A Programmer's Experiment

The first couple days were continuing with my Minecraft mod, finishing up the amethyst materials in order to make way for my next gem. That was the first hour - imaging the tools in GIMP, writing their code, generating the JSON files, and finally compiling it all together to make sure everything is running. 

And, as it has been the past week, everything went smoothly - until I really began testing the results.

[ To preface: my testing up until this point has been rather lackluster. Basically I would just make sure the items appeared in the creative tab, had the correct texture, and the proper crafting recipe. ] 

This time, I decided to do a full test in vanilla. Giving myself some basic items, I found a vein of the amethyst ore in a cave and started mining with my ruby shickaxe. Boom. Problems. 

Turns out the shickaxe tool I was so proud of a few weeks ago has a few issues with custom blocks. It will mine vanilla ones as expected, but anything beyond that leaves a little to be desired. 

Y U DO DIS!? 

So what was expected to span a day turned into a running tally. I spent two hours the next day working on it, trying everything I could think of in both the shickaxe and ore files to make it work. 

[ Side note: I find it amusing that, given how much I complained Minecraft didn’t allow me my usual preference of test-driven development, the only way I’ve been able to figure this out is make tiny changes to the code, compile and see if it works. ] 

So far nothing I’ve tried has worked. I’m considering reworking the entire code for the shickaxe to see if that fixes anything, or perhaps (if it comes to it) scrapping the idea all together and just making individual pickaxes and shovels (though I have no guarantee those will work either.) 

Ahh Minecraft. Just when I thought I had you figured out. 

But the next day, fueled by my frustration, was incredibly productive. I jumped back to AlexaWeb and started to fill in some of the missing elements. 

The first was my “About Me” section. Despite having to write (or speak) about myself at least a few times a year, I never quite know what to say. Hey, I was a journalist - I’m used to writing about other people’s accomplishments. 

But with the help of NPR’s This American Life and a nice, cold soda, I eked out a couple paragraph/sentences about myself. Again, the entire thing is still in it’s formation stage, but I’m rather pleased as it stands. 

TLDR: Look at me! I can adult! Or at least write coherrently coherently. 

To take a break from writing words and go back to writing code, I played around with the footer of my page. I’m still finding a good spot to put my email (mailto links aren’t working with my laptop, so I have no way of testing whether or not they work) but I was able to link my GitHub and LinkedIn pages nicely. 

All and all, I’m still chugging along. The next couple days will be rather interesting as I finish this project while simultaneously moving into my senior year of undergrad. My goal is to have my website completely finished, or at least to a final draft. 

While I’ve been thinking about new content to add to the game, this latest update has been mostly implementing ideas I had earlier this summer. Namely: amethysts!

Much purple. Very hard to spell correctly. Wow.

Continuing with my gems trend, I chose amethyst namely due to the colour. The process was exactly the same as creating the ruby materials, with the added pressure that I rewrote the word four or five times before finally caving and googling the proper spelling. It took a bit of effort.

amythest, amethest, amythyst, amythist, ...

The second hurdle was dealing with the custom ore generator - but once I (with help) realized I was trying to spawn an ungodly amount of ore blocks at the same time and crashing my game, that, too, passed easily.

Once these were dealt with, actually creating the components were rather easy, and I began poking at some of the methods to see what else I could change. Ah-ha! Ore drops.

Previously, the method responsible for the number of items (rubies, amethysts, etc.) was hard-coded to one. In vanilla Minecraft, diamonds (what I’ve been loosely modeling my gems after) drops between 1-4 diamonds per ore block. So I quickly set up a rng and reset the drops.

Not pictured here: the 10 minutes it took me to realize my original code allowed an ore to drop nothing, and my subsequent rush to fix it.

Next I started working on the tools made of the gems. As of now I have a ruby axe and hoe, and an amethyst sword.

Again not pictured: my joy in discovering I don’t screw up the JSONs anymore.

Even more than that, I finally got around to creating custom tool materials, rather than relying on the vanilla emerald (for some reason diamond-level materials are classified as emerald in the code.) Both ruby and amethyst lie between iron and diamond in terms of stats, with amethyst being just a twinge better.

As always there are a bunch of things I still want to add - the rest of the amethyst items, modifying the axe to mine faster onLeftClick, adding more gems such as sapphires and actual emerald tools.

When I last worked with this code, I was struggling with some aspects of CSS and how they interacted with my subgroups (Projects and Music.) So that was where my work began. As usual, it took quite a bit of trial and error, but once I figured out how to pull the proper strings, setting up the rest of the page was relatively easy. 

Look! Actual website! With pictures!

After finishing the rough layout, I set my sights next on the music section. While having some description of the instruments I play and the work I’ve done will be nice, I really wanted to incorporate actual media. 

After some searching, I found Cincopa. It seemed pretty easy to customize and even easier to add to the site, so I uploaded a couple of my songs to test it out. After pushing the update to Github, I refreshed the page and was pleasantly surprised to see everything working. More or less. 

While the colours do go with my overall theme, it’s safe to say I’ll never make a living as an interior designer. 

The update thus far. It still need some work, but I really like how it’s coming together. 

The original concept had us reading in the song information from a file, storing it in an ArrayList, and using that information (with some input from the user) to create a playlist based on genre, artist, or mood. 

I wanted to keep a lot of that structure and add more, so my first step was to grab the old code and see what I could run of it. 

Not much apparently. 

After fixing the initial mini-bugs that always come with importing an existing project, I realised there was a huge issue with the Player I used to actually play the file - in short, it did not exist in my library. 

Do you want bugs? This is how you get bugs.

Apparently in my haste to get something working for the Hackathon, I found an older library called javazoom that worked with media files. After some googling, I found the site but as it was last updated in 2010, I figured there had to be other, more recent libraries to explore. 

And I was right. Kind of. 

After much digging, I came across javafx. Truthfully, I’m not sure of its full scope, but it does have a media player and related functions, and a rather nice API. So I started tinkering around with it, eventually figuring out how to play a song. 

Yup. It was Taylor Swift. Because my taste in music rocks. 

At first things were going up. I rearranged the code into a semblance of structure and began poking at the other methods trying to see what else worked. (Including figuring out how to control the sound - after nearly blowing my ears out, twice!) 

Unfortunately that’s where my progress has stalled. With the exception of pause() and stop(), the methods are less than obvious and when I went to run a full song as my last test, I discovered it stops after about 30 seconds. 

At least it’s pretty? 

I’m going to continue working on it, though I am serious considering reverting back to javazoom. I’d rather have well-written code using a slightly out-of-date library than mediocre stuff written with more modern equipment. 

I did work more on the encryption project, finally adding Scrambling in with the rest of the methods. I managed to combine the two from my testing class into one, using nested for loops to 1. select the individual word of the input, and 2. run the scrambling technique on it. 

I also fixed the main a bit, adding a loop and doing some basic input-checking. Nothing to write home about, but it made the output look a little nicer. 

Now all I need to do is put it in a terminal window. Then it could be in a tv show about “hacking”. 

The efficiency is not exactly what I’d like it to be. Having those nested for loops means I’m running through every single bit of information -  O( n2 ). In the context in which I learned efficiency - primarily searching -  a O of that would be terrible (i.e. Selection Sort.) But is that still true in this case? Thinking about it, hitting every node when trying to scramble a word all but guarantees the original word won’t remain when it’s done. 

But despite that, the original problem still remains - as n increases, the runtime of such code will as well - in theory punishing the user for wanting to encrypt a longer message. 

I’ve been doing some research on this and some java OO design principles. I’m still looking to add on to this project, but at the moment I’m not sure what direction that will be in. Restructuring everything into classes? But how to implement that? Perhaps start adding more UI elements? 

After the fun that was Windows 10, I finally got my Eclipse workspace up and running and decided to continue work on my Encoding project. I’ve put the Public/Private key stuff to rest for now - the theory still fascinates me, but I haven’t quite figured out how to implement it yet. 

Instead, I decided to work off of my Transposition method - but rather than simply reversing the input, it scrambles every word. It’s a pretty common mode of encryption, albeit generally rather simple to decrypt, and can be implemented with or without using spaces. (I kept the spacing, though am planning on making it an option later.)

Formatting the input was rather simple, but I ran into some problems when it came to actually scrambling the string(s) efficiently. My first instinct (remnants of my introduction to programming a few years ago) was to create a bunch of variables and basically hard code the “randomness” into the shuffle. I know enough not to go with this idea, but coming up with something a lot better is still a skill I’m developing. 

After thinking of a few options and following up with some hunches, I stumbled upon the Fisher-Yates shuffle. It’s pretty simple, looping through a character array and using a RNG to grab a character at random and exchange it with another. 

All in all, it looks like this:

And all with actual comments!? I should frame this!

As I have been the past thirteen days, I opened up Eclipse this morning and chose the workspace that hosted my Encoding project, wanting to spend a bit of time on that. Error. 

Wait, what? 

I tried again, and once more just to make sure I wasn’t doing something crazy. Then I checked the folder itself - it was set to read only. Okay, not hard to deal with. I unchecked the box, waited for it to process, and opened the workspace again. 

Same error

Long story abridged, apparently there is a bug (feature) in Windows 10′s security settings that sets all of your folders to read only, and that can’t be undone simply by unchecking the box. So, after hours of googling the issue (no help) and poking around in the settings, I starting looking into the permissions of users on my laptop. Changed a couple of them that didn’t seem like they would affect anything and voila - everything magically works. 

Ironically in saving this picture, I realized I needed to do the same for my Pictures folder. Good job.

This issue aside, I’ve actually quite enjoyed Windows 10 thus far. The interface is simple and (once customized) really easy to use. I’m still getting used to finding all of the settings/control panel options, but that’s with any new system. 

Plus with the personalization options, I can make it look like this. 

After many hours of caffeinated research (and a little help from a wonderful person) the mystery of the Shickaxe has finally been solved. We ended up creating an entire new template class that extended ItemTool and customized it to break the blocks we wanted (namely those that are normally broken by pickaxes and shovels.) 

On the plus side, this gives us a lot of flexibility into extending this class to make new items in the future - such as including the properties of an axe.

But, as with much of Minecraft’s code, figuring out how to do this was a pain, which does not bode well for the other things I want to begin including in this mod.

What you don’t see is that these lines go on horizontally for awhile...

The second part of this was creating the ruby shickaxe. Extending the template class and putting it through the game registry is second nature at this point, but I once again ran into issues with the JSON generator. You’d think after two months I’d know how to get it right the first time, but eh, details. 

This was the final design - 

Plus a custom crafting recipe - I’m on fire!

I’m still finishing up the ruby items, but with this success, I’m considering branching out into other joint tools - a shovel/hoe combination perhaps? 

My next mod project will also take a bit of research to make possible - a toaster. It’s simple, but I like toasted things in real life, and I like the idea of a mini-oven that would take fewer resources than an oven, but would still provide more hunger bars than raw. 

AlexaWeb: 

I started Thursday continuing work on my website, really wanting to clean up the mess I left the previous days. The top section looked pretty good, so I turned my attention to the “About Me” section and began to play with the layout - and spent half an hour re-learning something important about CSS: margin and padding are not the same thing. 

Repeat: margin and padding are NOT the same thing. 

I really should get this as a tattoo... 

I was working on my little picture next to the text, trying to float it to the left and leave some space between the left margin and it. Cue half an hour of wondering why my code wasn’t working. Eventually (though googling) I figured out what was up, but not before this gem: 

If anything represents programming as a whole, it’s this: missing one stupid quotation mark. 

But after this little hiccup, things started looking up. Everything started fitting in to place, the CSS started looking less like spaghetti and more like semi-connected threads, and I actually made more progress than I expected. Woot!

It’s not perfect, but it’s getting there. 

Bonus: The picture behind Projects is one I took in our computer science research room during an event last spring. Yes, the wires usually look like that. 

It still needs work, but I’m rather pleased with the results thus far.  I’ve more-or-less finished the About Me section and have started working on Projects, meaning I’ll grab a few screenshots and quotes from this blog soon. Progress!

To preface, I am very much a test-driven developer. My SOP is to write a few lines (or occasionally a whole method) then run it to make sure it’s actually doing what I want it to. While it occasionally requires adding a ton of random print line that I need to get rid of, it’s saved me hours of redoing work. 

Unfortunately both web design and Minecraft modding take this idea and not only throw it out the window, but run outside to stomp on it a few times too. To say they don’t like it is an understatement.

Day Seven: 

I wanted to take a little break from the modding to work on one of my more serious projects, so I started working on my website again. Having made a rather crappy one last year, this project has had a goal of good design practices, namely having the site be mobile and browser friendly. 

Part of that has already been accomplished. Thanks to some images from a friend and a little Javascript, I’ve made my menu change based on the size of the screen - get to a certain point and the standard top word- based menu turns into a nice little clickable menu in the top left. 

Chrome has a really good DevTools option, including a button to see what it looks like on mobile. 

But the scaling that went so well in the menu has been proving a problem with other parts of the site, namely pictures. Scaling them has proven to be a pain in the ass, especially since I rarely can add a few lines, test, few more lines, test, etc. as I normally would. This isn’t as bad as developing Android Apps, but it’s up there. 

So that’s the limbo I am in for now. I don’t want to create a shoddily constructed site that I’ll have to go in and rip out the guts in a few months. Nope, I’m going to go slow and steady and make something both beautiful and functional. But really, emphasis on the slow part. 

Day Eight

Wanting a break from web design, I went back to (you guessed it) Minecraft. Originally I had planned to add in the things I mentioned last time, namely the ruby tools. So I went about creating all of the images for them and then started to work on the code. 

The first (a sword) went by rather easily. Code is pretty simple and took about half an hour to put it together, including fixing my little mistake:

Definitely the correct recipe for a sword. Yup. Totally. 

But that’s where my good fortune ended. Rather than making the standard shovel and pickaxe, I had the idea to make a combined “shickaxe”, one tool that did the jobs of both. Unfortunately since Minecraft’s code is a twisted plate of spaghetti, that was easier said than done. 

In total I spent about two hours today trying different ways to make it work. I found a canHarvestBlock recipe that looks promising, but making anything out of that is proving to be difficult, even after I broke and tried to find help online. (All of the sites I found were a few years old - does anyone know anything a bit more up-to-date I can look at?) 

And again no simple way to test, so bugs galore. 

Gahhhhhhhhh!

Rubies!

I know this seems rather basic, particularly since there are mods that add dozens of new ores to the game, but from this one cool-looking block (originally called something else,) I came up with the idea to challenge myself to actually make an ore. So from here I created the ore block:

Then it was a matter of what happens with ores in-game. After looking up some example code, I figured out how to make the ore block drop a little ruby (it may be just one at the moment, but I’m considering adding a random number generator to drop more.) 

Then came the most difficult part - learning how to make the ore spawn naturally. Through a bit of trial and error (see the picture below,) I found a good balance of rare but not too rare and had it successfully generate in my test world. 

This was perhaps a little too common. 

I was working on the next obvious step, having the rubies dropped from the ore blocks be able to craft into the ruby block, but there’s still an error in the code I’m trying to figure out. I did have a fun error come out of it though. In trying out different ideas, I accidentally set it so an empty crafting table was the recipe for 9 rubies - oops. 

Yes, give me all the free gems!

From here I have four goals for the mod: 

1. Fix the recipe for rubies becoming ruby blocks

2. Create ruby tools (sword, pickaxe, shovel, hoe, axe)

3. Create ruby armour 

4. Repeat for other interesting ores. Amethyst? Onyx? Sapphire?

It feels good to get back into this, and I’m really looking forward to see what comes of the next three weeks. (Though at some point, I should work on something besides Minecraft.) 

TacoPick: 

I’ve been thinking more of what to include in my mod and have come to the brilliant idea of creating tools out of food. I haven’t quite decided what the drawback will be (in terms of durability, blocks that can be broken, etc.) but it has been amusing explaining why I’m making a pickaxe look like a taco. 

On the positives, the item does work in-game and has a shaped crafting recipe (my first one!) Like any in-game pickaxe, it takes two sticks topped with three items, in this case tacos. Figuring out the shaped recipe took a bit of time, but it was rather cool to finally see it working. (Code below)

But that is where the good news comes to an end. In order to add the custom texture to the game, you need to generate a JSON file. Luckily there is a generator that will create one for you, but it is notoriously finicky and often does not like to work if everything isn’t exactly right in the code. When that happens, there’s basically a checklist of things to look at to try and find the error. And to put it lightly, I really do not have the patience for that right now. 

while tiny, the pick is rather pretty

______________________________________

On a tangential note, I did find an interesting article called 10 Books Every Programmer Should Read. I had heard of a couple of the books before, but I spent a bit of time finding copies of some of the more interesting ones - notably Coders at Work, Clean Code, and The Mythical Man Month. I’ve found that while I don’t often have a lot of time to break out the laptop and get to work, I do have chances to grab my tablet and read a bit. So why not make that time productive too? 

But in the times between me goofing off, I did manage to squeak out some work, namely on my small Encryption program. Thus far I’ve added a Caesar encryption that is hard coded to shift two to the right (I’m still trying to find a dynamic way to do this - any ideas?) and a Transposition cipher that reverses the input. I did clean up the “UI” a little bit (it’s kind of a menu) and created the framework for future additions, and standardized my input method. 

The research part came in when I decided to add Public Key Cryptography (PKC)  as an option. Having first heard about this years ago from Cory Doctorow’s book Little Brother, I’ve since found the concept fascinating and wanted to try and include it here. 

The theory is very simple. Two people (Bob and Alice) want to communicate privately. With PKC each person has a private key only they know and a public key that is broadcast to the world. So if Bob wants to send Alice a message, he encrypts it with her public key and only her private key can unlock it, and vice versa. And if you want to guarantee authenticity, Bob could encrypt it with his private key (that only he has access to) and Alice could decrypt it with his public key - a signature of sorts.

The complication comes with the keys themselves - they are generated mathematically and the “how” of that process is rather difficult to understand. (At least, I’m still trying to make sense of it.) At the moment I’m considering just showing the idea of PKC using passwords, but I would really like to make a small but actually functioning version. 

AlexaWeb is my attempt at a personal website, an online resume of sorts. Between working with various web tools (between yesterday and today I’ve used HTML, CSS, and Javascript thus far), designing something that actually looks aesthetically pleasing, and working in an environment that doesn’t quite match up to my test-driven habits, it’s been a rather interesting challenge. 

So here’s what I’ve created thus far:

Some of the code: 

A friend introduced me to GitHub’s IDE, Atom after seeing me work on my traditional Notepad++, so part of the past couple days I’ve also been setting this up and getting used to it. As I’ve only been working with simple files, I don’t have much to say at the moment, but I am interested in seeing how it’ll handle other, more complex projects. 

As for the look of the website, I know it’s rather plain and more than a little rough. I spent some time designing an outline of what I want the finished site to look like, but making that vision a reality is somewhat difficult when there are almost an unlimited number of things you can do in CSS. As a programmer I’m a very test-driven person, so having to make a number of changes before each test (particularly when tests are just visual) is maddeningly frustrating some times.

 (Very) Rough Draft:

But overall I’m rather happy with the progress I’ve made. The menu looks rather nice, particularly since it dynamically shrinks when viewed on mobile, and I’m happy with the designed layout that I hopefully will fully implement by the end of the summer. 

To be fair, I was super busy today and didn’t really have a chance to sit down and look at anything. But I did get distracted by video games after work, so I’m not entirely blameless for the lack of progress. Oops. That said, I did complete a full hour tonight all spent working on my “fluff” project - yup, Minecraft. 

In actuality, modding with Forge has been quite interesting. The source code can be confusing as hell, but the challenge it presents has been helping me get used to larger products that aren’t obvious at a quick glance. And while I did use a tutorial of sorts for my first custom food (a taco), the rest was done myself with a pinch of help at the end to make my custom sword even more entertaining. All and all, here’s what was accomplished: 

Custom Food + Custom Recipe (that contains another custom food!)

To give some context, there is an inside joke among my friends and I about tacos in general, so I figured they’d be the perfect first item to put in my mod. Plus tacos are delicious - who wouldn’t want them in Minecraft? 

In this case, I created a custom recipe so one wheat would produce two of my custom flour item, which in turn (when combined with a raw chicken) would make three tacos. They fill just shy of 8 “hearts” of hunger and can stack up to 64. 

Custom Sword

So with my new foodstuffs, I took the next logical step - make a weapon out of them! Introducing the Taco Sword which, thanks to my amazing image editing skills, looks absolutely nothing like a taco. Presently it has the same stats as a standard diamond sword, but that may change in the near future. But the coolest part is what happens when you attack a mob with them - with each hit, the mob drops tacos! It’s silly and ridiculous and absolutely the perfect addition to this mod. 

(Just for fun, here’s a quick shot of some of the code. I kept forgetting to render my items with the game registry which leads to very frustrating errors.)