Last week was a little more productive again, and I was able to fix a number of issues with the buildings and screen settings ahead of the experimental release.
As you can see from today’s picture, I was able to put the last of the buildings into the game, and I went through some more extensive testing on them. Based on that testing, I realised that the optimisation issues I had already noted a few weeks ago were more of an issue than I originally thought, and so I changed a few aspects of the way the buildings render at a distance to improve their performance. This isnt the full performance optimisation I am planning for as part of 0.9, but it should be sufficient to tide us over for the time being.
Another small change was in the way the buildings interact with the new AI navigation system - the previous way was a little easier to set up on my end, but gave rather inconsistent results depending on the placement of the buildings - so I switched my approach, and the buildings are consistently avoided as obstacles by our villagers now.
A final small change I want to talk about is based on John having recently bought himself a widescreen monitor: We’ve had a lot of complaints about issues on widescreen monitors, which I was unable to replicate when setting Unity to the corresponding resolutions. However, being able to see the issues “for myself” through screensharing with John has given me some new ideas for what may be wrong, and I have started to change a few UI settings in order to hopefully fix the issues we’ve been seeing.
At this point, I am planning on releasing 0.8 onto the experimental branch by Saturday (the 23rd). I am sure there will still be a lot of bugs at that point, and there are a few more features I want to include before the full release - but there are so many improvements over the currently available version with a mostly rounded experience that I think we need to start getting some feedback from the community at this point. I will release a post with instructions on how to access the experimental branch and all of the changes when it is ready on Saturday.
Until next week!
Weekly Update 116 - Intermission
Hi everyone,
In addition to being busy last week, I ended up also falling rather ill, so I have not managed to get much done last week, and I will just keep this short.
In good news, however, I am mostly recovered at this point, and aside from some catching up on other work I couldn’t get done last week, I should have a decent amount of time to work on the game this week. So, we are still on for a 0.8 experimental release before Christmas.
Until next week!
Weekly Update 115 - Plant Pots
Hi everyone,
As mentioned in the last post, last week and next will be pretty busy for me, so my apologies for getting this post out a little late and keeping it quite short.
Last week, I managed to add some foliage to the new building models, which also needed some kitbashing to start with. We’ve had various pots and potted plants in the game for a while, but I had to get them into a clean file to work with, clean up their materials, etc, and added some new ones I could add as well. The potted plants in today’s image are an example of that, and there are quite a few more (such as vines) that aren’t pictured.
I will try to get some more work done this week, and should be back to having more time after the weekend.
Until next week!
Weekly Update 114 - Metal Barracks
Hi everyone,
I finally finished the structure of the metal barracks last week, so today is a quick overview of the building, and what still is to be done on it.
As you can see from the image above, the metal barracks are a very substantial building, with three floors and multiple houses that are all coming together as one building complex. As the first of the metal tier buildings designed under the new rules, I think this creates a good idea of what the metal tier will look like once all of the buildings have been upgraded - and I think it provides a good end tier for the cities you may build.
At the moment, however, the building is still a bit barebones. It is lacking furniture, plants, ornaments, and similar - all of which I will be adding over the coming week. One of the goals for that step is always to break up the somewhat monotone colour scheme of each tier, which is desirable in the main design to make each tier easily distinguishable and fitting together, but obviously wouldn't be ideal for the final result. Expect many more vines, plant pots, and similar for the building in the game.
(The building from the back)
Once that is done, I will export the building into the game, and there will only be some finishing touches left before 0.8 experimental (some sound changes, some bug fixes, some optimisation). I will be fairly busy over the coming weeks, but even in a worst case scenario, I will be able to bring 0.8 to experimental before Christmas, so the break should provide a good opportunity for some of you to be testing it.
Until next week!
Weekly Update 113 - Intermission
Hi everyone,
Between some unexpected non-game work and travelling, I have not been able to finish the metal barracks yet, so there isn’t much to say today.
The metal barracks should be finished for next week’s post, however, so more about it then!
Until next week!
Weekly Update 112 - Pistoleer Watchpost
Hi everyone,
Last week also saw some solid progress - I managed to finish the pistoleer watchpost (which is a stone buildings), and I finished the remaining building blocks for tier 3.
Considering I already introduced the swordsman watchpost in detail, there isn’t much more to be explained for the pistoleer watchpost. The pistoleer watchpost is the direct upgrade to the swordsman’s watchpost, and as such will function much the same: The pistoleer stands in front of the watchpost and waits for targets. Once he sees someone, he runs out to fight them - initially by firing his pistol, then by using his sword.
One interesting design aspect was the addition of the trellis and vine. We generally really like the idea of keeping lots of plants in each of the buildings, and while the wooden tier will likely just have small trees and bushes next to the buildings, the stone tier will start incorporating them a little more in the deliberate ornaments of the buildings.
Finally, I also want to give a bit more of an explanation of the building blocks I mentioned earlier. To make consistent buildings, I had already created the tier overviews I showed a few weeks ago. However, those overviews were meant to just get a general idea of the tiers and how they would compare to each other. They didn’t show all possible building blocks - such as doors, columns, different sizes of windows, etc. However, there are many architectural aspects that overlap between the buildings, and creating a range of them allows me to put the buildings together more consistently (and faster!).
Above is an example of the progression I made for columns. The leftmost column is the first stone column, for tier 2b. The second column is tier 3a, while the last is tier 3b. With this progression, you can see that the ceiling gets a little higher with each tier, while the column becomes a little more elaborate. Now that the tier 3 building blocks are done as well, I can tackle the metal barracks next week, which should be the last building I needed to create before 0.8 experimental.
Until next week!
Weekly Update 111 - Books and Paintings
Hi everyone,
While things are progressing steadily on my end, John has managed to find some time to get some work done on the game as well last week. The result are the books and paintings you can see in today’s image!
When I started making the new placeholder buildings, I quickly noticed that we didn’t have too many ornaments to put around them. We had some crates, a barrel, baskets, and a few tools - but too little of them for their placement to not become repetitive. As a result, I asked John to make some new ones. Generally things that depended a lot on being able to competently draw - which made them impossible for me to do myself, but made them a good fit for John.
The new books and paintings are our first results from that effort. Both the books and the paintings have a simple script attached to them which means that their placement and look are semi-random, so no two buildings you place will have the exact same decorations. And while this is only a brief tangent from the more important, big updates that need doing before 0.8, it is also a bit of a preview for some of our plans for 0.9, when a lot more aspects of the buildings will be randomised to make them less repetitive (and hints at being able to find recipes and research across the island).
My next post will get back to showing you the progress on the next buildings.
Until next week!
Weekly Update 110 - Stone Barracks
Hi everyone,
Today I want to introduce you to this (close to final) draft of the stone barracks. The stone barracks are the upgrade to the wooden barracks, and in the current version of the game are home to the pistoleers.
I think the fleshed out version matches the rough draft is showed you last week fairly closely:
But of course we are still missing more finalised materials, and the terracotta roof will consist of actual tiles when the player is close enough to it.
Considering this is the first of the new stone tier buildings, it may be interesting to see that the way in which they are constructed changes a little from the wood tier. Where in the wood tier, the construction happens by first placing the various posts, beams, and frames, after which the weave walls are inserted, the stone construction progresses a little bit more from the ground up:
There are a few finalising touches I still need to make before I will put it into the game, but I hope that I will also get around to starting the metal tier barracks before the next post.
Until next week!
Weekly Update 96 - More Control
Hi everyone,
Last week was spent on a lot of bug fixing the new combat calculations, and in addition to testing for bugs, I was also able to test how the newly simplified calculations work gameplay wise.
First up: The new soldier behaviour is working a lot better. Not only is performance vastly faster, allowing for bigger battles, but the outcome of the battles has also become a lot more predictable and influenced by tactics, as you can more or less depend on your soldiers doing what you would expect them to.
However, having fixed these issues also meant that we could see some additional pieces missing from how we would “want” the soldiers to behave. The most important one being control: with all of the soldiers acting somewhat independently from their squad, it can easily happen that multiple squads get mixed in with each other during a bigger battle - so when you try to extract one squad to close in on some enemy ranged units, it is difficult to tell which unit you are going to select, and how many of its soldiers are actually at the front line.
Likewise, selecting said squad and telling it to close the distance to some ranged units potentially means that they end up re-engaging new enemies shortly after, as they get closer to them and perceive them as a higher threat - so while you can tell your units to retreat, telling them to get behind enemy lines is a little more difficult.
Overall, this has a lot to do with the fact that the soldier behaviour was originally designed for ranged units, and has since been adjusted to also work for melee units. As a result, a lot of the behaviour is something of a hybrid, which at this point works OK for both, but not great for either. While I don’t consider it to be necessary for 0.8 experimental, I now firmly plan on separating those two behaviours before finalising 0.8 - so we have two distinct behaviour types that both work great for their respective units.
For now, however, we are just missing sounds and the new building models, so I will focus on both of those to get to 0.8 experimental.
Until next week!
Weekly Update 95 - Combat Considerations
Hi everyone,
My work last week was mostly spent on combat calculation optimisation, but while a lot of that was technical, there are still quite a few things to talk about.
Let me start as I have a few times before, but mentioning that when we started to make this game, we essentially wanted to make an RTS game that was controlled from the ground - thoughts on the FP view side of things were very limited. In early versions of the game, the commander could neither shoot nor cut down trees - and it took us a little while until we realised that in order to have the first person view be a satisfying experience, we needed to give the commander options that any player would expect from a FP game.
With that in mind, a lot of what we do needs to be balanced between making the game a satisfying RTS game, and making the game a satisfying FPS game. With combat in particular, this is not an easy line to draw. While the public version of the game still has only the very simply musketeers, the internal version of the game has had more complex AI for a while. Pirates and Soldiers weren’t directly aware of enemies that were close-by in the jungle, for example. There are noise and stealth variables, consideration of distance and obstacles between the units, and when a Pirate finally noticed a noise, he would go to investigate whether it was a soldier, villager, or if it must have been the wind - acting differently according to what he did or did not find.
It was easy to get carried away building this system when I was primarily testing its feel by running into the jungle to see how pirates would react to me - and whenever something didn't quite make sense to simply add another more complex behaviour to make up for it. But this behaviour was of course more fitting for a first person RPG or a stealth game than it may have been for ours - and came with a heavy processing cost. While having it may have been enjoyable when playing as a sole commander exploring the jungle, it did not even give much benefit to bringing soldiers along (who are terrible at stealth to begin with), or when going into big battles - leading to hundreds of thousands of calculations about each enemy soldier’s stealth and noise level when battling them in open sight on a plain.
As a result, my optimisation efforts made me revisit what we actually wanted from the soldier behaviour. It was never the plan for a commander to sneak around the island on their own - a good commander should bring an adequate escort - and if a commander did end up investigating something by themselves, a single pirate or two should pose a challenge.
What we did want, however, was to create straightforward and predictable combat between large-ish armies, where a commander would be able to redirect a squad in the middle of battle in order to flank an opponent or to retreat with fewer losses. Something that requires an easy understanding of how soldiers will see and engage opponents.
With all of that in mind, I massively simplified the system with which soldiers now engage their enemies. Rather than assessing each opponent’s threat level and attacking the highest one, soldiers now simply engage the closest opponent they can see. If they can’t see a closeby opponent, they end up searching for the closest one they can hear. This is much easier to understand in the heat of battle than wondering why one of your soldiers has decided to run behind a rock to pursue one particular opponent (who he was particularly worried about), rather than engaging a clearly visible opponent slightly further away.
I only just finished rewriting this system, and while the performance has already improved (a 40 against 40 fight won't have a constant 3200 assessments per second, but will ideally only go through 80 assessments most seconds), I haven’t been able to test too extensively yet - so more on this next week.
