MowDown Developments

Let's be honest. No one's first app is good. But that's not the reason I pulled NagTasks.

To be honest, I never bothered marketing it, and never got more than 10 downloads in several months. But that's not why I pulled it.

No. The reason I pulled it was because I lost my keystore. Forgot the password. Embarrassing. You can't update an app without its keystore.

So, what happened to Vault? Well, I presented it just fine for my Finals at Brookdale. (I say, typing this from my dorm at NJIT.) It was pretty full featured, but pretty buggy from not having anyone but myself testing it in ways I considered normal. Amazing how much crap I missed when I had to do a demo. I tried to fix most of the bugs later, after getting my passing grade, but there was one that was giving me massive headaches. The damn stack trace wouldn't point to a problematic line of code that was throwing these NullPointerExceptions. Weird, that. Since then, it's been sitting on the market as a private (one-man) alpha because... you guessed it! I lost my keystore again.

It's probably sad to say this, but while I fooled around with various programming languages as a kid (some BASIC, making stuff in RPG Maker, etc.), the bulk of my general knowledge of programming comes from college courses. Community college courses, specifically. (I said it was sad!) And yet, until this semester, there was once concept in Java that was foreign to me, and it very clearly showed in any attempts at Android Development, and that was the Interface. Even with my Data Structures class leaning heavily on Interfaces, I still never got the use. So, they're empty classes that you use as a template? Umm... cool?

It took me a few hours staring at my own final project for my major and the auto-generated code therein for me to realize what the hell they can do.

Biggest example? Well, if I wanted to call an activity's method from an fragment, in the past, I might have done something like this:

((MyActivity) getActivity()).myMethod();

That works fine, if you know that the activity is going to have a method. But what if you don't? Well, if you make an interface within your class and implement that interface, that can't happen. Because they have to have that method.

Now, this is all well and good, except that this is what Google themselves prescribe right in the API guidelines. So, how is this news? Well, I had to figure this out on my own, for one. For another, I used this knowledge to create getters and setters for variables that may be in one activity but not another, with the assumption that a fragment could end up on one of many activities. Global variable not actually available? Just have the method fail without throwing anything. No muss.

So, to kind of give me an incentive to write in at least one of my Tumblr blogs, I thought I'd make up a little segment where I take an app that I'm using in my daily life or found interesting and give it a slight look over and critique it. Today? Well, I used to be an avid SugarSync user. (Key word: used to be. Dunno why I tapered. Probably a lack of Linux support after I started to use Ubuntu more.) And recently, they launched their big redesign. New apps, new everything. The old app was serviceable, but it lacked features that competitors have. So, how does it stack up? Well, let's start from the first thing people see...

So, a while back, before I started on Vault, I tweeted about the stupidest project I've ever started. Since that's clearly going to be shelved for the sake of focusing on Vault, I might as well let slip what it was about, in case I haven't. But first, watch this.

So, what was that? Well, I have a hobby of making YouTube vids of me playing various games, and this was actually something that got me thinking. I'd like to get my name out more. Everyone does. And to be honest, I'm terrible at marketing. So,this advice seemed like something I'd run with. Except I'm lazy. And trying to tag a bunch of XCOM: Enemy Unknown videos in the manner he describes in that video... well, there were too many variances. I didn't want to have to do 50 different entries. So I did what any good programmer would do: make something to automate the process. It's more work in the long run, but it'll save you time down the line.

So, a while back, I got in touch with the developer of Loot. I like Loot. I use it all the time, but the UI is definitely stale. Anything that still relies on the menu button in this day and age drives me batty. I knew for a fact that the dev was too busy to keep maintaining Loot (he had stated so on reddit). But someone needed to do something to do something to update the UI to be compliant with the current design guidelines. I asked if he would open source Loot. He said, "Sure." And then I forked it on Github and everything was fine and dandy for when I decide I have enough time to work on it

Flash forward to yesterday. I've never worked on someone else's project before, so I thought it'd be a good idea to make my own package for my additions. I thought it'd be as easy as changing the main package, switching the min and target SDKs from 3/8 to 14/17, clean out all the deprecated code for more modern equivalents, and then adding whatever stuff I thought would work better. Sadly, apparently, I don't know how to Android Development Tool. Anytime I'd try to add stuff to my package, it'd go into the Loot package. And eventually, the Git plugin I used for Eclipse refused to push to my repo.

Eventually, I just said "Screw it" and decided to start fresh. There were some bits of his code that, in spite of all the studying of it I did, I didn't quite get. Not to mention he made his own database class and I had no idea why. And to be honest, I'm planning on this as my project for my final semester of my CompSci major at Brookdale. (Yep. Still a student.) It feels more right if I'm doing this all by myself. So, while I'd like to throw this all into the fork of Loot I did, in the end, this is all going to be my code. If I end up borrowing code from Loot, I'll freely attribute it. But I'm probably just going to be looking at how I use Loot and try to replicate its functionality.

As someone with hopes and dreams of making my own games one of these days, it excites me to see the newest version of LibGDX released. With all my programming practice having gone into Java, LibGDX seems like it'd be my tool of choice for game development.

So, reddit just clued me in to this: a NaNoWriMo for app developers. This is something I can get behind. I signed up for their e-mail, so we'll see what comes of it. As for what I'll do? Well, keep reading.

I have an idea for what I want my next project to be. It'll be free and open source. Mostly, I just want to improve an app I use frequently. The problem is that I feel I need time to prepare for it. As I'm starting my next project with an existing open-source base, I kind of want to study his code. It's really old code, since the old maintainer stopped updating it two years ago, and I don't know whether I can't spot clean code or if it really is that sloppy. Or maybe it's because he's catering to Cupcake still, as that's labelled the min SDK. (WHY WON'T CUPCAKE DIE?) I'll probably give it a thorough 4.x makeover. It really needs an ActionBar to allow people to add entries.

I hate to use what's supposed to be essentially a developer blog for whiny rant-type stuff, especially since I took the time to shell out money to buy it a real domain name instead of leaving it as mowdowndev.tumblr.com. But lately, programming headaches just pop up. And typically, when programming headaches pop up, a programmer's first stop is Stack Exchange, particularly Stack Overflow. And therein is my headache.

You see, I have a problem with Stack Overflow. That problem is that no one wants to help me. Ever. My most recent attempt at getting assistance with a problem just created more problems. It felt like I was being trolled, being told the problem was something I did when I clearly provided evidence to the contrary. Even now, I invited the guy to chat about the issue, and he just doesn't bother.

Now, this issue that I'm asking for help with is a minor bug with the app I've been writing about so much here. I noticed it in the emulator and brushed it aside. My mother, who I managed to wrangle into being a guinea pig, also noticed the bug. Being as tech-illiterate as she is, that raised a small red flag when she noticed it. So, off I went to fix it. And every change I made in the XML for changing the EditText's InputType just led to more frustration. The documentation certainly made it seem like I was doing everything right. So, I was stumped.

This isn't the first time I've found Stack Overflow worthless. I've asked for help there numerous times, only to solve my own problem days later because no one was willing to help me. But most of those I would just chalk up to that tendency following me from my time when I used to play City of Heroes. (I would joke I was the bane of Customer Service.) This time was the most egregious, with the repeated "You did something wrong".

In light of Apple recently being granted an injunction on the Galaxy Nexus, and all the rage that followed, I felt the inspiration to update one of my two ailing blogs. So, with that in mind, here are the reason why I, as a lone gun developer, will never develop for iOS.

Cost. Let's start with the most mundane. In order to develop for iOS the proper way, you must own a Mac. There is no way around it. The iOS SDK only exists for Macs. I, literally, do not have $1000 in any of my bank accounts. A new computer, much less a Mac, is nowhere near feasible at this time. Then we get to the cost to publish your app. To be a licensed Apple developer, it's $100 a year. And Apple takes a 30% cut of any sales you make (which has become the standard). So, assuming I sell an app for 99¢ and that I buy a 13-inch Macbook Air to develop it on (we'll say the base $1200 model, even though its specs are laughable), in order to make up for my initial investment, I need to sell 1858 copies of my app. And I need to keep selling 130 copies of that app every year after that to keep making a profit. So, what about Android? I am developing for Android, so how's that? Well, the Android SDK runs on every platform. Mac, Windows, Linux. (Yes, Linux. I do a good deal of development on Ubuntu Linux. More on that later.) So, I don't need to invest money in new hardware. So, how much does it cost to be a licensed Android developer? Just a one-time payment of $25. Yes, they take a 30% cut of any sales you make, but do you see how much lower that bar for entry is? Instead of selling 1858 copies to be profitable, I only need to sell 34! And if sales flatline after that? No biggie! I made my profit. I don't need to keep making sales after that. And that's only if you want to sell via the market, which leads to my next point...

No Walled Garden. If you want to sell for iThings, it's iTunes or bust. There is no alternate market. (Well, Cydia, but since that involves doing some hackery things to your iDevice, that's probably not something most people will see.) With Android, I can post my apps to alternate markets if I want to. I still have Handango bookmarked in Firefox, because the developer for this one voice-recording app I used to use to record my lectures didn't have access to selling paid Android apps in his country at the time. Amazon's gotten into this space pretty big recently, and their fees for posting to their store aren't too bad. (Okay, so it's still the same $100 a year, but the first year is free! That's how you hook them in!) Or, I could even make my own website to sell it if I wanted to. Heck, if I was so worried about lawyers that I never wanted to publish my app at all and just wanted to hand it out to my friends, I could! Just have the SDK make me an APK, and I can e-mail that to all my buddies. There's absolutely no way to do that on an iPhone. (Okay, Cydia again has something for that, but again... not something normal people do.)

Apple treats their devs poorly. For the longest time, the approval process for getting your app into the iTunes app store was shrouded in mystery. Nowadays, the guidelines are actually published, but they'll still reserve the right to reject your app for reasons not mentioned in the guidelines. Like they want to steal your app and claim it as their own. Google knows better than to abuse their devs. After all, they have the ability to take their business elsewhere.

Open Source. This one probably doesn't bother the non-developer types out there, but from I've heard, and this is second-hand, is that alot of the legal stuff that you agree to when you become a licensed Apple developer runs against the majority of most open source licenses. So what does that mean? That means if I've written something cool and want to share what I did with every other programmer in the world, I can't. Which means that if I need help with something I'm writing for an iThing, it's much harder to find examples to work off of. Possible, but harder. Meanwhile, Android is 99% open source. (That 1% is all of Google's proprietary apps that are licensed separately from Android itself.) They allow you to develop for an open source platform (Android) on an open source platform (Linux). They even have their own repository service for hosting open-source code! That's awesome.