AusRoachman
Member
Somehow i think i screwed up my procedural map size algorithm. This map is just a little too cramped
-
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
HandsomeCharles
Member
Jacksinthe
Banned
So Unity RigidBody2D question for those of you using them to control your characters in game:
I'm toying with making a controller using RigidBody2D for the side project and have a question about moving platforms.
For MF1 I made my controller kinematic and move the sucker with translate and collision with rsyvasts so this is my first attempt at using RB2D. I can work with it just fine but for now I am having some issues with platforms.
My only solution is to move the platform using sqrmag with the start and end positions and speed, then when a character is on the platform simply add the Vector2 Velocity of the platform to the controller's velocity. Everything works but is there a more elegant solution?
Using translate in MF1 it moves in local space so I just child the player to the platform and Lerp the position of the platform and it works fine.
Curious if there's another solution for RB.
I'm toying with making a controller using RigidBody2D for the side project and have a question about moving platforms.
For MF1 I made my controller kinematic and move the sucker with translate and collision with rsyvasts so this is my first attempt at using RB2D. I can work with it just fine but for now I am having some issues with platforms.
My only solution is to move the platform using sqrmag with the start and end positions and speed, then when a character is on the platform simply add the Vector2 Velocity of the platform to the controller's velocity. Everything works but is there a more elegant solution?
Using translate in MF1 it moves in local space so I just child the player to the platform and Lerp the position of the platform and it works fine.
Curious if there's another solution for RB.
Guys,
I've started working on the world map and I've implemented my basic ideas, like moving to different places, different objects on the map, a dotted line that follows the player position etc.
Now the question is what style to go for. The typical cloth "treasure hunt pirates" map, or a modern simplistic map GTA style?
Or the futuristic blueish glowy thing?
Can you guys tell me what your favorite maps in games are? Any inspirational material?
Thanks
----
Note:
The game is set in a mad-max inspired universe of desert and cars
I've started working on the world map and I've implemented my basic ideas, like moving to different places, different objects on the map, a dotted line that follows the player position etc.
Now the question is what style to go for. The typical cloth "treasure hunt pirates" map, or a modern simplistic map GTA style?
Or the futuristic blueish glowy thing?
Can you guys tell me what your favorite maps in games are? Any inspirational material?
Thanks
----
Note:
The game is set in a mad-max inspired universe of desert and cars
I'm actually applying velocity to the rigidbody instead on walk and jump actions and it works fine as well, but we aren't using moving platforms so we don't have to figure that out, either.
I guess you could either take a look at the character controller's system, since it does support moving platforms in some way, or you could implement something for the MC such as:
Code:
private Rigidbody2D platformRigidbody; //Or script, whatever
private void OnCollisionEnter2D(Collision2D collision)
{
if (collision.gameObject.CompareTag("MovingPlatform"))
{
platformRigidody = collision.gameObject.GetComponent<Rigidbody2D>();
}
}
private void OnCollisionStay2D(Collision2D collision)
{
if (platformRigidbody.gameObject == collision.gameObject)
{
myRigidbody.velocity += platformRigidbody.velocity;
}
}
private void OnCollisionExit2D(Collision2D collision)
{
if (platformRigidbody.gameObject == collision.gameObject)
{
platformRigidody = null;
}
}Another way you could work with them would be by parenting the player to the platform while the two are touching, but to do that correctly I think you need the platform's root transform (the one that the player gets parented to) to have a scale of (1,1,1), because the last time I tried, other scales would distort the player's graphics in really weird ways while parented to a platform with a different set of scale values.
Jacksinthe
Banned
Childing a RB is a bad idea. It's fine if it is kinematic as it will follow the parent locally but a non-kinematic will break.
Right now I just use OverlapArea to dictate wether or not the player is on the platform. I don't trust Enter/Exit/Stay XD - Will probably just use raycasts on the controller to dictate what he's standing on since I already use them for grounded bools since enter/exit and OverlapArea flip true if colliders touch, let alone overlap which makes false positives.
I prefer raycasts, still. I'm just trying to create a good controller with RB. I still prefer my old method but I'd like to at least try using RBs.
LordRaptor
Member
Its definitely worth looking at the character controller system - you get a 'free' pseudo-raycast system to determine character collisions with collisionflags
If its post apocalyptic future, maybe something mundane like a tourist map of the area thats been modified with handwritten notes / landmarks added would be a nice way of showing that.
Sam & Max Hit The Road did a similar thing for laughs, where their map of the world was a thing they got free with a kids meal from a burger place.
If its post apocalyptic future, maybe something mundane like a tourist map of the area thats been modified with handwritten notes / landmarks added would be a nice way of showing that.
Sam & Max Hit The Road did a similar thing for laughs, where their map of the world was a thing they got free with a kids meal from a burger place.
That's a brilliant idea!
I've decided to go with the whole old paper idea and i looked at some other simple maps.
A quick scribble and some holes and I think I like where this is going.
Maybe it can be an old tourist / hiking guide from a time long past (with trees and lakes!)
Jacksinthe
Banned
Collisions aren't the issue. I just haven't found a reliable pixel-perfect system using colliders be it enter/exit or overlap. I built my own using Raycasts that's far more flexible.
Ex: nothing in my game is "wall" or "ground". At most I use a tag for moving platforms but I don't like parting out collision objects as I feel the controller should be smart enough to know a vertical or horizontal collision and slopes. Building this way, using colliders to detect overlap or enter/exit you can flip grounded by touching a wall in air - it's picky to point that the size of the collider in fractions will cause issues. This is why I prefer raycasts. Just easier to control.
My current issue is moving platforms since RigidBody doesn't work well when you slap them on things that move. My controllers have always been built with custom solutions and pseudo physics by translating the object vector in either world or local space as a kinematic object. I found this to be the most reliable way to handle any number of situations from platforms to slopes so I prefer that method. Just for giggles I'm trying RigidBody to kinda see what their limitations are. I use them for several enemies in MF1 just fine but those aren't controlled by the player and are strict when it comes to their functionality so they play by a much smaller set of rules.
I can just about do everything I need to, including the moving platform but I am just looking for a cleaner solution than the one I came up with. I much prefer Lerping their vector with the ability to apply Mathf.smoothStep when needed vs what feels like brute forcing them using a rigidbody and velocity while checking the square magnitude of their start/end positions to see if I shoot past them (I chose sqr so I can apply any vectors and any speed in any direction and have it work the overshoot).
It just seems rather clumsy to do it that way. Or I'm just incredibly dumb and haven't come up with a better solution
Edit: I suppose I can still lerp them and apply any change in vector to the player each frame to offset the new player position and ensure the player position gets updated after the platform but I would think that may not be bulletproof, either.
Not even sure if I want moving platforms in the side project as that removes some pacing from the player but I'm just trying it out to see how I can implement if needed. I try to cover all bases, tho.
Is Blender an option? Sorry if that's a scary name to mention.
Is Blender an option? Sorry if that's a scary name to mention.
I tried... To strange for me...
Do an 80s sci fi style neon glow
and your game looks really cool in that gif! where do I see more?
The game is set in a mad-max inspired universe of desert and cars
Love this
Does anyone have experience reaching out to websites to try and get coverage? I'm wondering what sort of wording to use in email when contacting them. Would it be considered a press release? Should I label it as such?
Also, if anyone knows any good websites or communities to advertise a story-driven 2D RPG, I would be extremely thankful to know!
Also, if anyone knows any good websites or communities to advertise a story-driven 2D RPG, I would be extremely thankful to know!
Jacksinthe
Banned
if you reach out to websites - keep the emails SHORT and link them to a press kit they can read on their own if your email entices them.
If it takes them longer than 30 seconds to read they probably will ignore it. They get tons of emails so the shorter the better.
That's a good point, thank you. How do people normally send out press kits? A downloadable zip?
Jacksinthe
Banned
You can either link them to your presskit() on your website, which is an online presskit or you create one for download.
I would also advise against sending any attachments unless requested. Just link to them where they can download or view your online presskit. There are do's and don'ts like don't inline screens in a PDF. Just include high-resolution screens, company logos, etc.
Presskit() is linked in the OP. Several sites have used its formatting since it's just XML to populate their pages with our game. It's a useful tool. Speaking of which, I should update mine http://press.absinthegames.com
Theanine3D
Member
Lolly Joe's store page is now live! The game can be added to your wishlist so you can be notified immediately when it becomes available.
RelentlessRolento
Member
#ScreenshotSaturday of an experiment I've been working on. Unsure yet if it will be a real deal or not.
I might be down for doing art at some point if you'd ever like to but I do like the crayon drawings.
Spoke to a local developer this week and am glad to find someone in a town that seems to have almost no indie devs. Feels less lonely.
Interesting.
I already know a bit of Blender (despite being a programmer instead of an artist). Do you think this tool could add something to my workflow?
I'm always looking for better ways to get assets.
BladeoftheImmortal
Member
Yea, anyone know a good place to get public assets?
I don't use it myself so I can't really say. I thought I'd mention the sale when I saw it since other people have recommended it.
TheHollowNight
Member
And the award to the most passive-aggressive email our team has ever received goes to...
Coherent!
It feels like I'm breaking up with them and they haven't got the message yet lol
AusRoachman
Member
Been inspired by all these no mans sky leaks so thought i would add abit more variation to some of my procedural maps.
Random Skybox / Colours / Floating objects (Clouds / Rocks)
Random Resources
Random Skybox / Colours / Floating objects (Clouds / Rocks)
Random Resources
Teknopathetic
Member
"I don't use it myself so I can't really say. I thought I'd mention the sale when I saw it since other people have recommended it."
While that's a bit of an old recommendation, there's probably worse ways of spending 13$ if you can tolerate the wonky glitches. Realistically, that's a bit much to ask of someone, because that could result in a lot of lost work and time. Thus, I probably wouldn't recommend Silo any more now that there's NVil which is roughly the same price (when Silo isn't on sale), doesn't have the bugs and glitches, and is in active development. Though it's interface isn't as nice as Silo's, it's still not as awkward as Blender's.
While that's a bit of an old recommendation, there's probably worse ways of spending 13$ if you can tolerate the wonky glitches. Realistically, that's a bit much to ask of someone, because that could result in a lot of lost work and time. Thus, I probably wouldn't recommend Silo any more now that there's NVil which is roughly the same price (when Silo isn't on sale), doesn't have the bugs and glitches, and is in active development. Though it's interface isn't as nice as Silo's, it's still not as awkward as Blender's.
Teknopathetic
Member
Tabs and pages on top of tabs and pages. While that's arguably a necessity because of the sheer capability of Blender, it always feels like everything I want (that's not a command I can simply type in) is a tab or a page away. Also the vertical tabs on the left hand side by default are the worst.
For example, when you look at the primitives under the create tab, each individual primitives has a bar button with the icon and the text and it results in a long scrolling menu with all of the options. Is there any reason why those aren't just small icons without the text and condensed to take up less space, similar to Modo or 3DS Max?
It's a lot of little things like that that dissuade me from using it. Especially when there are cheap alternatives. Not cheaper than Free, of course, and for everything Blender can do (and the massive community it has is also a major bonus) it's extremely impressive. I just don't want to use it.
For example, when you look at the primitives under the create tab, each individual primitives has a bar button with the icon and the text and it results in a long scrolling menu with all of the options. Is there any reason why those aren't just small icons without the text and condensed to take up less space, similar to Modo or 3DS Max?
It's a lot of little things like that that dissuade me from using it. Especially when there are cheap alternatives. Not cheaper than Free, of course, and for everything Blender can do (and the massive community it has is also a major bonus) it's extremely impressive. I just don't want to use it.
I don't know if I got you right, but I had a similar situation for some of
the effects I did a couple of month ago, where I switched form translation +
smoothstep to something more physical giving me more flexibility.
Well, I was looking at something more advanced than smoothstep to vary the
position of "objects" on the screen in some more sophisticated, non-uniform
ways. Meaning; I was in need of a user-defined function f(t) which varies in
non-uniform steps with t varying in equal steps. Sure, smoothstep and friends
are fine for many things but you have to adhere to the tangency condition
which basically defines how smoothstep behaves (blends) in-between (Hermite
interpolation). To get some interesting non-uniform varying positons I
switched to "rigidbody". Basically, I gave each objects a piece-wise velocity
function defining the velocity of the object for different sub-ranges of t.
Integrating out yields the non-uniform varying position as needed, also
matching the defined velocity along the way, for uniform steps of t.
Hence, being able to define the velocity in a piece-wise fashion gave me some
good control on how the objects move in-between, something you aren't in
control given many of the interpolation schemes. So basically, the physical
approach described here is a way to build your own blending function with you
being in control of how the function behaves in-between.
Don't know if this is of any help to you considering your "translation +
smoothstep" vs. "rigidbody" encountering.
Theanine3D
Member
Blender's great. I made every single model in my game with it, and I also prototyped all of my shaders in Cycles first before making them in Unreal. My skyboxes were prerendered in Cycles too.
As it goes with all software, when you've already spent years learning a certain program, it's so incredibly hard to get used to a completely different one. That's why Autodesk gets so much traction, they hook people in by offering free student licenses. Once they've learned 3ds Max or Maya for free, they're usually Autodesk customers for life, because most people are unwilling to shake away from what they know and try something like Blender.
I've always wanted to use GIMP as I prefer open source software alternatives whenever they're available, but every time I've gotten down and tried to learn it, I end up too frustrated. I'm just too used to Photoshop - I've been using PS since my early teen years. I probably would feel the same way about Blender as I do about GIMP, if I had learned Maya first. It's hard to kick old habits!
As it goes with all software, when you've already spent years learning a certain program, it's so incredibly hard to get used to a completely different one. That's why Autodesk gets so much traction, they hook people in by offering free student licenses. Once they've learned 3ds Max or Maya for free, they're usually Autodesk customers for life, because most people are unwilling to shake away from what they know and try something like Blender.
I've always wanted to use GIMP as I prefer open source software alternatives whenever they're available, but every time I've gotten down and tried to learn it, I end up too frustrated. I'm just too used to Photoshop - I've been using PS since my early teen years. I probably would feel the same way about Blender as I do about GIMP, if I had learned Maya first. It's hard to kick old habits!
TheHollowNight
Member
Finally locked down how I wanted the Chilopods to sound!
https://youtu.be/Zc26w57sLs0
Still got a couple of sounds to make, like the end roll animation, and a few other barks and impact sfx, but I'm pleased with the progress.
https://youtu.be/Zc26w57sLs0
Still got a couple of sounds to make, like the end roll animation, and a few other barks and impact sfx, but I'm pleased with the progress.
Thanks!
as a matter of fact I just finished all basic game systems! (Is my game in alpha now?)
Here is a basic video that briefly touches on
- world map / view
- dialog / event system
- normal gameplay (combat)
- inventory / looting
-> check it out!
https://www.youtube.com/watch?v=tvRTHaPNTyk
About the game - Bounty Road is a party RPG set in a distant future, where brave warriors fight in the desert heat. Haha.
Do an 80s sci fi style neon glow
and your game looks really cool in that gif! where do I see more?
Since you asked I think it's ok to post some links
Most story stuff here:
https://forums.tigsource.com/index.php?topic=55376.0
Most tech stuff here (not only Bounty Road)
https://kosmonautblog.wordpress.com/
and everything here:
https://twitter.com/kosmonautgames
I am a 1 man, hobby dev, so yeah, don't expect the world
followed
man that dust effect is out of this world!
Thanks
Here is a breakdown
And here is a blog article about the shadows : https://kosmonautblog.wordpress.com/2016/04/10/smooth-shadows-for-geometry-dust-clouds/
Note that I disabled the soft smoke in the video, because i was testing some stuff. Meh.
AusRoachman
Member
Thanks
Here is a breakdown
And here is a blog article about the shadows : https://kosmonautblog.wordpress.com/2016/04/10/smooth-shadows-for-geometry-dust-clouds/
Note that I disabled the soft smoke in the video, because i was testing some stuff. Meh.
Was worried you had scrapped it as you had gone quiet of late.
Glad to see the project is still going , such a cool setting and gfx are looking awesome .
Thank you again! That really helps a ton!
Where you seeing that from? The only thing on their roadmap that is slated in an actual Unity version (5.5 in this case) is a compiler update. But it explicitly says in the notes that it is not updating the runtime version.
Please only take this if you actually want to give it a try.
Oh, and report back to the thread your impressions!
Oh, and report back to the thread your impressions!
ModBot said:Instructions for participants:
I am giving away a Steam Gift. To enter this giveaway, send a PM to ModBot with any subject line. In the body, copy and paste the entire line from the message below containing the game you want to enter for. Confused? Watch this GIF tutorial or ask for help.
Want to make your own ModBot giveaway? Click here for a quick tutorial thread. Please give generously.
ModBot Basics:
- This is a giveaway for the thread url: http://www.neogaf.com/forum/showthread.php?t=965454.
- I really appreciate thank you messages, but please send them to me (Popstar, not ModBot!) via PM instead of in thread.
- Do not trade keys you win off-site to enrich yourself. Don't try to claim games you have no interest in collecting or playing. Don't claim games to give them to friends off-site.
- If the key is already taken you will not receive a reply. Replies may take a minute or two.
Rules for this Giveaway:
- If you are a lurker--if you have both fewer than five posts in this thread--you are not eligible for this giveaway.
- This giveaway is a raffle. The winners will be selected by random draw 24 hours after the draw was created. Any games not claimed after that point will be given away first come first serve.
Silo 2 --
MB-97BD8D7732E8C06A- Taken by mStudios. 2 entrants total.
Secret beta build
http://forum.unity3d.com/threads/upgraded-c-compiler-on-5-3-5p8.417363/
Definitely makes a difference
http://forum.unity3d.com/threads/holy-moly-alloc-appreciation-thread.417765/
TheHollowNight
Member
Messing around with some composition shots for a few 'vista' spots. Spent about 30 minutes constructing this.
Hey guys, I'm back after being a bit quiet recently.
The reason for this is, well, that big changes have been made to the project I'm working on. And by big changes I mean that we did a major overhaul to the concept and naming of said project. As of such, may Pulsar Raiders rest in peace, here is...
We also have a new trailer, which looks way better than what Pulsar Raiders ever did and now looks like something what I would consider buying off of Steam/PSN/Xbox Live (finally, it took us a while).
https://www.youtube.com/watch?v=ikZ7Y-O6HEY
Our store page on Steam has finally been accepted too (with a bunch more screenshots and info), and we've updated our IndieDB page and created a new Facebook one.
Hope you guys enjoy it!
The reason for this is, well, that big changes have been made to the project I'm working on. And by big changes I mean that we did a major overhaul to the concept and naming of said project. As of such, may Pulsar Raiders rest in peace, here is...
We also have a new trailer, which looks way better than what Pulsar Raiders ever did and now looks like something what I would consider buying off of Steam/PSN/Xbox Live (finally, it took us a while).
https://www.youtube.com/watch?v=ikZ7Y-O6HEY
Our store page on Steam has finally been accepted too (with a bunch more screenshots and info), and we've updated our IndieDB page and created a new Facebook one.
Hope you guys enjoy it!
Jacksinthe
Banned
This is what I'm leaning towards. I can control the rate of acceleration and deceleration much better on platforms.I don't know if I got you right, but I had a similar situation for some of
the effects I did a couple of month ago, where I switched form translation +
smoothstep to something more physical giving me more flexibility.
Well, I was looking at something more advanced than smoothstep to vary the
position of "objects" on the screen in some more sophisticated, non-uniform
ways. Meaning; I was in need of a user-defined function f(t) which varies in
non-uniform steps with t varying in equal steps. Sure, smoothstep and friends
are fine for many things but you have to adhere to the tangency condition
which basically defines how smoothstep behaves (blends) in-between (Hermite
interpolation). To get some interesting non-uniform varying positons I
switched to "rigidbody". Basically, I gave each objects a piece-wise velocity
function defining the velocity of the object for different sub-ranges of t.
Integrating out yields the non-uniform varying position as needed, also
matching the defined velocity along the way, for uniform steps of t.
Hence, being able to define the velocity in a piece-wise fashion gave me some
good control on how the objects move in-between, something you aren't in
control given many of the interpolation schemes. So basically, the physical
approach described here is a way to build your own blending function with you
being in control of how the function behaves in-between.
Don't know if this is of any help to you considering your "translation +
smoothstep" vs. "rigidbody" encountering.
Another weird issue I am running into with Rigidbody is collision/groundchecks.
Trigger colliders are too unreliable and checking an OverlapArea produces false positives when bumping a wall in mid air. I'm not about to mark walls or floors since corners can have both making the distinction moot when both intersect - so I chose to raycast as I usually do.
The issue I see now is that on extreme corners where I see a collider not touching a corner, the raycast produces no hit but I am still standing on top of the edge, thusly removing any actions I can take while I should be grounded. I'm curious if adjusting the skin width will help.
For my custom controller I create my own skin widths for colliders as starting points for the rays and have created exceptional corner detection using my own methods but rigidbody just seems slightly awkward in this case. Kinematic bodies can clip colliders which is fine - which is why I prefer using the casts but I was REALLY trying to roll simpler solutions using a Rigidbody instead of my currently implemented controller for MF1.
I will probably rewrite my methods from scratch, looking at how I handle that character controller I can probably write one slightly more efficient.
There's a zillion other cheats I can employ with Rigidbody like velocity detection and using some rays to dictate collision left/right/up/down based on Y velocities and work proper states this way but I was just hoping some quick rigging of rigidbody objects would just "work" using some simple tools.
I can still probably make it work but not directly out of the box. I should probably rewrite a controller from scratch using my old methods, but slightly more robust and optimized and GitHub it for everyone although that does slightly defeat the purpose of learning for some. Perhaps some lead-in and fill in the blanks, leaving room for customization.
I've been interested in game design for a long time and I think I'm finally gonna push myself to commit to doing something. I really want to make a tactics style game. I downloaded Blender and Unity today (I've dabble with both in the past in a minor way) and I've been messing around for a few hours.
It's really nothing yet, but after I finally got everything working the way I wanted to I am unusually proud of this:
Just having an idea for something I wanted to accomplish today (moving a "unit" around on a grid) and then being able to implement it is a cool feeling.
Next I want to think about basic pathfinding, so that when I click on a tile it draws the path the unit will take and change the color of the blocks along that path, and then changes them back to normal as the unit moves past them.
It's really nothing yet, but after I finally got everything working the way I wanted to I am unusually proud of this:
Just having an idea for something I wanted to accomplish today (moving a "unit" around on a grid) and then being able to implement it is a cool feeling.
Next I want to think about basic pathfinding, so that when I click on a tile it draws the path the unit will take and change the color of the blocks along that path, and then changes them back to normal as the unit moves past them.
I know starting with a big project isn't a great idea, but this is something I've been pretty passionate about the past few days so I gotta go with that, right?!
- Status