We made some good progress this week. Most notably, I am nearly finished with planning out the next several months of development. My pages of notes are now laid out in a rough schedule which I’m currently translating to Crea’s project plan.

As for other progress this week, I managed to fix a huge rendering issue that has plagued me for quite some time. When anything would be slightly transparent it would get a white tint added to it. This was caused by how the lighting was rendered. The main reason for fixing this finally was because I wanted to add in a trail effect when you dash.

Notice the trail effect when he dashes backwards from the monster

Another small feature we got in this week is the ability to do colored text in our chat. This was done by Matt, the new developer, to help get his feet wet.

This next week I am finally finishing up combat changes and cleaning various things up. Matt will be looking into getting multiplayer to a more stable place. Kelley is working on the new skill animations. We’ll have a lot to show off next Monday!

The next beta update will be ready soon. It will include a major overhaul on the skills, with many new skills being added. There will also be improvements to the underground biome.

Once that update is released and proven to be stable, I am thinking I will finally release Crea beta phase #3 and start our Greenlight campaign! Beta phase #3 will include the first boss and a few new features, such as a type of magic called “remna.” Then we will finally see the addition of NPCs! Lots of exciting stuff coming up. Just have to get over this development hump.

The beginning of last week was rough. Being a programmer at heart, balancing is often hard for me to do because there is no obvious right answer. My goal has been to improve combat; however, this is much trickier than I would like.

There are two major constraints that are making this much more challenging than it should be. First is that the terrain is randomly generated and players can change it around. Secondly, adding animations to our characters can be troublesome. Despite these constraints, combat is starting to come around. Some things I have done so far:

• Refine timing of some monster attacks
• Oil slime second attack
• Sand Runner (now called Ortomi) second attack
• Scrapped shields and dashing
• Brought back dashing with double direction key press
• Started working on new monster spawning system dubbed “Conflict”

The Oil Slime’s new ranged attack

Beyond combat, modding API reference documentation has finally began to take shape. Even though it still has a long ways to go it still is looking good. Here is a sample of what it looks like.

Right now nearly all functions look like this Player.canEquip() with no named arguments and no description text. I would like to eventually have all functions documented (or at least all of the frequently used) like Player.loadInfo() with nice arg names and docstrings. This documentation will also provide how-to mod tutorials. However, this work that can quickly become outdated, so I will cautiously work on it and may seek outside help. If you want to work with me on writing documentation, let me know. Once complete, this will undoubtedly prove to be an invaluable resource to modders.

Lastly, I finally convinced myself to plan the next few months of development which involves several features and much content. For most of Crea’s development I just have been designing as I go, but lately I have found myself getting slowed down by having to context switch too frequently. I am just about done planning things and am quite excited with the results which I will divulge later. Now I just have to implement them!

This next week I am hoping to wrap up these combat changes and the conflict system, which will probably take up the majority of the week. I am anxious to get a new beta build out so I will be working towards getting things cleaned up and ready for a release.

I have been wanting to get back to posting the recaps on Monday so I’m skipping this week and will post a full recap on Monday. I’m currently waist deep in planning and designing upcoming updates. I’ve had my head in the code for far too long and started to lose sight of what exactly needed to be done and how it should be done. See you on Monday!

Phew. The end is in sight for this mod support feature. I knew it was going to take awhile to get through all of the changes I wanted to make. Originally I planned to do it in stages while also doing more visible gameplay changes but shortly after starting I realized that would not be the most optimal.

The majority of these changes were quite intrusive and had the potential to completely break the game, which means I would need to go through and thoroughly test everything for each release. I decided to do this thorough testing after all of the changes have been made. I’ve already encountered some fun bugs like this one (lighting issue around world seam).

Death awaits

To expand on this “mod support” feature, it really encapsulates many small to large changes related to providing a better experience for modders. There have been some great enhancements made such working on support to disable/enable mods on a per world basis without having to restart Crea (UI is not implemented yet). Also I have been reading through the Steamworks API and laying the foundation to be able to easily add support for Steam Workshop. However, most changes have revolved around reworking and simplifying the interface modders will be working with.

A great example of simplifying the modding interface is what I have spent almost all of my time on last week, which is rewriting all of the content scripts. I was surprised how much we already have – 290! This includes monsters, items, plants, tiles, etc. This took several days and much effort/boredom but I am quite pleased with the end result. As an example, here is what the old Greatsword content script looked like versus the new one.

Old Version

name = "Greatsword"
render = Render('mods/base/item/weapons/swords/greatsword.png')
add(render)

item = Item(delay=600, unique=True, use=Substitution('sword', render.image))
item.setUseAnimation("sword", "arm_front")
add(item)

craft = Craft(category='Weapons', subcategory='Swords', level=10, experience=25, service="Anvil")
craft.material('mods/base/item/weapons/swords/broadsword_plus.ce', 1)
craft.material('mods/base/item/recipes/smelting/gold_ingot.ce', 5)
craft.material('mods/base/item/recipes/smelting/iron_ingot.ce', 5)
craft.onCraft = upgradeEquipment
add(craft)

equipment = Equipment('active')
equipment.onCreate = createEquipment(0, 0, [ChangeStatAttribute.create('knockback', (2, 3))])
add(equipment)

add(Weapon("Sword", 65, AttackType.Melee, DamageType.Physical, useMeleeWeapon))

New Version

from core.template.item import Material, Sword, StatAttribute

Sword(
    name = "Greatsword",
    power = 65,
    level = 10,
    experience = 25,
    serviceRequired = "Anvil",
    upgradeFrom = "broadsword_plus",
    materials = [
        Material('gold_ingot', quantity=5),
        Material('iron_ingot', quantity=5),
    ],
    attributes = [StatAttribute('knockback', valueRange=(2, 3))],
)

There are two new aspects here that greatly change things. The first is that there are no imports at the top of each script so you know exactly where everything is coming from. The other is that there are now content templates. Here you see a Sword template that is specialized for creating swords. Templates help setup the boiler plate code and try to provide a more readable interface.

Over this next week I am going to be working on getting Crea completely back up and running as well as fixing up high priority bugs. After that I will be working on getting combat to a better state and also continuing to get multiplayer truly functional. We’ve just about made it through this technical excursion!

Another week already!?! Since quitting my job I have completely lost track of time. Everyday is the same – work on Crea. Well, that is not completely true…

Last week Kelley and I got to met up with the wonderful guys at bigsushi.fm. We did an interview with them that will be going up on their podcast in the future (around Crea’s Greenlight launch). If you have not listened to them before then you should! Lately they have been hitting up dozens of indie game devs and getting a unique perspective on the life of being an indie. Some of my favorites consist of them talking to Handy Man Studios – Jake (Edge of Space) and Robot Loves Kitty (Legend of Dungeon) but they’re all good! (HINT: You should listen to them)

While there we joined them for a game night which consisted of eating delicious pizza and playing Ouya games. Good times were had. Kelley kicked some butt on Towerfall but unfortunately that is not featured in the videos that you can find by going to this bigsushi post. If you enjoy that then you should subscribe to bigsushi’s twitch and join future game nights. I know I will be!

Progress on Crea has been good. I have been continuing to focus on engine work (mod support). My goal is to try to get the engine to a stable place. For a long time now I’ve been trying to balance between doing engine work, features and content. This is extremely tricky since anytime an engine change is made there is the potential for anything break. Additionally, engine changes often result in needing new worlds and/or characters which I know we all don’t enjoy. And on top of all of that, engine interface changes will break mods. In other words, stable engine equals good.

Engine changes are like removing a card from bottom of a tower of cards

The majority of the week was spent doing a MASSIVE refactoring of the engine to support realms. Previously there was an assumption made that there was one virtual physical space. Wrong! If a torch is placed at position (50, 50) then a light source would be emitted at that position. With the world realm (the main world area) and another realm, lets say a dungeon, there are suddenly two positions known as (50, 50) but we only want to display the light source in the realm the torch is in and not the other one. If not handled correctly a light source could appear in both.

Where is that light coming from? Bug Reporting Time!

This is no good. Also this is not only the case for light sources but for all spatial systems such as physics, dropped items, monster AI, and so on. You don’t want an invisible oil slime from another realm attacking you, do you? …I didn’t think so.

When I had finished adding realm support in I took a diff between where I started and where I finished. It was an astounding 25,000 lines long! To reiterate, this is a MASSIVE change. To give some perspective, the entire engine is only about 40k lines of code. In other words, I touched and changed about half of the entire code base that I’ve been working on for a year in a week span. Best of all, it works!

This next week should be the last week of engine work. Afterwards the plan is to fix bugs and finally get combat and monsters to a good place so we can add some more monsters we planned.

The bug fixing continues! More than two dozen bugs have been fixed this week. We also managed to make some impactful gameplay changes: Researching was overhauled, blocks now can go on both the ground and wall layers, the tools (mattocks) were redone, and we implemented some other feedback from our testers. Oh, and multiplayer is supposedly working now!

It was all too common to drop a pile of materials into your research window, click “Research” and go on your merry way. This is not how I wanted researching to function. I wanted it to be more hands on – you’re engaged and making decisions. Consequently, researching needed some heavy modification.

Now in order to research items you must first craft a “Research Desk” and be nearby it to research. When researching a material you must select the item category to focus on. For each research you can now learn multiple recipes instead of always one. The number of recipes you learn is based off of your craft level and the level of the item(s) you will learn.

Researching is feeling much better than before! One remaining change we want to make is to avoid having a single item used in too many recipes. As it stands lumber is used in 56 recipes! That is way too many. In the future we will be limiting a single material to be used in at most 20-25 recipes.

Another massive change is that now blocks can be placed on either the ground layer or wall layer. Before we had “Dirt” and “Dirt Wall”. Now it is just “Dirt”. There are a few major benefits to this approach.

• The first and foremost is that we have nearly twice as many blocks in the game – meaning more customization.
• This saves on inventory space.
• Now all wall tiles are consistently shaded and more easily distinguishable from the ground layer.
• We can now place ores and other rare resources on the wall that will make for more interesting world generation.

Complimenting this change to blocks, we did a complete makeover of our mattocks (tools) to be more in line with our equipment. Now mattocks have equipment attributes and are upgraded from one to the next. Also now all blocks can be destroyed by any mattock. I believe this will have a very positive impact on gameplay. Previously we needed the rigid block restriction rules to gate players; however, this is no longer the case. Instead we have the nice side-effect of gating progress by having most items upgrade from earlier versions.

Anyone got any lumber?

Even though I don’t always have the time to reply to all our beta players feedback, I’m reading it! Here are a few things that you can see that already made it in the game that was player feedback.

• Allow placing items directly from the inventory, without having to put them in the action bar first
• Clicking on a material in the craft UI should take you to that material’s recipe
• Make it possible to remove wall blocks from behind the player, trees and other things.
  
• Better world generation with more caves and ores

Finally, after much toil, I have gotten multiplayer to a working state. I am sure it still needs much love before I would consider it complete. Why did multiplayer take so long and what was wrong with it? Sooo many things… I had solely tested multiplayer on my local connection and things worked perfectly fine; however, once in the real world it immediately broke down. I investigated for days and continued to bug the good people on the IRC for assistance in testing and then yesterday I finally discovered the cause for all my troubles.

One feature I will be working on soon is a dedicated server. Imagine how nice it will be to have dedicated server software for a game for Linux!

Over the last two weeks since we launched beta, I have been working hard to get Crea to a good point with all features to be working well. I imagine there will be several more bugs for multiplayer to come in, but now it is finally time to turn my focus to do some of this work to get official modding support in. I will likely be going back and forth between this and bug fixes/small enhancements over the next few weeks.