a look.

So, this brings me to what I’ve been thinking about over the past few days. Given x-callback-url and App URL schemes in general it would be extremely cool to use those to create object hierarchies using JavaScript. Why JavaScript? Well, it’s native to iOS and applications can use the runtime. Given the advances made by the Workflow team why not take it one step further?

Allow applications to specify a  Scripting Dictionary or Type Library as part of the application bundle, this should allow runtime creation of objects. I know this isn’t rocket science and it’s been done many times over.

Short of adding support to the OS it would be pretty sweet if an App like Workflow, Launch Center Pro, or Pythonista would standardize on a way to parse a URL Scheme into an Object Hierarchy.

I’m going to use Evernote, Bear, Overcast, and Arrgly as examples.

What do you mean by Object Hierarchy?

That’s a good question. Here’s what I’m thinking. Since the Apps mentioned above all support URL Schemes we can derive an Object Hierarchy from them. Basically the beginning of a URI begins with a scheme. The scheme is the name. In the case of Evernote it’s evernote. Pretty simple, right?

Given the scheme name we follow that with a path. In the case of x-callback-url based URL schemes we will skip over that part and move to the second item in the path. This will be the action, or function, or the object we’re going to execute.

evernote://x-callback-url/new-note?type=text&title=EC%203

The above URL will tell Evernote to create a new note of type text with a title of “EC 3”. If we had a way to parse that in a runtime application we could present the user with an Object that has methods that take arguments, like this.

evernote.new-note(type, title)

Let’s do a couple for Bear. First the URL Scheme.

bear://x-callback-url/create?title=My%20Note%20Title&text=First%20line&tags=home,groceries

Now translated into code

bear.create(title, text, tags)

Overcast URL Scheme.

overcast://x-callback-url/add?url=

Code

overcast.add(url)

And finally, my favorite, Arrgly URL Scheme.

arrgly://shorten?url=

Arrgly Code

arrgly.shorten(url)

Pretty simple to turn all of those into objects. When I say you can create a hierarchy it means you could, by convention, lump groups of actions into objects, or like the above examples have a set of actions that all live on a single object.

Here’s what a object might look like as a URL Scheme.

thing://x-callback-url/document/add?title=
thing://x-callback-url/document/delete?id=

That would result in using it like this

thing.document.add(title)
thing.document.delete(id)

Of course this need more fleshing out and it would require app developers to decide on a well known convention to make it work as expected, but it could be done with a bit time and effort. It could be these become an extension of the x-callback-url specification?

MacBook Monster

I listen to a podcast called Accidental Tech Podcast. If you’re a Mac or iOS Developer you’ve probably heard of it or you’ve heard of one or more of its hosts. One of their ongoing topics of conversation is Apple’s apparent lack of focus on the Pro market. It’s true Apple has become very focused on more consumer oriented products, like the iPhone. I mean, who wouldn’t? When you look at the numbers it makes total sense.

When Apple announces results sites like Six Colors do a great job breaking down all the numbers and, in Six Colors case, they make really awesome charts! Just look at this one from January of 2015.

Awesome Six Colors Revenue Chart
Awesome Six Colors Revenue Chart

Who can blame Apple for spending most of their time on the iPhone? Look at those numbers. They’re stunning. It’s not to say the Mac or iPad are losers, they’re not. Most companies would give anything to have one product doing so well, Apple has at least four; iPhone, iPad, Mac, and Watch. Most likely someone will write in to let me know I’ve missed one, or I’m wrong about something, but you get the point. Apple is killing it, on multiple fronts.

What about Pros?

Much of the consternation from the ATP guys revolves around the Mac Pro. Who can blame them? Apple hasn’t shipped a new model since 2013. The hardware is now embarrassingly outdated and is due for a much needed refresh.

As a developer I can understand their need for Phenominal Cosmic Power! But they’re stuck with an Ity Bity Living Space. Not fun.

I’d love to see Apple pull together a great new piece of hardware that includes all the latest, greatest, internals and really appease the Professional Apple Workstation crowd. This would include filmmakers, photographers, designers of all kinds, the CAD folks, and, of course, Software Engineers of all kinds. On my list of nice to haves would be a box large enough to hold multiple multiple core processors (say 256 cores, why not? Windows can do it), tons of RAM, and a killer external bus system to allow folks to chain together external GPU’s. That would be a really great computer, don’t you think? Maybe the base configuration is the most popular, but it sure would be nice to be able to take a Mac and macOS to an extreme level.

Having said all of that, as a Professional Developer I’d prefer a really great Portable Workstation.  What is that? Well, it’s a desktop replacement in a laptop form. Apple is obsessed with making everything thinner and lighter. I do appreciate that, I really do. My first MacBook Pro was a lovely 17-inch beauty, but it was big. Not only heavy, but it was long enough that it was difficult to find a decent backpack to carry it in. Let’s just look at the weight alone, it was 6.6 pounds.

I remember carrying that thing around my first WWDC in 2011. Early morning to late afternoon. Yes, I got tired of carrying it, but it wasn’t a hardship. I would happily trade a bit of weight for a super powerful MacBook Pro.

I read an article recently on the new Dell Precision 5520 on Windows Central. This thing is a beast.

“Besides the big Core processors, all of which are the last generation “Skylake” variants, there is the big kahuna with the Xeon E3-1505M v6. Introduced this year, this Intel Xeon is a quad-core processor with support for ECC memory, 8MB of cache (up from the usual 6MB in a consumer Core i7), and a slightly higher base clock rate of 3.00 GHz with Turbo up to 4.00 GHz. These specs make it one of the fastest mobile processors around, besting the Core i7-7700HQ found in the XPS 15 (9560) by 200MHz.”

Yes, you read that right. This laptop can be configured with a Xeon processor, not to mention 32GB of RAM. That is professional sized horsepower in a small package. This is what I’d like to see from Apple. A true high-end Portable Workstation. Oh, and do you remember the weight of that old 17-inch MacBook Pro I mentioned above? Yes, 6.6 pounds. This Dell weighs 4.56 pounds. Compare that to the new 15-inch MacBook Pro which weighs 4.02 pounds. It’s not that big a difference.

I’d love to have this kind of power wrapped in Apple’s design ascetic. Oh, it should only come in one color. Black.

Work Note: git flow feature start fails

This morning I was creating a new branch to start work on a new feature and I use git flow.

To start a new feature you type…

git flow feature start FeatureName

After typing that and pressing enter git flow complained that it was initialized and I should run the command…

git flow init

I’ve been working in this same environment for two years. I know git flow is configured, but like a good automaton I followed the directions. Of course, you can guess what happened, right? Yep, git flow init says “Hey, you’re good to go!”

I wasn’t. My commands still failed with the same message.

Luckily there is a fix.

Yep, that’s right, cd into the .git directory for that repository and remove all git flow references from your git config file. Then run git flow init. All better.