Citan
04-19-2017, 09:35 PM
This is just a quick update to last week's blog post, because I'm still working the same things! Progress on the Serbule revisions and the Brewing skill are coming along, but we're still a week or so away from being ready to launch.
Serbule Updates
This particular iteration of Serbule is focusing on three things: the textures and models in Serbule city (some of which are already live), the terrain (the grasses, trees, and so on), and the sky. We develop iteratively, so I'm not trying to make the "perfect Serbule" here, just a "better Serbule". And it turns out to be easy to make Serbule look better... the trick is making it look better AND have better framerate -- or at the very least, not-worse framerate.
Graphics work always seems to take vastly longer than I hope it will. Part of that is because it needs to be tested on lots of hardware -- and we don't have a full set of testing hardware, so we also have to try to "simulate" low-end machines, which requires some development work. But it's coming along, slowly but surely. Here's a few screenshots. They're a couple days old, and the game already looks a bit different from these, but it will give you an idea of where we are in the process.
http://i.imgur.com/EaqkDl6.png
This first screenshot shows the sky system. Dawn is just breaking, and it looks to be a foggy morning outside of Serbule Keep. The sun, moon, fog levels, and clouds are all modeled, and later on rain will be modeled too, with the Weather Witch skill intertwined into the weather systems!
http://i.imgur.com/Y3NA1h5.png
Here's late at night. I turned off the nearby lamps to get a feel for how dark it can get. I'm still not happy here -- it's a work in progress. (That building is the "cow barn" in Serbule, which is undergoing renovations, as you can see.)
http://i.imgur.com/vvFMrhv.png
The grass and trees can be rendered at a farther distance now, with higher detail density. (This is just before dusk.)
http://i.imgur.com/oGpyBjb.jpg
Another foliage shot. I just liked this one. We still need to do more work on the grass details that you can see in the medium-distance, but I think it's already a notable improvement from the old foliage. (This is the afternoon.)
Brewing Skill
Working on graphics optimization is very time-consuming, but it has a lot of downtime. It takes 30 minutes to compile and copy each test build, and I need to make a whole lot of tests each day. While I wait for it to compile, I've been working on the brewing skill. Brewing has been one of the most-requested crafting skills, so I want it to be fun.
The basic ideas of brewing are the same as I described last week, but the details have changed many times. In fact, I think this is the most times I've ever iterated on a craft skill before it went live! The first few versions were prototypes, trying to figure out what the system's goals were and how it would achieve them. I blogged about the system last week based on a fairly fun prototype version. But then I needed to future-proof the system, which turned out to require a full rewrite.
I don't usually bother trying to future-proof crafting skills, because rewriting it later doesn't usually cause alpha-testers too much pain. When I rewrite a craft skill, you keep your old level and recipes, but the contents of those recipes change. No big deal. But brewing is different: brewing recipes have randomized results which will require a lot of player experimentation, so I want to protect that time investment. For instance, if I later decide that apples aren't a low-level fruit anymore, and replace them with, I dunno, kiwifruit, what happens to brewing recipes that can take apples? Obviously, kiwi should be a drop-in replacement in those brewing recipes, doing the same things as apples used to do, so that you don't have to re-try every brewing ingredient combination. But that's not an automatic feature -- it had to be coded that way. I brainstormed other ways that the skill might change in the future, and I tried to make sure those changes wouldn’t mess brewers up too badly. I can't guarantee that everything will work out right -- who knows, a bug might screw everything up. But I've given it my best shot.
After that, I realized the skill was way TOO random: every time you learned a new recipe, you had to start your experimentation all over again. That's fine at low level when there's not too many ingredient combinations, but by level 50 there's over a hundred brewing outcomes! If levels 50, 60, 70, 80, etc. each had 100% different random outcomes, it wouldn't make economic sense to experiment at lower levels. Instead, players would just grind as fast as they could to the highest level and experiment with only the high-level recipes. That's a boring design! I needed a system that lets you "carry over" some of your brewing-experiment knowledge from mid-levels to higher levels. I've got a system that does that now -- although it might carry over a bit too much info... I'm still fiddling with things here.
And in between all these revisions, I've been trying out tons of different possible buffs that could come from drinking booze. I had to answer some tricky questions, like: how many drink-effects can you have at once? If you can "stack" too many booze buffs then each individual drink would have to be weak and uninteresting. But if they don't stack at all, there's no chance to mix-and-match drink effects. After some experiments, I've decided that you can have three beers (or glasses of wine) at once, plus one drink of hard liquor, for a total of four stacking alcohol buffs. That's a lot of buffs! I'm trying to make the drink effects somewhat useful by themselves, but also make them more impressive when stacked together. It's a delicate balance. However, the specific buff effects can be changed later as balance demands, so I'm not TOO worried about this part.
I've also made sure there's room for various systemic interactions later on. For instance, I was thinking it might be fun if beer brewed during the full moon is more random, with a chance to have higher or lower stats than normal. But what would it mean for a beer to be "extra effective" or "less effective" than normal? I had to work all that out too.
In other words, there's lots and lots of design questions! Nothing earth-shatteringly hard, but it's been keeping me busy.
So what's the time table for this next snapshot? Probably the middle of next week. I really wanted to get it out the door by Friday, but there's still too much left to do. But the results will be worth the wait. I'll talk to you again soon, with more screenshots -- maybe some brewing screenshots next time!
Serbule Updates
This particular iteration of Serbule is focusing on three things: the textures and models in Serbule city (some of which are already live), the terrain (the grasses, trees, and so on), and the sky. We develop iteratively, so I'm not trying to make the "perfect Serbule" here, just a "better Serbule". And it turns out to be easy to make Serbule look better... the trick is making it look better AND have better framerate -- or at the very least, not-worse framerate.
Graphics work always seems to take vastly longer than I hope it will. Part of that is because it needs to be tested on lots of hardware -- and we don't have a full set of testing hardware, so we also have to try to "simulate" low-end machines, which requires some development work. But it's coming along, slowly but surely. Here's a few screenshots. They're a couple days old, and the game already looks a bit different from these, but it will give you an idea of where we are in the process.
http://i.imgur.com/EaqkDl6.png
This first screenshot shows the sky system. Dawn is just breaking, and it looks to be a foggy morning outside of Serbule Keep. The sun, moon, fog levels, and clouds are all modeled, and later on rain will be modeled too, with the Weather Witch skill intertwined into the weather systems!
http://i.imgur.com/Y3NA1h5.png
Here's late at night. I turned off the nearby lamps to get a feel for how dark it can get. I'm still not happy here -- it's a work in progress. (That building is the "cow barn" in Serbule, which is undergoing renovations, as you can see.)
http://i.imgur.com/vvFMrhv.png
The grass and trees can be rendered at a farther distance now, with higher detail density. (This is just before dusk.)
http://i.imgur.com/oGpyBjb.jpg
Another foliage shot. I just liked this one. We still need to do more work on the grass details that you can see in the medium-distance, but I think it's already a notable improvement from the old foliage. (This is the afternoon.)
Brewing Skill
Working on graphics optimization is very time-consuming, but it has a lot of downtime. It takes 30 minutes to compile and copy each test build, and I need to make a whole lot of tests each day. While I wait for it to compile, I've been working on the brewing skill. Brewing has been one of the most-requested crafting skills, so I want it to be fun.
The basic ideas of brewing are the same as I described last week, but the details have changed many times. In fact, I think this is the most times I've ever iterated on a craft skill before it went live! The first few versions were prototypes, trying to figure out what the system's goals were and how it would achieve them. I blogged about the system last week based on a fairly fun prototype version. But then I needed to future-proof the system, which turned out to require a full rewrite.
I don't usually bother trying to future-proof crafting skills, because rewriting it later doesn't usually cause alpha-testers too much pain. When I rewrite a craft skill, you keep your old level and recipes, but the contents of those recipes change. No big deal. But brewing is different: brewing recipes have randomized results which will require a lot of player experimentation, so I want to protect that time investment. For instance, if I later decide that apples aren't a low-level fruit anymore, and replace them with, I dunno, kiwifruit, what happens to brewing recipes that can take apples? Obviously, kiwi should be a drop-in replacement in those brewing recipes, doing the same things as apples used to do, so that you don't have to re-try every brewing ingredient combination. But that's not an automatic feature -- it had to be coded that way. I brainstormed other ways that the skill might change in the future, and I tried to make sure those changes wouldn’t mess brewers up too badly. I can't guarantee that everything will work out right -- who knows, a bug might screw everything up. But I've given it my best shot.
After that, I realized the skill was way TOO random: every time you learned a new recipe, you had to start your experimentation all over again. That's fine at low level when there's not too many ingredient combinations, but by level 50 there's over a hundred brewing outcomes! If levels 50, 60, 70, 80, etc. each had 100% different random outcomes, it wouldn't make economic sense to experiment at lower levels. Instead, players would just grind as fast as they could to the highest level and experiment with only the high-level recipes. That's a boring design! I needed a system that lets you "carry over" some of your brewing-experiment knowledge from mid-levels to higher levels. I've got a system that does that now -- although it might carry over a bit too much info... I'm still fiddling with things here.
And in between all these revisions, I've been trying out tons of different possible buffs that could come from drinking booze. I had to answer some tricky questions, like: how many drink-effects can you have at once? If you can "stack" too many booze buffs then each individual drink would have to be weak and uninteresting. But if they don't stack at all, there's no chance to mix-and-match drink effects. After some experiments, I've decided that you can have three beers (or glasses of wine) at once, plus one drink of hard liquor, for a total of four stacking alcohol buffs. That's a lot of buffs! I'm trying to make the drink effects somewhat useful by themselves, but also make them more impressive when stacked together. It's a delicate balance. However, the specific buff effects can be changed later as balance demands, so I'm not TOO worried about this part.
I've also made sure there's room for various systemic interactions later on. For instance, I was thinking it might be fun if beer brewed during the full moon is more random, with a chance to have higher or lower stats than normal. But what would it mean for a beer to be "extra effective" or "less effective" than normal? I had to work all that out too.
In other words, there's lots and lots of design questions! Nothing earth-shatteringly hard, but it's been keeping me busy.
So what's the time table for this next snapshot? Probably the middle of next week. I really wanted to get it out the door by Friday, but there's still too much left to do. But the results will be worth the wait. I'll talk to you again soon, with more screenshots -- maybe some brewing screenshots next time!