that makes sense thanks. About input, should it be updated every frame and accumulated for the logic update?
-
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.
Indie Game Development Discussion Thread | Of Being Professionally Poor
- Thread starter chubigans
- Start date
- Status
that makes sense thanks. About input, should it be updated every frame and accumulated for the logic update?
IMHO input should be checked on the fixed frames (logic update), not on variable frames (graphics update). There's no reason to accumulate the input for later use, check it just before you actually intend to use it. The only thing you should be doing at the graphic framerate is interpolation and other things related to rendering, everything else should be on the fixed logic framerate, including polling for input. You'll thank yourself later if you try implementing network multiplayer or any kind of replay system!
The only reason I could think of for polling input on the variable framerate is if your fixed framerate is really low.
Hey all! Quick question for you all. I want to try out making something in Unity for iPad, but I don't have a mac (just an iPad).
Is there a way to do this without having to buy a mac? Apparently Unity only allows you to build an iOS version in the OSX version of Unity, so couldn't I just install OSX on my windows machine and install Unity on the OSX side of the machine? Is there a way to do that, similar to bootcamp? Anyone have any experience with this? I want to make sure it's possible before I start installing a ton of stuff.
Is there a way to do this without having to buy a mac? Apparently Unity only allows you to build an iOS version in the OSX version of Unity, so couldn't I just install OSX on my windows machine and install Unity on the OSX side of the machine? Is there a way to do that, similar to bootcamp? Anyone have any experience with this? I want to make sure it's possible before I start installing a ton of stuff.
There is something called Hackintosh but you will need a OSX Licence and then i am not sure if this is legal at all. Furthermore you will need a Dev Licence which cost 100$.
If you dont want to spend much money, then just get a cheap Android Tablet or use your (if you have one) existing Android Smartphone.
I installed OSX on a PC once, about three years ago. I don't remember how to do it, but I remember that it wasn't a pleasant experience. The OS was missing a lot of little things, like drivers for the screen and internet adapter. Chances are that drivers for your hardware do not exist, since it's not expected to be needed. So I had a 800x600 res, no-internet piece of crap for about three days before deciding to undo my frankensteined Hackintosh.
It's possible that things have improved in the mean time, but from my experience I can't recommend it.
I once asked a tech wiz that's a friend of my dad's about that, and he directed me to this page: http://www.unibeast.com/ . There's also a step-by step explanation here: http://www.tonymacx86.com/374-unibeast-install-os-x-mavericks-any-supported-intel-based-pc.html
I haven't looked much into it, though, but I guess I eventually will since publishing on iOS would be a huge boon.
EDIT: Like Gundalf said, this does require an OSX license, so I don't know how legal it'd end up being if you'd probably end up using a single installation/license for both the Mac you downloaded it from and the PC as well.
I am polling input every frame.
(HTML5 game, though).
The only question is whether I actually use the data of the input states or not.
If there are things already set in motion, then obviously that input will not be used. But as soon as I have an idle game state (i.e. where a player can make a difference with input), it better react to it the first frame available - in this case, that is a ~16ms input polling.
(HTML5 game, though).
The only question is whether I actually use the data of the input states or not.
If there are things already set in motion, then obviously that input will not be used. But as soon as I have an idle game state (i.e. where a player can make a difference with input), it better react to it the first frame available - in this case, that is a ~16ms input polling.
tonymacx86.com
Possible, but you really need to have the right hardware for it to be a good experience.
Thanks guys! I'll look into those options, but I think my best bet may be to develop for android phone, and then once I have the game working pretty well, make the decision if it is worth it to get a mac to port to iOS.
I have some friends who have macs, do I need access to a mac every time I want to run the game, or is it just for exporting a final build?
I have some friends who have macs, do I need access to a mac every time I want to run the game, or is it just for exporting a final build?
You can't make an iOS build of the game at all on PCs, since Unity has to do some cross-compilation through OSX's XCode. Also, I think you can't create your own keystores like in Android, so you have to pay up and become an authorized developer before you get to do any iOS builds.
The fastest button presser in the world could do 16 presses a second. Thus as long as your fixed framerate is at least 30fps, the chances of missing a button press while polling during the logic update is extremely unlikely. It should only be a problem when for some reason the game is running slow, and at that point your variable framerate is also going to be delayed as well.
Unless you have an unusually slow fixed framerate for your logic/physics updates, I just can't see a compelling reason to poll input outside of your fixed logic frame updates... but I'd like to hear the reasoning for it if someone feels strongly otherwise!
I got myself an android phone to start testing my game. I had to correct only two simple things and the game was running. I confess I was a little bit surprised. I have some java -- c++ mumbo jambo going on I was fearing that would give me trouble.
Next step: polish the game, test it, add sound and it's ready!
Next step: polish the game, test it, add sound and it's ready!
Be aware that a hackintosh has to be really close to actual mac parts to work, and if a part doesn't match, you'll get random stuff that doesn't work.
Finally have a good idea and a decent plan. Resurrecting a movie idea from 10 years ago (where we filmed half of it and storyboarded the whole thing). Using the old footage and story boards for cutscenes and the story idea for the levels and what not.
Thanks for the Unity suggestion - started on the engine tonight!
Thanks for the Unity suggestion - started on the engine tonight!
Good point, and I'll be checking input at 60 fps, so it should be fine.
Since it's now Saturday. Just completed a new map for my tower defense game. I call it Twin Souls. Call it that as they rotate around each other with bots rotating around the bases themselves as well as the outer circle too.
Also made it so if one base is destroyed it will completely throw off the rhythm. The other base will stop in it's place, it's spinning disk of 2 bots will slow down. And the outer circle/disk will come to a complete stop and act like a "broken" clock. Still have some work to do balancing the map. Cause of the enemy speed, need to give player lots of starting money.
Also made it so if one base is destroyed it will completely throw off the rhythm. The other base will stop in it's place, it's spinning disk of 2 bots will slow down. And the outer circle/disk will come to a complete stop and act like a "broken" clock. Still have some work to do balancing the map. Cause of the enemy speed, need to give player lots of starting money.
holy crap I just looked at my game's server list
that's a lot of fuckin servers goddamn
What game is this?
some game about burgers or something I dunno http://kritz.net/games.php
Game's in a pretty rough open alpha. To be honest? Not sure how the game's gonna work out for me, considering the game is free to play as an alpha, but at some point I kinda do intend on selling the game (or selling stuff inside the game and keeping the base free, who knows).
... anyone have experience with greenlighting a f2p game?
Game's in a pretty rough open alpha. To be honest? Not sure how the game's gonna work out for me, considering the game is free to play as an alpha, but at some point I kinda do intend on selling the game (or selling stuff inside the game and keeping the base free, who knows).
... anyone have experience with greenlighting a f2p game?
Input is tricky, and well worth some study!
One problem is "presses". When buttons go from not pressed, to pressed. The other is poll distribution - trying to achieve a regular time between polls, for a consistent feel.
If your game is running slowly, or you set the fixed time step to something tiny, you may generate more than one keyframe in one loop of your game. They are generated right next to each other, so the input is most likely going to be the same. It's okay for presses, but not so great for being able to do 1-frame actions like in Street Fighter for example. Disaster would be if your input reported a 'press' for all the frames you generate in a single iteration of your loop, (which will happen if you just poll at the start of the game loop!).
When your game is running really fast you get loops of your game which don't require new keyframes to be generated. Buttons could be pressed and released in that time, and you may miss them!
I find the best system is to track input in the main loop of your game and consume logged presses and releases in the key frame generation. So, the keyframe generator gets a 'press' and cancels it from the system, so it doesn't see it twice.
If this kind of thing is crucial to your game, like in the example of Street Fighter, then bin the keyframe thing altogether. Lock to 60 and poll the input at a regular 1/60th. Otherwise your game will never have a consistent feel. It's no good writing a game that responds to input differently on different machines!
In something like a driving game it's not important as input is much slower.
Hope that helps!
its very helpful!Hope that helps!
the game/project is a top-down or third person turn based strategy. Discrete turns are played out in semi real time(you spend "ap" for movement/attacks/etc, and animations play out as you issue commands), maybe even with some fully interactive parts like the aiming in valkyria chronicles.
Perfectly consistent input is probably not very important in this kind of game. I'd still like it to feel smooth and responsive though. Polling in the "update" seems like the best choice.
I guess what confused me a bit is how to properly control and update the camera with a fixed timestep loop. somehow I felt it should get some special treatment but it looks fine being treated just like any other object so I'm probably overthinking it.
as for the intricacies of input.. every time I decide to write a proper context sensitive input system I end up procrastinating.
Mystic River
Member
Another week is gone, and accomplish a lot lately. This is what I've been working on the past few days...
Just docking my ship to a new redesigned space station for repairs, trading and other things.
Also, you can upgrade or sell modules, sometimes you may be able to recover debris or modules from defeated enemies, plug em into your ship and sell them in your next stop and get some quick cash, in this screen you will be able to sell your excess modules.
Just docking my ship to a new redesigned space station for repairs, trading and other things.
Also, you can upgrade or sell modules, sometimes you may be able to recover debris or modules from defeated enemies, plug em into your ship and sell them in your next stop and get some quick cash, in this screen you will be able to sell your excess modules.
Another week is gone, and accomplish a lot lately. This is what I've been working on the past few days...
Just docking my ship to a new redesigned space station for repairs, trading and other things.
Also, you can upgrade or sell modules, sometimes you may be able to recover debris or modules from defeated enemies, plug em into your ship and sell them in your next stop and get some quick cash, in this screen you will be able to sell your excess modules.
looking pretty great man -- I admit I don't know a lot about this game yet. is it a space roguelike (spacelike)?
psst if it's saturday you should post it to the screenshot saturday thread.
Yeah, you should be fine then.
And yep, everything's the same. If it moves, keyframe it and interpolate for display.
Congrats, you've leveled up!! ;D
UI development
Resizing a window is now also possible by dragging its edges. I further worked
on the states of each element -- which can be quite cumbersome at times.
For example, when pressing a button it should also turn into hover state
(cursor hovering the button) if the cursor leaves the button with the mouse
still holding down/pressed. Moving the cursor back again over the button
(while still holding the mouse down) should change the button's state back to
pressed. Now since I allow to press a button without bringing the framed
window atop, which is cool, the button should also change from pressed into
hover state (mouse button still down) if the cursor runs into another window
on top of the button, yet without turning the window atop into hover state
while keeping the hover state for the window below even if the cursor is above
another window. See the animation above.
A similar issue occurs for a frame window with a client area and another such
window partially covering it, in determining the correct hover state. Anyhow, I
got it all working without compromising the ui-elements.
All the oddities explained above happen without any ui-element knowing about
it. Every ui-element can have a specialized mouse handler (next to a standard
one) which does the specific action. Additionally, the handler can be changed
at runtime allowing to reconfigure the element's mouse behavior on-demand. I
want to do something similar with the callback system of each element. I want
to make it possible to be able to reconfigure an element's callback system at
runtime. This makes it possible to have multi-function buttons which can
change their behavior on-demand. For example, you may press a nearby (slider-)
button who will switch to another callback system for the button in
question, or of an entire group of buttons. I have some ideas using this for
the spectral analyzer tool. An obvious need is also for debugging. The entire
callbacks may be routed through a debugging hub displaying additional
information. All on a per ui-element basis. Sound too good... xD
Resizing a window is now also possible by dragging its edges. I further worked
on the states of each element -- which can be quite cumbersome at times.
For example, when pressing a button it should also turn into hover state
(cursor hovering the button) if the cursor leaves the button with the mouse
still holding down/pressed. Moving the cursor back again over the button
(while still holding the mouse down) should change the button's state back to
pressed. Now since I allow to press a button without bringing the framed
window atop, which is cool, the button should also change from pressed into
hover state (mouse button still down) if the cursor runs into another window
on top of the button, yet without turning the window atop into hover state
while keeping the hover state for the window below even if the cursor is above
another window.
See the animation above.A similar issue occurs for a frame window with a client area and another such
window partially covering it, in determining the correct hover state. Anyhow, I
got it all working without compromising the ui-elements.
All the oddities explained above happen without any ui-element knowing about
it. Every ui-element can have a specialized mouse handler (next to a standard
one) which does the specific action. Additionally, the handler can be changed
at runtime allowing to reconfigure the element's mouse behavior on-demand. I
want to do something similar with the callback system of each element. I want
to make it possible to be able to reconfigure an element's callback system at
runtime. This makes it possible to have multi-function buttons which can
change their behavior on-demand. For example, you may press a nearby (slider-)
button who will switch to another callback system for the button in
question, or of an entire group of buttons. I have some ideas using this for
the spectral analyzer tool. An obvious need is also for debugging. The entire
callbacks may be routed through a debugging hub displaying additional
information. All on a per ui-element basis. Sound too good... xD
legacyzero
Banned
Saturday!
Just got a simple pick-up-and-play game finished and released on Android today called flow roller. I was wondering if people had some critique I could use in improving the game further or any impressions, haven't had a lot of playtesting done yet and could use some more opinions! Game is free to play and the music track was done by the wonderful and talented sadsic
https://play.google.com/store/apps/details?id=com.emicogames.flowroller
I'm seeing lots of simple to make games popping up on smartphones these days, anyone think this is an actual trend right now or is it just something that's been there all along? Depending on how well this is received I might just continue doing these type of games since it doesn't take long to finish (in under a month basically) and I can actually finish them instead of working on something until I get a burnout and just shelve it along with other unfinished projects, it really feels good to have something out for once. I'm sure plenty of you guys have unfinished projects just sitting there as well
https://play.google.com/store/apps/details?id=com.emicogames.flowroller
I'm seeing lots of simple to make games popping up on smartphones these days, anyone think this is an actual trend right now or is it just something that's been there all along? Depending on how well this is received I might just continue doing these type of games since it doesn't take long to finish (in under a month basically) and I can actually finish them instead of working on something until I get a burnout and just shelve it along with other unfinished projects, it really feels good to have something out for once. I'm sure plenty of you guys have unfinished projects just sitting there as well
I made a video of that map I mentioned earlier today. Really like the flow of it. Esp when it gets disrupted.
https://www.youtube.com/watch?v=BCJGrlS5n28
SeanNoonan
Member
Probably my dullest #screenshotsaturday yet, but hey, the game needed a pause menu...
However I did remake my character sprite in Resogun for a laugh.
However I did remake my character sprite in Resogun for a laugh.
SlipperySlope
Banned
I have a silly-stupid Unity question. I'm using the free version. I'm making a pixel art game with a low resolution of 384x216. The goal is to have the screen rendered at this resolution, then upscaled to whatever the device is using. This will keep the pixel art visual style of the consoles of old.
Doing this in Unity does not seem to be overly easy. RenderTexture seems to be the only way? But that requires Unity Pro. I'd rather not have to upgrade to Unity Pro for something silly like this.
Any advice?
Doing this in Unity does not seem to be overly easy. RenderTexture seems to be the only way? But that requires Unity Pro. I'd rather not have to upgrade to Unity Pro for something silly like this.
Any advice?
SlipperySlope
Banned
You can do a PC release without Pro.
But IOS, Android, consoles, etc. all need Pro and each needs a separate license fee if my understanding is correct.
Edit - IOS can be especially difficult because you need MacOS and XCode - and likely a Mac computer which I don't have.
Dynamite Shikoku
Congratulations, you really deserve it!
iOS and android are free too. But you can't use pro features of course.
Teknopathetic
Member
"I have a silly-stupid Unity question. I'm using the free version. I'm making a pixel art game with a low resolution of 384x216. The goal is to have the screen rendered at this resolution, then upscaled to whatever the device is using. This will keep the pixel art visual style of the consoles of old."
You don't need to use RenderTextures, you don't need to use a low resolution to have a pixel art style. Set the orthographic camera to the right size so that your sprites will be "pixel perfect." In this case, you'd use 216 as your target height. Also make sure to set all of your sprites filter modes to "point" so no filtering is applied. Everything will automatically scale up to higher resolutions, though you will need to keep in mind displays with different aspect ratios will have more visible, but you can just put some kind of overlays "outside" of the camera to block what you don't want to be visible (think stuff Backbone Entertainment's various Capcom ports).
Then you can probably just do some math in script to snap moving sprites to each pixel, otherwise you'll get that icky scrubby looking movement "between" pixels.
You don't need to use RenderTextures, you don't need to use a low resolution to have a pixel art style. Set the orthographic camera to the right size so that your sprites will be "pixel perfect." In this case, you'd use 216 as your target height. Also make sure to set all of your sprites filter modes to "point" so no filtering is applied. Everything will automatically scale up to higher resolutions, though you will need to keep in mind displays with different aspect ratios will have more visible, but you can just put some kind of overlays "outside" of the camera to block what you don't want to be visible (think stuff Backbone Entertainment's various Capcom ports).
Then you can probably just do some math in script to snap moving sprites to each pixel, otherwise you'll get that icky scrubby looking movement "between" pixels.
Mystic River
Member
thanks!, in fact it is a space roguelike game, random events, procedural things like huge procedural derelicts randomly generated to be explored for loot and knowledge , procedural guns, procedural crew , procedural universe and many more things, so hopefully you won't be playing the same game in your life time
, something like FTL but you will be able to pilot your ship or let the AI pilot it while you take care of the inner stuff like boarding other ships trying to save the universe. Build or upgrade your ship, all ships are made of small modules so you can shape your ship in many different ways and customize it, I will let player build their ships in a very flexible way so if you want your ship to look like a millenium falcon you will be able to do that, or like the Normandy in mass effect, so basically,I give the tools and let the users build the things they want to fly, but there is a catch, all is based on physics, so strange shapes will be harder to maneuver than standard ship shapes in space so be careful where you put your ship weight and distribute your thrusters, I really don't know for sure what shapes will have more problems than others yet since I am still building many more ships and see how physics affect my designs, having a lot of fun doing that now.Thanks for the link! I will post next saturday since it is already Sunday
Here's a #screenshotsaturday screen for my game, Turnover. Just running from some molotov fire.
What a better way to demonstrate progress if not with a progress bar?
Nice. Are you writing that UI yourself? The idea of doing UI stuff makes my chest heavy.
I really like these sort of stealth games! Keep going!! :+
Entirely. A couple of pages ago I wrote about how the UI will evolve. There
will be at least one version which can be run from a bootloader (PC and PS3
for example) before kicking off using more sophisticated (hardware
accelerated) techniques. Hence, the primary focus of the UI is to be platform
independent, initially, as well as being independent of any rendering backend,
i.e. Set{Pixel/Character} should be enough. Problem with doing a UI from scratch
is that you have to do a lot of work yet get very little in return, initially.
But let me tell you; once you are able to combine your elements, the curve
goes exponentially. :+ Personally, I don't have much problems going through
this lean period of not having much boombastic stuff on the screen simply
because I have an auto-intrinsic interest in doing what I do.
SlipperySlope
Banned
Oh wow. Perfect! Thank you sooo much.
SlipperySlope
Banned
Nice. What platform is it designed in? Are you planning to sell it?
Are you talking about pixel perfect movement? If so he would have to accumulate the fractional until it is a whole number and then use that for the velocity to get pixel perfect movement. I forget how to exactly do it because haven't played with it in 2 years since I stopped using XNA.
I spent some time with your game, and in that time I managed to cook a bunch of items until they were completely black, punch the shit out of ingredients, cook money, put a rat in a burger, shove a customer completely out of the store, and watch a rat grab money and run out of the front door.
I think I just found my GOTY.
Looking for some good written Unity tutorials for just starting out? I worked in Unity about 4 years ago to make a simple game, and I would like to relearn. Back then it had a really simple scripting language. I was wondering if it still has the same as Im not amazing at programming
EDIT: I want to start with scripting and playing around with movement before doing anything graphics wise.
EDIT: I want to start with scripting and playing around with movement before doing anything graphics wise.
- Status