Polyphasic Programmers

Recently, I’ve been toying around with a little web framework called Zappa. Well, I guess you could call building two fully-functional websites, one of which is running for a live client with decent traffic “toying”. I must say, I’m very, very impressed by what I’ve seen so far.

For those of you who haven’t heard of it yet, node.js is a recent serverside framework that uses Google’s V8 Javascript engine serverside. That’s right, you run JavaScript on your server, and it serves up your html pages and can do low-level sockets. If that sounds a little backwards, it kind of is at first. Node.js itself is very much a low-level framework, essentially just an asynchronous sockets library and the standard javascript functions, so trying to do much of anything using just node.js is a pain.

Enter Express. This wonderful little framework imitates Sinatra (and several other web frameworks) and turns the process of defining routes, handling html templates, and setting up a proper MVC into childsplay. On top of this, we run Zappa, which takes everything that makes Express great, and lets you code it in Coffeescript instead of straight JavaScript. Suddenly, you’re using a super-clean language that’s just a joy to use.

This leads to something interesting. Zappa ships with CoffeeKup as the templating engine, so you write your HTML templates, serverside logic, and clientside behavior all in CoffeeScript. No more funky data formats and weird convention differences– your client and server can literally share libraries and source code if you so desire. That alone speeds up the process of development tremendously. On any given day, I touch the database, controllers, several views, and the CSS stylesheets. The fewer times I need to re-adjust my slow, human brain to parse a different language syntax, the happier I am. (The only part of my codebase that isn’t in CoffeeScript right now is the CSS, because Stylus is cleaner, and CSS is sort of weird to begin with.)

This is not without some oddities of course. node.js is primarily an asynchronous framework– this means that all of the library calls pause execution while waiting for IO, and then resume that execution later. Rather than using functions that, for example, return data from a database, you instead pass a callback function as an argument, and deal with the data in the callback whenever IO gets around to finishing. For anyone who’s used to a traditional procedural language, this is a royal pain to adjust to. The main benefit of course is performance (because node.js can serve other pages while it’s waiting for your IO to complete) but it’s a rather unconventional way to code coming from a background in PHP. CoffeeScript does make anonymous functions and callbacks very easy, which helps a lot.

The other major benefit is performance. The way node.js works, when you start up your server, it parses and JITs your code (V8, Google, totally awesome) and leaves the instance running in memory. Because node.js can queue up requests and handle everything asynchronously, it doesn’t need to spawn a new processing environment for each pageload, but does it all using the instance that’s already
there. That means response times are quick and memory usage is minimal. It does mean that all of your code is always loaded, but that’s hardly an issue for the vast majority of sites. I find that, running no less than 4 node.js instances during development, the small PHP site that my server runs still uses 4x the RAM as all of the node instances combined, and that’s running through lighttpd with some rather restrictive process spawning limitations. I’m not sure how well node.js can balance high traffic, and I’ve yet to try load balancing, but I’m hoping this new project gives me an opportunity to do both.

Yesterday, I started doing gruntwork on a new project in Zappa, one that I’ll reveal later this month as it starts to come together. In the course of a single day, I was able to put together a fancy ajax navigation system, a real-time chat window, and a forums system. All that in a one day using Zappa! In any other framework that might have taken me several days. It is some seriously rapid development. The community is already strong around nodejs as well, even though it’s so young and not widely adopted. I’m using Sequelize to make MySQL work a breeze, Cradle for CouchDB when I’m feeling adventerous, node-validator to sanitize inputs, and several other odds and ends that make things breezy.

Would I recommend node.js? Absolutely. Is it the best language for any job? Of course not. No language is. It’s certainly capable of being a general-purpose server, especially with its low level socket APIs, but right now the support isn’t quite there for everything. As an HTTP server it totally rocks. If you’re willing to take the time to learn it, I think it’s well worth the effort.

I decided to start a minecraft challenge map called SkyBlock 1.1, which can be found at the other end of this elegant and finely crafted link. I’ll be playing this in Normal mode at all times. That’s not a big issue at the beginning, because the map isn’t exactly large:

That's a tiny island!

That tree is *precious* at this point, as it’s my only source of renewable anything. Oh, before I forget, this map comes with a set of challenge objectives:

Challenges:
1.Build a Cobble Stone generator.
2.Build a house.
3.Expand the island.
4.Make a melon farm.
5.Make a pumpkin farm.
6.Make a reed farm.
7.Make a wheat farm.
8.Make a giant red mushroom.
9.Build a bed.
10.Make 40 stone brick’s.
11.Make atleast 20 torches.
12.Make an infinite water source.
13.Build a furnace.
14.Make a small lake.
15.Make a platform 24 blocks away from the island, for mobs to spawn.
16.Make 10 cactus green dye.
17.Make 10 mushroom stew.
18.Make 10 Jack ‘o’ lanterns.
19.Build 10 bookcases.
20.Make 10 bread.

These are going to be interesting. I lack iron or any ores, so I’m stuck with wood and stone tools. I also cannot make pistons or redstone mechanisms of any kind. The pumpkin worries me the most, as I’m still running 1.8. The only way to get pumpkin seeds is to break the stalk, which has a chance to drop 0 seeds. That’d kill the run right there, so it looks like I’m stuck with just one pumpkin plant for now.

I started out by flattening out the first 2 levels of the island. In the process, I discovered 3 sand blocks buried inside the floating island. I chopped down the tree, picked up the 1 sappling that fell from the leaves (just one? Really? -_-) and looked in the provided chest to see the following:

A Chest!

The first thing I build is a bed, so that I can set the night back to day at will. I use the 12 string provided for the wool. This is not necessary, but it makes my nights less worthless at the start because I have no torches to grow trees. That bonemeal is an option, but until I have a steady, safe supply of skeletons, it’s precious like everything else in that chest.

Challenge 9: build a bed, complete.

My attack plan at this point is to get a tree farm running pronto. This is slow going, but it’s steady progress and I’m collecting a lot of wood. (I watch quite a few youtube videos while this is going on.)

It speeds up considerably once more sapplings drop and I can grow more than one tree at a time. I split the logs into planks and expand the island outwards. I fill in the bottom of the island to create a dirt square, and then expand the island 8 blocks in every direction. Finally, some room to move! This is still less than 24 blocks in width, so no mobs will spawn. (I have a plan for those guys later.)

Tree farm + Youtube for hours = Platform ^_^

Challenge 2: Expand the Island, complete.

I built an infinite water source using the two ice blocks, and a cobblestone generator using the provided lava bucket. This gives me infinite cobblestone, the first 8 of which go directly into a furnace. Hooray progress!

Water Pool

Cobblestone Generator. Semi-automatic too, using just one water source block. I replaced the dirt with cobblestone after I took this screenshot.

Hooray! A furnace finally!

Completed:
Challenge 1: Build a Cobblestone Generator
Challenge 12. Make an infinite water source. 
Challenge 13. Build a furnace. 

Next up on my list is a series of farms. I decide to tackle the pumpkin, watermelon, and reed farms all in a small area. The pumpkin is especially slow because I can’t harvest seeds, but I eventually get 3 melon plants going, and I can refill my food bar again. That’s a relief– I don’t like running around with half a heart for that long. While I’m at it, I set up the 3 sand blocks and get cactus growing. I don’t see much point in making an automatic cactus farm for just 3 plants, so I harvest them manually. Once I’ve got 10 cactus, I use the furnace to make my 10 green dye. So far so good. ^_^

It's so lonely...

Completed:
4.Make a melon farm.
5.Make a pumpkin farm.
6.Make a reed farm.
16.Make 10 cactus green dye. 

I’ll eventually have enough reeds to make the 10 bookcases, but that’ll take a fair bit longer to grow. I just have to remember to collect every once in a while as I work. While I’ve got the furnaces going, I cook up a large amount of the logs and finally have coal with which to make torches. Huzzah! Now my place is lit up at night, and it’s safe to venture forth outside the 24 block radius without fear of wandering creepers.

My tree farm still isn’t particularly fast, and there’s a limit to its speed because I only have so much dirt to work with. I decide to use the cobblestone generator to give myself some more building materials, and then get to work on building a mob pad. I use a design by ethoslab which is quite effective, and which also happens to avoid spawning endermen, which would cause me all sorts of trouble. I build this 27 blocks in the air, with the bottom block at height 92. Mobs are directed to an opening in the bottom and fall to their deaths, leaving a nice pile of goodies for me to collect.

This gives me an infinite supply of sulfur, string, arrows, bones, and rotten flesh:

That was a *lot* of cobblestone. Those generators are not fast.

The inside has water channels that push the mobs to a central hole for processing.

Rotten Flesh! YES!

By the time I finished building this mob tower, I had collected way more than enough reeds to make the 10 bookcases. (I was running to my farms in and out while grinding cobblestone.) Since I’m going to need lots of mushrooms, I take the leftover wood and build a simple mushroom farm. Just a semi-dark room really.

I missed a golden opportunity to conceal the door behind one of the shelves. Alas!

Completed:
11.Make atleast 20 torches.
15.Make a platform 24 blocks away from the island, for mobs to spawn. 
19.Build 10 bookcases. 

I use my newfound source of bonemeal on the grass and create seeds, allowing me to create the wheat farm. Hooray! This also gives me a limitless supply of roses and dandelions, should I find a use for them. I’ve since expanded my melon farm a bit, and my poor pumpkin is still lonely. I’m up to 6 pumpkins at this point, not quite enough. My mushrooms are also taking *ages* to spread. Ah well.

Bread! Delicious.

Completed:
 7.Make a wheat farm.

The major challenges that remain are the House and the Lake, which will expand the island even more. I decide to work on the Lake first, as I want a lakefront house. I’m not sure how extravagant this is going to get. I’m rolling over in string, so I also start collecting wool and various dyes for a side project.

I get started on the lake first. I figure, with limited time, resources, and patience, it’ll really end up being a small pool, but it’s a lake in spirit. While I’m at it, my brown and red mushrooms finally spread. Yay! I grow a brown mushroom on the corner, then I get the brilliant idea to kill two birds with one stone: A giant mushroom house. This is the end result:

A small pond with big dreams

I spent ages getting the red mushrooms to all spawn at their smallest height, and ended up with nearly 2 stacks of them by the end of it.

The interior's a bit... lacking. Ah well.

Completed:
 2.Build a house.
8.Make a giant red mushroom.
14.Make a small lake.

I’m finally nearing the end of this challenge! The only things remaining are number-based puzzles, and I quickly get to work on those. I’m up to 8 pumpkins so far, so I should have those done by the time I finish collecting everything else. A couple of hours later, I stash the goods in a chest in my new house:

The Pumpkins took nearly the entire challenge to grow!

Completed:
10.Make 40 stone brick’s.
11.Make atleast 20 torches.
17.Make 10 mushroom stew.
18.Make 10 Jack ‘o’ lanterns.
20.Make 10 bread.

All in all, this was a fun challenge map. ^_^ For grins, this was my final inventory in the other chest, with all the items I collected during the entire challenge:

Mob spawner -> Bonemeal -> Roses -> Red Dye + Bonemeal -> Pink Dye + (String -> Wool) -> Pink Wool!

The mob spawner didn’t seem to be very good at killing spiders (they kept climbing up the drop tower, which I might have fixed if it were made of glass, as I don’t think spiders can climb transparent blocks) so the string production was low. I might have collected enough wool to make some fine art, but alas! Maybe another day.