It looks like a scarf to me but pretty good nonetheless
-
Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.
GAF Indie Game Development Thread 2: High Res Work for Low Res Pay
- Thread starter Jacksinthe
- Start date
- Status
It looks like a scarf to me but pretty good nonetheless
Dynamite Shikoku
Congratulations, you really deserve it!
Who was the gaffer here that is part of the team that made 'Shiftlings'? I just saw on my twitter a ripoff of it being made called 'Shift Happens'. Ugh.
taku
Member
It's unfortunate but I don't think it's a rip-off.
First of all, this specific gameplay mechanic has been seen several times before.
Secondly, it seems that Klonk Games developed a game called Mercury Shift in 2012 with the same premise and gameplay as their new one, which incidentally is very much like Shiftlings.
Dynamite Shikoku
Congratulations, you really deserve it!
Just wanted to say thanks to all the dev's that post here.
You have most likely given me the inspiration to try and start making games.
I've done some simple coding in my time but as I never really had a need for it I never really took it any further.
With the availability of accessible game engines its given me the courage to try to put something together and see how I go.
I highly doubt I will ever get to the stage of releasing a product but you never know stranger things have happened.
Finding it very interesting as you post little insights to you progress and enjoying picking up tips here and there.
You have most likely given me the inspiration to try and start making games.
I've done some simple coding in my time but as I never really had a need for it I never really took it any further.
With the availability of accessible game engines its given me the courage to try to put something together and see how I go.
I highly doubt I will ever get to the stage of releasing a product but you never know stranger things have happened.
Finding it very interesting as you post little insights to you progress and enjoying picking up tips here and there.
How do you approach alpha/beta testing? As you may know, I've uploaded two alpha demos of my game to itch.io (the first in April and an updated build in June) for public testing, both containing the first chapter. I included ingame-links to a feedback (Google) form. Since then, I got around 500 downloads and...... 7 of them filled out the form (only 3 of them with any valuable feedback).
Now I've finished the core campaign (time mode, bonus levels, level editor still missing), estimated play time about 4 hours. How do I find testers for this? I mean, I've got a couple of friends that will play through it, but that's not enough. I don't want to upload a large part of my game for free to test obviously, but I don't think a "contact me if you want to test, then I'll send you a link" will work either.
The only thing I can think of right now is putting it up on Steam Early Access for a month or so before release. Any other ideas?
Now I've finished the core campaign (time mode, bonus levels, level editor still missing), estimated play time about 4 hours. How do I find testers for this? I mean, I've got a couple of friends that will play through it, but that's not enough. I don't want to upload a large part of my game for free to test obviously, but I don't think a "contact me if you want to test, then I'll send you a link" will work either.
The only thing I can think of right now is putting it up on Steam Early Access for a month or so before release. Any other ideas?
I wonder if you could lower the barrier of entry. Right now, people need to leave your game, login to Google, fill in the forum, click stuff for some time.
In some game, I don't remember which one but I think it was a Ubisoft one. After each mission it would popup something where you could rate it from 1 to 5 stars. That would give you some very quick feedback if you could split those ratings into multiple categories like "difficulty", "controls", etc.
Also, given that you're taking "valuable" time from your testers, you should make it clear there is some reward once the game is finished (credit line, free item, etc.).
Geminosity
Member
If you're looking for testing on different hardware I can't suggest anything all that helpful, but if you're trying to refine gameplay or user experience I don't think anything beats being there in person to observe
As such, I've mostly just been taking my game on my laptop to events and watching people play. I tend to find what people do and how they react in the moment to moment while playing is a lot more valuable than asking them to try and remember how things went as lot of things will go forgotten or unnoticed.
For example, watching people play my puzzle game made me realise that no matter what I do with difficulty curves and whatnot, people will eventually find something they get stuck on and where/what brings them to a halt differs from person to person. Usually I'd see this and despite usually employing a "dont interfere while they're playing" policy, after a while I'd drop a hint that would get them going again. Realising that I don't come packaged with my game, it led me to adding a time-release hint system that only adds futher hints if you refer to it (so it will only give you a vague hint the first time you check it, regardless of when you check it and only progress to counting down to adding the next obvious hint after you've looked)
It's been extremely well received by anyone who's used it despite not being something anyone would've thought to have asked for and I don't think I would've even realised it without those observational sessions :3
jimmythepage
Member
I thought they were wearing a knotted red scarf round their face and kneck, kind of like a bandit
Thanks!
Thankyou!
Oh, now I see the scarf too, can't unsee it! xD
Geminosity
Member
Well, scarf or beard, it still looks good :3
SpacePirate Ridley
Member
This. is. fanatstic!
wow Pehesse, this was exactly what i was looking for in photoshop, those brushes are fantastic. Also I got used to line in 5px and 9px (becuase of what we used in the scribbelnauts games), but I always found there could be a better method. Going to try to use high resolutions you are using for the collectible cards art we are doing for our game.
So thanks a bunch!
Welcome speps!
jimmythepage
Member
ahah thanks again, I'll see if I can improve it a bit.
I'm using an old intous 3, where I am now I can't use my cintiq 22 HD :/
Finished this guy last night.
http://www.gfycat.com/FavorableBitterDonkey
The bullets ricochet, but they don't do damage after they bounce on something. Felt a bit unfair if they did. Sound cues are also important here.
http://www.gfycat.com/FavorableBitterDonkey
The bullets ricochet, but they don't do damage after they bounce on something. Felt a bit unfair if they did. Sound cues are also important here.
Hahaha, I know, right? Felt like christmas when I stumbled upon that set!
Cintiq! (Smallest size, can't remember the number). It's actually one of the first things I invested in when I tried to enter the industry! (That reminds me, I'm actually running low on pen tips now, I'll have to order new ones soon.. anybody have any experience with that?)
Jacksinthe
Banned
+1 for Cintiq
Bloodember
Member
That looks pretty awesome. How big are your sprites?
electroflame
Member
Thanks for sharing, that one actually looks pretty decent.
I found two other ones that I really like: MessageKit and Frictionless.
MessageKit is probably the one that I'll use as a base, I think. I really like the ideas behind Frictionless, but it seems to be a bit too much for what I think I'll need. Having said that, though, I think I'll probably write my own EventManager class as a wrapper for whatever I choose, that way if I decide that I need something else later I can just swap out the guts of that class and not go code-hunting.
Thanks for the help!
Jacksinthe
Banned
#ScreenshotSaturday tomorrow, folks. Dress your best and go to town on Twitter. I'll be RT and quoting with extra hashes throughout the day!
Also, if you followed @DevGAF, are a dev and I haven't hit a follow back - PM ME so I know who you are. I want to make sure I do some extra reach for everyone tomorrow.
I think our main character is 7 pixels tall. If the pixel scaling were 1:1 our entire game would fit on a 192x108 screen. We use a scaling of 10:1 for about 90% of our objects and some are 1:1 blended in (when needed to limit color banding for fades).
Also, if you followed @DevGAF, are a dev and I haven't hit a follow back - PM ME so I know who you are. I want to make sure I do some extra reach for everyone tomorrow.
I think our main character is 7 pixels tall. If the pixel scaling were 1:1 our entire game would fit on a 192x108 screen. We use a scaling of 10:1 for about 90% of our objects and some are 1:1 blended in (when needed to limit color banding for fades).
How effective is that stomp attack of his? It doesn't seem like there's any need for little blue to get close enough to the monster that she'd be in range of it. It's very pretty as usual.
Thanks
If you're touching the ground the stomp will stun and heavily damage you
http://www.gfycat.com/FixedLiquidKoala
And you do need to get close, the monster can only be harmed if you're gassed up, and the only source of gas near him is.. him.
The "gas up" mechanics are specific to this region and already introduced to the player in a more controlled way before fighting anything hard.
Thanks
If you're touching the ground the stomp will stun and heavily damage you
http://www.gfycat.com/FixedLiquidKoala
And you do need to get close, the monster can only be harmed if you're gassed up, and the only source of gas near him is.. him.
The "gas up" mechanics are specific to this region and already introduced to the player in a more controlled way before fighting anything hard.
Oh, I can see that now. I noticed the green glow effects but didn't put two and two together. Very nice!
Update: I might actually quit sooner for the sake of my sanity.
Don't do it, time goes fast when you are developing and your savings dissappear quickly with it.
In my case I was fired so I didn't had a choice but you still have one. Unless your game generates massive hype before release, don't bet your future on it.
electroflame
Member
So, uh, yeah. Getting Git submodules and symlinks working with Unity is kind of a pain.
electroflame
Member
Phew, finally got it working! Now I can have my common imports all in a side repository that I can update whenever and pull the changes into whatever project I want. This has the added bonuses of A.) being easier to manage for multiple projects and B.) making .unitypackage imports way easier, as I can let it dump all of the files into the root Assets folder of the submodule repo.
So, basically, I set it up like this:
(My project directory, from Windows)
Next, I use a Directory Junction/symlink to link the top-level Submodules/Assets directory to a Submodules directory inside of my Assets folder. (This'll make short work of this step if you're not command-prompt savvy)
So, my Assets directory looks something like this:
You may be wondering why I have it inside of a Standard Assets folder. Standard Assets is one of Unity's Special Folders, and is compiled before anything else is compiled. This is great for dependancies that need to be ready for other scripts, but the main reason is that Standard Assets is only compiled if its contents are changed. If you have a lot of things in your Submodules directory, this can drastically speed up your compile time.
It took more than a few stabs in the dark, but I'm very pleased with how it's working. I highly recommend it if you have a lot of project-independent code that you tend to import/share a lot.
So, basically, I set it up like this:
(My project directory, from Windows)
- Assets
- Library
- ProjectSettings
- Submodules
Next, I use a Directory Junction/symlink to link the top-level Submodules/Assets directory to a Submodules directory inside of my Assets folder. (This'll make short work of this step if you're not command-prompt savvy)
So, my Assets directory looks something like this:
- Resources
- Scripts
- Third-Party
- Standard Assets
- Submodules
- Standard Assets
You may be wondering why I have it inside of a Standard Assets folder. Standard Assets is one of Unity's Special Folders, and is compiled before anything else is compiled. This is great for dependancies that need to be ready for other scripts, but the main reason is that Standard Assets is only compiled if its contents are changed. If you have a lot of things in your Submodules directory, this can drastically speed up your compile time.
It took more than a few stabs in the dark, but I'm very pleased with how it's working. I highly recommend it if you have a lot of project-independent code that you tend to import/share a lot.
Icyflamez96
Member
I was looking a bit through the thread. There is so much awesome stuff posted. There's a lot of talent here on GAF.
This seems like a solution to a problem I didn't even know I had. Will keep this in mind for my next project. =)
Ooo.. I like the pony tail version.
I fixed her knee and ankle, made the skirt look less stiff, added the weapon cases, and tweaked the delay in her arm. Also ponytail version.
I still think the right leg (her left) looks off. The knee isn't rotating enough for that foot, and it doesn't feel like it's bending. A lot of the up and down movement looks like it's coming from the foot/ankle. It also feels like the whole leg is detaching from the pelvis.
Maybe simplify the movement and keep the feet planted. The art itself is great, just the animation needs a little more tweaking.
And I agree with Jobbs. That's a whole lot of detail and overlapping elements. Looks awesome, but man, that's gonna be a lot of work.
It took pretty much all week, but I completely overhauled the map generation system in our game - before it was a very rigid system where you said "use this algorithm with these parameters" and it always spit out 2 floors, a shop, and a boss level with no real ability to customize it.
The entire system is now a pipeline of discrete steps, and each zone has a data file, where each level has a block that looks something like this. This lets me change big stuff like map size per floor, down to little stuff like tweaking what set of decoration steps to use. And it's gone from being a bear to wrestle with to something that I really really enjoy using.
I've made a bunch of other changes as I go, like adding a first pass of a snow effect to our london/town level:
And the ability to add interior columns to our mansion level rooms for more permanent cover options:
The last big thing I did was a change to our core layout algorithm for the mansion - before it was always a directed acyclic graph which works great but it starts to feel a bit same'y when you always have one long path with offshoots and no loops. So I modified the algorithm to go back after the initial layout has been done and add loops wherever it can as long as it doesn't cause a shortcut that significantly reduces the distance from start to finish. Example:
So these maps started with the same generated room layout. On the left you can see the directed acyclic graph version, which gives you a main path (in blue) and offshoots (in green) that branch off and never really overlap. Good for placing objectives at dead ends, not so great for anything else. The version on the right is from the new system, where there's still a main path (although even that supports branching now), but the entire level is far more interconnected. One downside is this makes doing 'locked door' type objectives significantly more difficult - not sure how I'm going to handle that yet.
So, yeah. Lots more to do on the map generation front but really happy with how the new system is working out so far.
The entire system is now a pipeline of discrete steps, and each zone has a data file, where each level has a block that looks something like this. This lets me change big stuff like map size per floor, down to little stuff like tweaking what set of decoration steps to use. And it's gone from being a bear to wrestle with to something that I really really enjoy using.
I've made a bunch of other changes as I go, like adding a first pass of a snow effect to our london/town level:
And the ability to add interior columns to our mansion level rooms for more permanent cover options:
The last big thing I did was a change to our core layout algorithm for the mansion - before it was always a directed acyclic graph which works great but it starts to feel a bit same'y when you always have one long path with offshoots and no loops. So I modified the algorithm to go back after the initial layout has been done and add loops wherever it can as long as it doesn't cause a shortcut that significantly reduces the distance from start to finish. Example:
So these maps started with the same generated room layout. On the left you can see the directed acyclic graph version, which gives you a main path (in blue) and offshoots (in green) that branch off and never really overlap. Good for placing objectives at dead ends, not so great for anything else. The version on the right is from the new system, where there's still a main path (although even that supports branching now), but the entire level is far more interconnected. One downside is this makes doing 'locked door' type objectives significantly more difficult - not sure how I'm going to handle that yet.
So, yeah. Lots more to do on the map generation front but really happy with how the new system is working out so far.
electroflame
Member
I needed to cook something like this up due to the way Unity 5 broke importing/upgrading Asset Store stuff (i.e. you have to keep it or move it back to the root to have the .unitypackage overwrite successfully). It's also handy to have my common-use frameworks all in one place where any change I make can be pushed back to the common repo and propagated to whatever project makes use of it.
It was a pain in the butt to work out, but it'll save me so much time. It's also pretty easy to add to an existing project (basically just set it up like I had it, but before symlinking the Submodules folder delete any copies of the assets/code in your current project that's present in the Submodules repo).
One caveat is that this doesn't seem to work with things that make use of Unity's "Plugins" special folder. Due to the nature of the folder (i.e. only one per project, and it has to be in the Assets directory root) there's no easy way to handle that. Since "Editor Default Resources" works similarly, I would imagine that that won't work very well either.
Still, for most things (bar those two folder scenarios) it works really well. Now that I know how to do it it won't be nearly as difficult next time. I should probably document this somewhere for posterity, though. Maybe it's time to update that Github pages site I claimed but never update...
It took pretty much all week, but I completely overhauled the map generation system in our game - before it was a very rigid system where you said "use this algorithm with these parameters" and it always spit out 2 floors, a shop, and a boss level with no real ability to customize it.
The entire system is now a pipeline of discrete steps, and each zone has a data file, where each level has a block that looks something like this. This lets me change big stuff like map size per floor, down to little stuff like tweaking what set of decoration steps to use. And it's gone from being a bear to wrestle with to something that I really really enjoy using.
I've made a bunch of other changes as I go, like adding a first pass of a snow effect to our london/town level:
Those pulled-back shots don't do your game justice. It looks really great up close. Though it's neat to see the systems at play. Procedural stuff fascinates me extensively. I thought it'd get old already (which worried me, as I'm also making a procedural game), but it keeps on being interesting. I think it's here to stay.
I haven't touched the map generation code in my game in almost a year. I'm super nervous to approach it because I barely have any idea of how it works anymore or which strings to pull to retool different aspects.
It took pretty much all week, but I completely overhauled the map generation system in our game - before it was a very rigid system where you said "use this algorithm with these parameters" and it always spit out 2 floors, a shop, and a boss level with no real ability to customize it.
The entire system is now a pipeline of discrete steps, and each zone has a data file, where each level has a block that looks something like this. This lets me change big stuff like map size per floor, down to little stuff like tweaking what set of decoration steps to use. And it's gone from being a bear to wrestle with to something that I really really enjoy using.
I've made a bunch of other changes as I go, like adding a first pass of a snow effect to our london/town level:
And the ability to add interior columns to our mansion level rooms for more permanent cover options:
The last big thing I did was a change to our core layout algorithm for the mansion - before it was always a directed acyclic graph which works great but it starts to feel a bit same'y when you always have one long path with offshoots and no loops. So I modified the algorithm to go back after the initial layout has been done and add loops wherever it can as long as it doesn't cause a shortcut that significantly reduces the distance from start to finish. Example:
So these maps started with the same generated room layout. On the left you can see the directed acyclic graph version, which gives you a main path (in blue) and offshoots (in green) that branch off and never really overlap. Good for placing objectives at dead ends, not so great for anything else. The version on the right is from the new system, where there's still a main path (although even that supports branching now), but the entire level is far more interconnected. One downside is this makes doing 'locked door' type objectives significantly more difficult - not sure how I'm going to handle that yet.
So, yeah. Lots more to do on the map generation front but really happy with how the new system is working out so far.
This is all very cool looking in motion, and interesting to read about!
I fixed her knee and ankle, made the skirt look less stiff, added the weapon cases, and tweaked the delay in her arm. Also ponytail version.
Agreed with above, the main (only) issue I have is with her left leg bending. As it is, the knee rises up and down, giving the feel that the thigh is coming out from the hip, or that the shin is getting shorter.
I'd suggest rotating the shin+knee towards the inside instead of lowering it. I think you're going for a shortening perspective thing where the leg bends towards us, but this kind of animation using pieces doesn't lend itself to that specific trick, at least not in my eyes, since the different pieces don't visibly change size (the armor leg parts and the knee, most notably). To make it work, maybe you could try shadowing the shin progressively as it's supposed to bend further near the ground - otherwise, I'd just take the simplest route and rotate the leg on the other plane: on the side, rather than towards us, though that might mean you'll need more hip movement.
Here's an overly exagerated attempt, focused mainly on hip movement rather than leg movement. Don't know if that's what you're going for, but just in case. (I didn't really bother with the top part/shoulders/head/hair since you've got that locked down pretty tight!)
The main thing here is that to convey hip movement, it's mostly the thighs that are moving around - the shins stay relatively immobile, apart from some rotation on their own axis, which you do quite well.
It's very rough and dirty, so I still had to use some shortening to convey the hips lowering, but I try to use that mainly on the thighs, not the shins (though there's still a bit of that on the right leg). Those should be still be able to rotate without too much trouble for a less exagerated sway/squat. And as for shortening the thighs, you could always cover them gradually more with her midriff ribon, simulating that kind of effect while not actually changing the length of any part? I also tried rotating the torso a bit more for each sway, but I'm not sure it's actually possible past a certain point with your setup. Again, just in case!
SpacePirate Ridley
Member
I fixed her knee and ankle, made the skirt look less stiff, added the weapon cases, and tweaked the delay in her arm. Also ponytail version.
Looks so lively and full of detail!
It took pretty much all week, but I completely overhauled the map generation system in our game - before it was a very rigid system where you said "use this algorithm with these parameters" and it always spit out 2 floors, a shop, and a boss level with no real ability to customize it.
The entire system is now a pipeline of discrete steps, and each zone has a data file, where each level has a block that looks something like this. This lets me change big stuff like map size per floor, down to little stuff like tweaking what set of decoration steps to use. And it's gone from being a bear to wrestle with to something that I really really enjoy using.
I've made a bunch of other changes as I go, like adding a first pass of a snow effect to our london/town level:
And the ability to add interior columns to our mansion level rooms for more permanent cover options:
The last big thing I did was a change to our core layout algorithm for the mansion - before it was always a directed acyclic graph which works great but it starts to feel a bit same'y when you always have one long path with offshoots and no loops. So I modified the algorithm to go back after the initial layout has been done and add loops wherever it can as long as it doesn't cause a shortcut that significantly reduces the distance from start to finish. Example:
So these maps started with the same generated room layout. On the left you can see the directed acyclic graph version, which gives you a main path (in blue) and offshoots (in green) that branch off and never really overlap. Good for placing objectives at dead ends, not so great for anything else. The version on the right is from the new system, where there's still a main path (although even that supports branching now), but the entire level is far more interconnected. One downside is this makes doing 'locked door' type objectives significantly more difficult - not sure how I'm going to handle that yet.
So, yeah. Lots more to do on the map generation front but really happy with how the new system is working out so far.
The gif looks amazing, fast and pixelart looks really cool.
We are making similar type of game apart from the beat 'm up sections, with procedural enviroments also, so its pretty interesting reading about this stuff.
Wow this is gonna take a while I guess.
Thank you so much Pehesse, I really needed advice from someone with your experience and talent. Thanks for taking the effort to help me with those references
My goal was moving the knee closer to the screen, since the other option (as showcased in your gif) has too much movement.
I'm gonna do two things.
First, making the skirt-less version so I can see better where the legs begin, start, and blend.
Then, I'll try making the knee bigger, rotate a bit, and bring it closer to the ground to achieve that depth effect. I'll see what I can do.
Thanks again, to you, and to bkw, your comment also helped me a lot
Thank you so much Pehesse, I really needed advice from someone with your experience and talent. Thanks for taking the effort to help me with those references
My goal was moving the knee closer to the screen, since the other option (as showcased in your gif) has too much movement.
I'm gonna do two things.
First, making the skirt-less version so I can see better where the legs begin, start, and blend.
Then, I'll try making the knee bigger, rotate a bit, and bring it closer to the ground to achieve that depth effect. I'll see what I can do.
Thanks again, to you, and to bkw, your comment also helped me a lot
jimmythepage
Member
idle/walk
And this is how the character should look, I couldn't figure out a nice way to draw the eyes in top-down view :/
And this is how the character should look, I couldn't figure out a nice way to draw the eyes in top-down view :/
idle/walk
And this is how the character should look, I couldn't figure out a nice way to draw the eyes in top-down view :/
I thought the mustache was a scarf lol
Thanks for the advice. You're right.
I was just having one of those days. I slept it off. Back to my game now.
jimmythepage
Member
ahah I need to find a way to fix this I guess xD
Glad you're feeling better! It's indeed not a good decision to make without taking a good night's sleep first, at least...
idle/walk
And this is how the character should look, I couldn't figure out a nice way to draw the eyes in top-down view :/
I confess, I kinda see the scarf more than the beard as well, even though I understand the intention. Don't really have any advice as to how make it more legible, or the eyes more visible... but even if it's a scarf, I think it works!
No worries! I'm curious to see if shading the shin might help change my perception on what happens here, so I'll be waiting for the next version of it :-D Don't kill yourself over one animation, though... Gotta move on at some point!
Did a quick and dirty implementation of some of my animations for Big Blue, just to see if it all lines up more or less correctly. I've already spotted some color balance stuff to fix, and most of the anchor point problems will be dealt with when I integrate for real... but I'm pretty happy it seems to be working otherwise!
Looks sexy
That was something I improved in the second alpha: You only had to roll into a specific "teleporter" that opens the feedback form in your browser. Ingame rating after each level would be great, but would probably take me weeks of work. The reward thing is something I'll keep in mind, I wouldn't mind giving somebody a free Steam key for the final game if the feedback was helpful. Maybe I should do a private alpha for GAF users.
Yeah, that's the best kind of feedback you can get. A few YouTubers played the alpha and that really helped a lot. Getting to events takes a lot of time though and you probably wouldn't get feedback for a full campaign.If you're looking for testing on different hardware I can't suggest anything all that helpful, but if you're trying to refine gameplay or user experience I don't think anything beats being there in person to observe
As such, I've mostly just been taking my game on my laptop to events and watching people play. I tend to find what people do and how they react in the moment to moment while playing is a lot more valuable than asking them to try and remember how things went as lot of things will go forgotten or unnoticed.
---
Screenshot Saturday, right?
This is from one of the largest puzzles in the game (the screen shows about 1/3 of the laser grid). It has an "easy" route for lazy people and a more complex one for those that want to collect all points (yellow). May cause permanent brain damage.
Mystic River
Member
I fixed her knee and ankle, made the skirt look less stiff, added the weapon cases, and tweaked the delay in her arm. Also ponytail version.
legs look fine , but something that bothers me is the chest , looks still too stiff or moving actually ahead , delaying the movement a frame or two will fix that so it won't move in sync when returning back
industrian
will gently cradle you as time slowly ticks away.
Does anyone have any advice for creating a menu system in Monogame?
Finally have painting of trees working:
Still not perfect though, most importantly spline mesh collisions currently have a bug in UE4 and I'm not sure if I will go with growing for the trees instead of instant painting:
I'm also curious if I can make it work to let the trees randomly grow on their own.
The Blueprint I created can then also be used for painting all kinds of other spline based stuff like pipes, ropes, cables, tubes etc. And with some tweaks hopefully tracks so I can draw me some rollercoasters.
Just painting strokes that you can walk on should be also cool.
Still not perfect though, most importantly spline mesh collisions currently have a bug in UE4 and I'm not sure if I will go with growing for the trees instead of instant painting:
I'm also curious if I can make it work to let the trees randomly grow on their own.
The Blueprint I created can then also be used for painting all kinds of other spline based stuff like pipes, ropes, cables, tubes etc. And with some tweaks hopefully tracks so I can draw me some rollercoasters.
Just painting strokes that you can walk on should be also cool.
Yeah I was in a similar boat, I wrote the bulk of the map code about 18 months ago and any changes since then had just been hacks to shoehorn in whatever bits I needed. Is it working well enough that you can just leave it be? Or do you feel like you'll have to rework it eventually?
Finally have painting of trees working:
Still not perfect though, most importantly spline mesh collisions currently have a bug in UE4 and I'm not sure if I will go with growing for the trees instead of instant painting:
I'm also curious if I can make it work to let the trees randomly grow on their own.
This is really cool! Have you looked into using L systems for having them grow on their own?
I fixed her knee and ankle, made the skirt look less stiff, added the weapon cases, and tweaked the delay in her arm. Also ponytail version.
I'm just gonna say what's been on my mind for a long time because I think it's probably constructive, but if I'm wrong I'm wrong.
You're spending an enormous amount of time and effort on all of these tiny details and all of this sub-animation, which is definitely an impressive thing on its own, and it's worth pointing out. However, I think your energy is not being spent right. The overall outline of your character motions are not as good as they should be, especially for an action brawler, and I think the need to preserve all of these meticulous detail is boxing you in with these outlines and contributing to the problem. Your rendering is good, but your outline and the conceived poses need work, and it's hard for you to address those problems when there's such a high burden on rendering.
I don't enjoy typing that out, because of all the work you've put in, but I think the outline is the problem, not the detail and sub animations, and the need to include all of the detail and sub animation I feel is leading to a "cart before horse" problem with your animation and ultimately holding it back.
(When I say the outline is the problem I don't necessarily mean the idle animation in the quoted portion, which is not that bad, I'm talking about the overall moveset of the character)
Yeah, it's working well enough, for sure. It's just kind of simple. Right now it builds a critical path based on a dynamic list of goals (that I can easily change). By goals, I just mean something like checkpoint room, finish room, miniboss room or any of the various puzzle rooms. Using a constant, the generator builds X rooms in a string and then builds Y goal, until all of the goals are accomplished. Then the code runs back through the critical path and tries to find places it can punch a hole in the wall to build a branching path, which itself can branch, for a max 3 distance from the critical path. Finally, it puts a store in one of the terminals close to the start and a special token in one of the terminals near the end (for unlocking the secret ending).
I don't know if that sounds simple, but there are no loops, and you don't have to do any backtracking technically, assuming you're able to guess the critical path at each fork. The goals that are on branches are optional and not super sophisticated. For now, this is fine. I'm focusing on making all of the direct experiences enjoyable without some overarching metagaming, but there might come a point down the road when I want that too.
edit: I should add that for my purposes map generation and room generation are two separate beasts. For room generation, I've built almost 3000 templates that get filled with a separate procedure that I'm completely happy with.
I've thought this too. It's a common trait to super-detailed animation, where the details are more important than the animation.
What's hard about criticizing it is that the details look great in each of the frames. So I haven't said anything either, because choosing pretty details over expressive animation is just a style choice to me. It's not necessarily better or worse in my opinion.
- Status