What are you working on? - Developers

[QUOTE=Pat.Lithium;49634793]This game jam is fun, 1 day in game is comong along nicely. Slept on the couch in the office its set up in. Cant wait to go home in 24 hours[/QUOTE] I wish mine was fun. We're ~12 ish hours in and we're struggling to maintain our group, it's in shambles. One guy has taken over the story and made it ridiculous, he insists a point-and-click with this Unity package called Adventure Creator which literally strips all form of using C#. I can't wrap my head around this and neither can the other developer. We both know Unity but we're literally tied to this as the "leader" is one of the developers, and expects us to pick up his knowledge of the package to his level in this very short time frame we've got. Development Hell ¯\_(ツ)_/¯ [editline]30th January 2016[/editline] Oh yeah, there's three developers and when I suggested source control or paired programming: "We can just do it from my laptop" man.

if you guys nap or try to sleep you'll probably hear "scope creep" in your nightmares speaking of "scope creep"- at the boeing project management seminar thingy I went to a while back I could swear that one of the engineers talking about his failed projects downed to scope creep looked like the [URL="https://media.giphy.com/media/Hy1U5snVwWAvu/giphy.gif"]vietnam flashback dog[/URL]. That entire seminar was just "DON'T LET SCOPE CREEP HAPPEN" :v:

[QUOTE=paindoc;49633174][...] mmm bb. function is in such fancy font too ~[/QUOTE] I wish I wasn't so bad with font editors, since that one programming font with ligatures got me thinking. I'd still use [URL="http://www.fontsquirrel.com/fonts/merriweather"]Merriweather[/URL], but add ligatures for [code]== ≈ != ≠ <= ≤ >= ≥ => →[/code] (though the last one isn't quite correct, since the font doesn't have the right arrow with that little vertical dash at the beginning). I'd also like to add a ligature [code]{ } ∅[/code], but that symbol is completely unavailable in this font and I highly doubt I could add it.

So I've been in a programming rut for about half a year now. I have a mostly working road generator which I've been trying (unsuccessfully) to expand to generate buildings, or at the very least land plots for future buildings. This was going to go towards a procedurally generated "game" that I had an idea for. The whole concept was basically a top down RPG, which had shooter elements to it. So the guns, the armor, etc would all be randomly generated, but every unit would also be generated so you have people who joined different factions for different reasons, just a game with a huge scope which I knew would be very hard to model. Think Dwarf Fortress but 3D and modernistic. It was going to be set in a post apocalyptic snow environment, because I thought it would be super cool to have snow falling, using snow tracks, having the atmosphere be a factor of staying alive. Some of these elements would be borrowed from my favorite games, like Fallout and Stalker. The reason I put game in quotation marks is because I thought, at least in version 1, you would only watch the AI play with itself. You didn't actually interact with any units, but you could see what they were doing, what their motivations are, etc. I thought it would be very fun to code all of this emergent gameplay, and I still think it is. The problem with all of this is even getting started. Generating buildings has been such a huge pain, and because I was less experienced than I am now with when I started the project, I have refactored many components so often that is has become more of a chore than a hobby. I was going to switch off from world generation to actual gameplay, given how uninterested I've become with that aspect, but I just feel like there is a mountain of work. I would have to get PhysX working again (which was a pain the first time around in another project), then I have to create 3D models, which I am complete ass at, and don't get me started on trying to get animations to work (which I never have accomplished), or even font rendering. At this point I'm thinking about whether this whole project is even feasible. The only reason rougelikes have so much procedural content is because it's easy to create cool names for swords you never have to model, and make intricate descriptions that don't look like crap because your imagination is working for you. Hell, even making 3D cubes shoot at eachother seems like an insurmountable task that will be entirely discarded once I get real human models holding real guns. I don't know what to do really, the barrier of trying to program again is just so high right now. My school assignments are easy and at least they keep me in touch, but since we can only use C++03 I weep every time I have to manually delete. Sorry about this huge rant, I don't really know what I meant to accomplish with it. I guess just getting my frustrations out is nice.

[QUOTE=WTF Nuke;49635994]So I've been in a programming rut for about half a year now. I have a mostly working road generator which I've been trying (unsuccessfully) to expand to generate buildings, or at the very least land plots for future buildings. This was going to go towards a procedurally generated "game" that I had an idea for. The whole concept was basically a top down RPG, which had shooter elements to it. So the guns, the armor, etc would all be randomly generated, but every unit would also be generated so you have people who joined different factions for different reasons, just a game with a huge scope which I knew would be very hard to model. Think Dwarf Fortress but 3D and modernistic. It was going to be set in a post apocalyptic snow environment, because I thought it would be super cool to have snow falling, using snow tracks, having the atmosphere be a factor of staying alive. Some of these elements would be borrowed from my favorite games, like Fallout and Stalker. The reason I put game in quotation marks is because I thought, at least in version 1, you would only watch the AI play with itself. You didn't actually interact with any units, but you could see what they were doing, what their motivations are, etc. I thought it would be very fun to code all of this emergent gameplay, and I still think it is. The problem with all of this is even getting started. Generating buildings has been such a huge pain, and because I was less experienced than I am now with when I started the project, I have refactored many components so often that is has become more of a chore than a hobby. I was going to switch off from world generation to actual gameplay, given how uninterested I've become with that aspect, but I just feel like there is a mountain of work. I would have to get PhysX working again (which was a pain the first time around in another project), then I have to create 3D models, which I am complete ass at, and don't get me started on trying to get animations to work (which I never have accomplished), or even font rendering. At this point I'm thinking about whether this whole project is even feasible. The only reason rougelikes have so much procedural content is because it's easy to create cool names for swords you never have to model, and make intricate descriptions that don't look like crap because your imagination is working for you. Hell, even making 3D cubes shoot at eachother seems like an insurmountable task that will be entirely discarded once I get real human models holding real guns. I don't know what to do really, the barrier of trying to program again is just so high right now. My school assignments are easy and at least they keep me in touch, but since we can only use C++03 I weep every time I have to manually delete. Sorry about this huge rant, I don't really know what I meant to accomplish with it. I guess just getting my frustrations out is nice.[/QUOTE] I, too, wanted to make a fully procedural city. I switched to a static world because I realized after a while that I was going to spend too much time on the generator when my most developed ideas revolved around gameplay that didn't depend on a procedural environment. I'd just be wasting my time for little payoff and probably a lot of manual tweaking and bugfixing. What kinds of models do you need? If you're considering continuing your project, I would be happy to contribute a few models as I keep refining my skills. I can make a [I]mean[/I] toilet.

Thanks for that. But even having actual models seems daunting. Having a real 3D representation of what I would procedurally generate is just such a departure from the core of procedural generation, you know? Like, two guns that look the same but perform entirely differently? Who would want to see that? I don't know, I'm going to go to sleep. Just feeling drained and I haven't even lifted a finger.

[QUOTE=WTF Nuke;49636137]Thanks for that. But even having actual models seems daunting. Having a real 3D representation of what I would procedurally generate is just such a departure from the core of procedural generation, you know? Like, two guns that look the same but perform entirely differently? Who would want to see that? I don't know, I'm going to go to sleep. Just feeling drained and I haven't even lifted a finger.[/QUOTE] At some point you need to set your expectations toward something more realistic. I think we all have ideas for things we'd like to work on that we simply can't do because they aren't feasible projects, I know I've personally dealt with "AAA syndrome" a lot, and have published exactly 0 games as a result. I think it would be better for you to try and pick out the ideas and aspects from your project that you CAN get done (i.e. certain gameplay aspects, story elements, etc.), prototype them with primitives and programmer art, and then try to recruit artists once you've done the heavy lifting to let them know you're serious. If you keep ramming your head into a wall trying to make a million dollar game during your lunch breaks and weekends, you're just going to go insane. Work with what you have, dial it back a bit and work on the concept rather than the assets. Who knows, you might make the next big indie hit.. [editline]30th January 2016[/editline] Keep in mind that, to my knowledge, there have been close to zero large-scale, high quality games released by independent, singular developers. It takes a team to build a game of a certain scale, and no amount of Maya tutorials will change that. The fastest way to attract a team is to prototype and demonstrate your ability as a [I]programmer[/I] and give artists a reason to help you. It's not a personal shortcoming of yours being unable to do the jobs of 10-20 highly paid professionals with years of experience in their respective fields.

[vid]https://zippy.gfycat.com/InsecureInexperiencedEastsiberianlaika.webm[/vid] VR duck hunt for Global Game Jam tee hee

[QUOTE=Silentfood;49635546]I wish mine was fun. We're ~12 ish hours in and we're struggling to maintain our group, it's in shambles. One guy has taken over the story and made it ridiculous, he insists a point-and-click with this Unity package called Adventure Creator which literally strips all form of using C#. I can't wrap my head around this and neither can the other developer. We both know Unity but we're literally tied to this as the "leader" is one of the developers, and expects us to pick up his knowledge of the package to his level in this very short time frame we've got. Development Hell ¯\_(ツ)_/¯ [editline]30th January 2016[/editline] Oh yeah, there's three developers and when I suggested source control or paired programming: "We can just do it from my laptop" man.[/QUOTE] At that point it'd just bail, and just make something of a smaller scope in the time I have left. Game jams are supposed to be for your benefit. To have fun, and end up with something you feel proud of having created. There's no point to it otherwise.

[vid]https://a.uguu.se/ytymnr_2016-01-3009-23-41.mp4[/vid] bhahahaha so fun

[QUOTE=Downsider;49637128][vid]https://a.uguu.se/ytymnr_2016-01-3009-23-41.mp4[/vid] bhahahaha so fun[/QUOTE] In the original the squares show up only where the duck is (since that's how the controller checks if it's pointed at one), but I suppose making it the hit flash makes more sense while staying close enough to the original.

[QUOTE=Downsider;49637128][vid]https://a.uguu.se/ytymnr_2016-01-3009-23-41.mp4[/vid] bhahahaha so fun[/QUOTE] I love the attention to detail with the flashing square. [url=http://electronics.howstuffworks.com/question273.htm]for anyone curious[/url]

[img]http://i.imgur.com/bXWzuWG.gif[/img] i almost scrapped this enemy type because my collision code is so jank

[QUOTE=No_0ne;49638437][img]http://i.imgur.com/bXWzuWG.gif[/img] i almost scrapped this enemy type because my collision code is so jank[/QUOTE] Use quadtrees! Sorry I can't stop seeing quad and Oct trees everywhere

You don't need quad trees for simple collision detection.

You need quadtrees for reducing collision detection tests. Not that it is the only choice and for simple 2D game it's not needed.

I'd implement quadtrees if I was ever going to have more than a dozen physically active objects at a time. The problem is that the new enemy is they are both physically active and can collide with other objects, so when two of them knock together bugs can occur as my original code didn't pay attention to stuff like that. I'll probably implement a special case or something.

[QUOTE=srobins;49636215]At some point you need to set your expectations toward something more realistic. I think we all have ideas for things we'd like to work on that we simply can't do because they aren't feasible projects, I know I've personally dealt with "AAA syndrome" a lot, and have published exactly 0 games as a result. I think it would be better for you to try and pick out the ideas and aspects from your project that you CAN get done (i.e. certain gameplay aspects, story elements, etc.), prototype them with primitives and programmer art, and then try to recruit artists once you've done the heavy lifting to let them know you're serious. If you keep ramming your head into a wall trying to make a million dollar game during your lunch breaks and weekends, you're just going to go insane. Work with what you have, dial it back a bit and work on the concept rather than the assets. Who knows, you might make the next big indie hit.. [editline]30th January 2016[/editline] Keep in mind that, to my knowledge, there have been close to zero large-scale, high quality games released by independent, singular developers. It takes a team to build a game of a certain scale, and no amount of Maya tutorials will change that. The fastest way to attract a team is to prototype and demonstrate your ability as a [I]programmer[/I] and give artists a reason to help you. It's not a personal shortcoming of yours being unable to do the jobs of 10-20 highly paid professionals with years of experience in their respective fields.[/QUOTE] Yeah, you're very much right. I think I'm just gonna try to get PhysX working now and then add some very simple AI that shoots at eachother then dies. That sounds like fun :)

[QUOTE=Tamschi;49637336]In the original the squares show up only where the duck is (since that's how the controller checks if it's pointed at one), but I suppose making it the hit flash makes more sense while staying close enough to the original.[/QUOTE] Yeah, I tried the other way but it was just ugly. I might do both.

[QUOTE=No_0ne;49638553]I'd implement quadtrees if I was ever going to have more than a dozen physically active objects at a time. The problem is that the new enemy is they are both physically active and can collide with other objects, so when two of them knock together bugs can occur as my original code didn't pay attention to stuff like that. I'll probably implement a special case or something.[/QUOTE] Oh, no I totally agree. I've just learned about a new topic and can't stop thinking of it in terms of "You get a quadtree! and you get a quadtree! and you! and you! you!" etc. If it's not needed for you, or won't save you time in the long run, then its not worth it.

The first step of using the PhysX SDK is downloading the PhysX SDK, which is actually a lot harder than you would think because all the links on the Nvidia website lead to the GameWorks download center, which only seems to list plugins for Maya and 3DS and PhsyX APEX but not any actual SDK downloads. Wait did they move the SDK to github? Asking for access right now, but that is a bit of a weird choice since literally all the links point towards the GameWorks download center. Oh wow ok so that is how it works now. Well, now I have to figure out how to build it. However, this is nice since now I don't have to migrate SDK version everytime visual studio updates (since they love breaking their ABIs).

physx should stop being used honestly, its a nice physics engine but nVidia is being really anticompetitive with it.

What would you suggest? I had huge issues with bullet in the past, mostly because the character controller they provide you wasn't complete. Also, when I imported things from blender, it used the wrong axis for gravity so everything got pulled to the left rather than down.

Unity also has problems with Blender (switched axis)

I feel like I'm out of the loop on this one, can someone inform me? (both intel and nvidia).

[QUOTE=WTF Nuke;49639617]I feel like I'm out of the loop on this one, can someone inform me? (both intel and nvidia).[/QUOTE] Something something PhysX runs really bad on AMD.. something something Intel compilers optimize for Intel processors which makes programs run bad on AMD.

[QUOTE=WTF Nuke;49639617]I feel like I'm out of the loop on this one, can someone inform me? (both intel and nvidia).[/QUOTE] Well at least you ain't stuck in infinite loop.

Basically AMD just can't catch a break lol

Oh, that's a shame :/ Is there a nice alternative? Because from a developer perspective PhysX is just so so nice.

Octrees [editline]30th January 2016[/editline] I am so sorry