Is it still realistic to hope to “make it” as a solo/duo indie dev? |
- Is it still realistic to hope to “make it” as a solo/duo indie dev?
- I made a video about 10 most common misconceptions about gamedev I've noticed people believe
- Pixel Art Tutorials for July 2018
- FBX Export/Import from Blender into Unity: this is our workflow (may not fit with your setup)
- The small things that helped us out at our first larger con
- Class structure adivice for a card game
- Made an SDL port of Keen Dreams, with never before seen smooth scrolling (tm) and more
- How are profits typically split between an Indie Game Team?
- 2D lights and shadows algorithm review
- Hey all! We just released a quick video on the Emotional Power Behind Sound Design! I hope you all can get something out of it :)
- OOP design question for RPG games
- This part 7 of my Unity ECS series which covers generating 3D objects from materials and meshes
- I made a little character creator as part of a upcoming project I've been working on
- How do you approach prototypes? And some questions specifically for me. (somewhat of a discussion as well)
- C++ career advice?
- Which 'minSdkVersion' are you using in the android games?
- How to show controller layout or buttons, without using walls of text?
- Should feedback be given during meet ups in person or in small increments online?
- Unreal Engine 4 : Paragon Enemy Damage
- How do I make the screen zoom out when two players in a party game go to opposite directions? Alternatively how to make split-screen, but that isn't as cool.
- Drivetrain simulation constraints
- Gamedev Sound Effects
- Help looking for lua game engine :)
- Building Hype On Social Media
Is it still realistic to hope to “make it” as a solo/duo indie dev? Posted: 30 Jul 2018 04:15 AM PDT With the huge amount of indie games launching these days, is it too crazy to hope to be able to actually make a living doing this? I have some ideas that I feel are decent and could be turned into interesting and fun games, but it seems to be a Herculean task these days to get your game noticed. Of course I enjoy designing and developing, I also just want a realistic idea of what to expect before I pour months or even years of my life into such an endeavor. [link] [comments] |
I made a video about 10 most common misconceptions about gamedev I've noticed people believe Posted: 30 Jul 2018 06:54 AM PDT |
Pixel Art Tutorials for July 2018 Posted: 30 Jul 2018 06:56 AM PDT |
FBX Export/Import from Blender into Unity: this is our workflow (may not fit with your setup) Posted: 30 Jul 2018 12:47 AM PDT |
The small things that helped us out at our first larger con Posted: 30 Jul 2018 07:43 AM PDT We spent the past 4 days showing off our current in development game, The Last Hex. We were at Replay FX in Pittsburgh, PA. It was amazing and we learned quite a lot about our game. We also learned quite a lot about showing at a con. So I figured I would share some of the small things that we have learned and did that really helped with the con experience for us. Think about the small things to bring, and then bring a bunch of those things Yes, cons are about showing your game but its the small things that can really help and make your days slightly easier. Things like a bunch of pens, paper, business cards, markers, wipes, extra cables, scissors, water, snacks, table cloths, sometimes extra chairs, shoe inserts (gels), notebooks, hand sanitizer, mouse pads. It felt like I was in an office, you want to keep your office stocked and clean for when people visit. People enjoy clean and put together offices, it makes you look better. People will also ask for things, and if you can help them they more often than not ask about what you are showing. We forgot scissors and ended up running around to vendors to find one and it was a larger hassle than it really needed to be. Your health is important Cons can take a toll on your body, working a con will take a toll on your body. Constant standing, constant talking, constant early mornings/late nights. Bring a couple cases of water, bring snacks, make sure to have chairs that are always available for your team, make sure to have some kind of insert in your shoes, have a small pharmacy of stuff for your team, remember to eat, remember to shower (both hygiene and also it feels great). Look out for each other and if its only you then make sure to take breaks, set alarms and make sure to sit every once in a while. Get information Get as much information as you can about the con. Who is in charge, who can you contact with issues or questions, get to know vendors, get to know the exhibitors around you. Find out how much space you have, find out how many chairs and tables you have, find out where the power cords or sockets are. Find out if they give you table cloths, learn the con schedule (events, concerts, competitions, ect ect). The more you know the better you can plan, the more vendors/exhibitors you get to know and work with the more acknowledgement you will get and partnerships will form. Have more than one setup We had 4 computers setup and I wish we had more. Sometimes every station was taken and people were waiting to play or said they would come back later and other times everything was empty. This is an opportunity to show off your game, might as well get as many people playing at once. Crowds draw more crowds. It gives people more to look and can get them playing your game faster. We saw a couple of devs only have a single small laptop, yah it is amazing that you are showing your game, but this is your opportunity to create a fan base and get interest in your game. With one small setup you can potentially have a lot of people pass you by because you only have one setup or its to small to see, or already a line to play. Be nice, be attentive, be excited, smile and learn how to read people. People skills are really important, and sometimes it can be hard (I am terrible with people) but it is super important at a con. Be nice, smile, be talkative, be attentive, be in front of your booth (not behind the table), and people will come. You may be tired, you may be hurting but it is an exciting thing to be showing your game off at a con. Learn when to talk and when to just let someone do their thing. Always ask if they want help when they sit down, ask if they want a quick primer or if they just want to go at it themselves. If someone is walking by and looking, ask nicely if they would like to know some info about the game. Yes it is scary but you gotta do what you gotta do to grow and get attention. We also made sure to carry around a small notebook to write down feedback, bugs and recommendations from players. I think this also makes them feel like they helped as you are giving them your full attention and writing in the book means that you are taking their thoughts and ideas seriously. It is ok to step away I was afraid to step away from the booth. What if I missed something or someone! But stepping away helped quite a bit. I think it also helped that we had shirts with the game logo on it. Walking around and talking with vendors and other exhibitors was great. We learned tips and tricks, we learned more about the industry, more about the con and even made some partnerships and started to cross promote games. I even went around and played some arcade games and managed to see people who played our game. Sometimes we started talking and friends would stop by and they then proceeded to talk about the game and told their friends to check it out. Walking around was also a great mental break from things. Start preparing early This is a big point i think. Start gathering all your things way before the con, like weeks. Get your marketing stuff, get your station stuff, extra cables, newsletter sign ups, anything you can think of. This could save some money as well. Find sales from printing sites, find deals on computer equipment and accessories. Will be cheaper in the long run. Ok I think I rambled quite enough. If there are any questions or if there is anything I can expand on please let me know. Thanks! [link] [comments] |
Class structure adivice for a card game Posted: 30 Jul 2018 10:42 AM PDT So I am at an boring webdev job and php is getting tiresome and I started to design and program a game to learn Java. Kinda a mix of slay the spire with the health system of battletech. Right now I want to create many cards probably in an xml or json but still want them to function very differently. Do I write the play function in the xml? Or should I get my functionality thru parameters and construcors only? How do I handle the different types of cards (interface vs inheritance)? I am looking for advice on how to structure the classes Thanks in advance [link] [comments] |
Made an SDL port of Keen Dreams, with never before seen smooth scrolling (tm) and more Posted: 29 Jul 2018 01:51 PM PDT Heya, remember back in the day.. All of the Keen games had this rather terrible every-few-pixels scrolling? Guess what, it's an optimization id Software built in to get better performance out of the CGA/EGA hardware that existed back in the day. With some sorcery it's possible to get smooth scrolling on modern hardware though! Gameplay video: https://youtu.be/lCS4Ec-ciS0 Github: https://github.com/marcel303/keen Now imagine all of the Keen games had smooth scrolling like this. ;-) I'd love to work on this, but I guess getting my hands on the source code won't happen anytime soon. Anyway, It was a nice experience diving into some of the old code. The straightforwardness of DOS programming, the lack of object oriented programming and contrived architectures and polymorphisms. Just some clever code, packed in pragmatic routines. Using assembler for some core routines where performance matters, and high-level C for the rest. Also, oddly, I sort of like EGA's planar display modes. I never liked EGA much, and certainly not CGA, but I feel a bit sad planar display modes and images (textures) have mostly been forgotten. There's a few definite niceties about them! In addition to the straight up porting of the code (and fixing a bug or two while at it), I added software OPL synth emulation, smooth scrolling by hacking the EGA routines and 'sprite shifting' code in there, added support for XInput gamepads (Windows only), and various new command line options like enabling/disabling vsync, manually setting display and logic tick rates and overriding game files. [link] [comments] |
How are profits typically split between an Indie Game Team? Posted: 29 Jul 2018 10:46 PM PDT Say for example you have a 7 man team: 1 Developer 1 Game Designer 1 Artist / Animator 1 Guy doing sound FX 1 Guy doing music 1 Guy doing voiceovers 1 Marketer / PR guy Typically how would profits be split? I realise it may vary depending on the relative talents of each person, the amount of man hours required from each, as well as their negotiating skills, but roughly what would a typical breakdown look like? And what about a 3 man team. How would profit split look for this: 1 Developer / Game Designer 1 Artist / Animator 1 Sound / music / voiceover guy With everyone sharing the load for marketing/PR. [link] [comments] |
2D lights and shadows algorithm review Posted: 30 Jul 2018 09:38 AM PDT I have a little 2D lighting renderer I am making that uses infinite extruded shadows. Right now I am just creating the shadow geometry and blending in semi-transparent black over the lit scene. Instead of alpha blending black into the scene I want the shadows to subtract out the light value. I'm not really a graphics expert, so before I change a ton of code I wanted to make sure I was on the right track. First the pixel value on a rendered scene should be the sum over all lights of the unlit pixel value multiplied by the light intensity and modulated by the light color. The idea for shadows is to subtract off any light influence in the shadow, and this will be scaled with alpha for fuzzy penumbras. The algorithm I want to use is like this. Render the scene to the frame buffer with lights but no shadows. Render the scene again without any lights to a texture buffer. Using the unlit scene as input, render the shadows to another texture buffer initially filled with full alpha. The shadow geometry will be rendered with a shader that colors the scene with the negative value of the individual light that the shadow is blocking, with some alpha blending for fuzzy penumbras. The idea is that the shadow should exactly subtract the light influence from that pixel. Draw the shadow texture over the main scene in the framebuffer with alpha blending. The main drawback is that I have to draw the scene twice. But if I sample the framebuffer of a lit scene to calculate shadow values, then I would have to calculate the sum of influence of all lights at a pixel, in the shadow's pixel shader, which would be bad. Any advice or recommendations would be highly appreciated. [link] [comments] |
Posted: 30 Jul 2018 09:20 AM PDT |
OOP design question for RPG games Posted: 29 Jul 2018 11:56 PM PDT This is probably solved 1000 times, but I'm not sure how to google it. Imagine we have RPG game. We have inventory and character. Which is better and why: 1. store information in item object whether it is equipped or in bag 2. store information in character (i.e.player) object what was equipped and let the item store only item specific data. tia [link] [comments] |
This part 7 of my Unity ECS series which covers generating 3D objects from materials and meshes Posted: 30 Jul 2018 09:58 AM PDT |
I made a little character creator as part of a upcoming project I've been working on Posted: 29 Jul 2018 05:21 PM PDT |
Posted: 30 Jul 2018 12:30 PM PDT Sorry for the generic question in the topic, there is actual substance to the question, but I kept the title basic. This applies to both applications and games, realistically. See numbered questions below for specifics. I understand prototypes, their purpose, etc, but I'm curious to hear how everyone else does/approaches the prototype. Maybe you have a good article that helped lead you to a method you could follow? I understand you 'design, figure out the core, and just recreate the core. Try and create the 'fun' in your prototype'. I personally always run into the following issues when prototyping anything: 1). How do you force yourself to keep it simple? I have an incredibly difficult time leaving something incomplete, poorly structured, not 'as I would like it in the final build' for the most part. If that makes sense. (ex. Rather than hard coding things to test with, I'll create a configurable xml file immediately and already support loading the files as well as tools, etc). 2) How do you go about gauging the success of a prototype? Yourself? Family? Online forum of some kind (some sub reddit? this one?) 3) What software do you prefer to use for prototyping? I personally prefer to use C# for coding in general, but I know that sometimes you go to a simpler/faster method to create prototypes. Which engine do you use (if any), for prototyping? [link] [comments] |
Posted: 30 Jul 2018 12:25 PM PDT After doing gamedev work in higher level scenarios (C# in Unity, Gamemaker, and Mobile stuff with Lua), I want to jump back into C++ and become an expert at the language, working towards being able to land a great job at a bigger studio. I'm curious as to what path to follow after getting familiar with the syntax and overall concepts again. If I go with raw OpenGL I'll most likely learn the most, but it'll take a very long time to actually create something shippable. If I use SFML, it'll be much quicker but I will most likely not have delved deep into the inner workings of everything and it seems like I might as well have just used something like C#. SDL seems to be in the middle. What would you guys recommend if I really want to dig deep, yet also actually create something that doesn't take forever? [link] [comments] |
Which 'minSdkVersion' are you using in the android games? Posted: 29 Jul 2018 11:20 PM PDT Just reviewed the android fragmentation and noticed that the following versions: 4.1-4.3 have only 3.6% of the total amount of the store users. As far as I know the minSdkVersion = 16 (Android 4.1) it's something like a gold standard. But at the same time the per cent is too low and as practice shows the most badly reviews are coming from the old devices. So what do you think about 'minSdkVersion' = 19 (Android 4.4)? And which 'minSdkVersion' are you using? [link] [comments] |
How to show controller layout or buttons, without using walls of text? Posted: 30 Jul 2018 07:33 AM PDT |
Should feedback be given during meet ups in person or in small increments online? Posted: 30 Jul 2018 07:30 AM PDT TLDR; Should the majority of feedback be given when the team meets up in person or should feedback be given progressively between meet ups. Why or why not? So I'm currently developing a game in a group and i was wondering when should we as a team give feedback to each other. We're currently following a close modification of an agile SCRUM approach but because we only see each 1-3 times a month we don't have a huge amount of meet ups. However, when we do have meet ups they can last hours and we really get into the nitty gritty of everything we've done since last meet up. It's great and we accomplish a lot. Of course between meet ups we'll post and update our work as we work on them but we don't necessarily give in depth feedback. So for example, a teammate "finishes" some concept art and we may give some very MINOR feedback like "looks good" or "can you make this colour darker?". Do you guys think we should give in-depth feedback once it's posted or should we save most of the feedback for meet ups in person? My team has had no problem saving and giving the majority of their feedback when we meet up so far but I've been wondering if this process can be improved. I've been thinking that perhaps we should start giving feedback as we update our team in between in-person meet ups because it allows, for example the artist to adapt to the feedback, continue working and I guess overall it may speed up the developmental process. However, by doing this we may not get a really in depth and more thought out team feedback compared to when we save the feedback for meet ups. I reckon with this process we'll still have meet ups but a lot of the discussion may have dwindled down in terms of quality and quantity due to progressive feedback being shared. What do you guys think. Should feedback be given almost all at once or should feedback be given progressively? It'd be great to see both pros and cons of both sides. [link] [comments] |
Unreal Engine 4 : Paragon Enemy Damage Posted: 30 Jul 2018 10:33 AM PDT |
Posted: 30 Jul 2018 10:18 AM PDT I'm making a little clone of Tron and want to spice it up a little by making it scroll. So now I want the scale to decrease when the players go to two opposite edges of the screen, so that one can always see them. Could anyone give me some tips on how this usually works in other games? [link] [comments] |
Drivetrain simulation constraints Posted: 30 Jul 2018 09:50 AM PDT I'm working on a video game that involves the simulation of drivetrain for player - created vehicles. The drivetrain system is composed of a network of sources/loads (motors, wheels, propellors, etc) and constraints(transmissions, clutches, differentials, and fluid couplings), all connected by driveshafts. Each frame, these things happen in sequence.
Right now I am in the process of figuring out the actual constraint functions. Each constraint is expressed as a function of incoming rotation rates, masses, gear ratios, and time step for time dependant constraints(clutch and viscous coupling). I have sorted out the functions for the transmission. I basically just started with constrained rotation rates V1 = G*V2 and conservation of momentum M1(V1-V1_0) + M2(V2-V2_0)/G =0 And then solved for V1 and V2. The part I'm stuck on is formulating the system of equations to describe an open differential. So far I know constrained rotation rates V1=(V2+V3)/2 But I'm totally at a loss for the other two equations. I'm pretty sure one is conservation of momentum where M1(V1-V1_0) + M2(V2-V2_0) + M3(V3-V3_0) = 0 But im at a total loss for the third. Any insight on this would be greatly appreciated. [link] [comments] |
Posted: 30 Jul 2018 09:25 AM PDT Do the majority of video game developers create their own sound or do many purchase royalty free sounds/packs? Also most of the "free sounds" online do not sound very good, any recommendations for places to purchase high quality sounds (near AAA quality)? [link] [comments] |
Help looking for lua game engine :) Posted: 30 Jul 2018 09:15 AM PDT Hi! So recently I've been really looking into Lua, and I think I'd like to dive in and learn it. I have experience with making both basic programs and games with visual basic, and regular programs with python, so I'm not a complete programming beginner. I tried Godot because it seemed great, but it wasn't for me atm, so I may come back to it in the future. For now, I'm looking at Defold, Corona, and Gideros for learning overall game development in lua (dabbling in both desktop and mobile targets). Which one would you suggest a Lua beginner? Thank you in advance! [link] [comments] |
Posted: 29 Jul 2018 03:26 PM PDT Just wondering the different ways you guys are growing your social media pages for your games and building hype! I am currently using instagram but having a hard time getting traction. Any strategies or tips? Feel free to post your social media pages as well! I would love to look and see what type of content people post. [link] [comments] |
You are subscribed to email updates from gamedev - game development, programming, design, writing, math, art, jams, postmortems, marketing. To stop receiving these emails, you may unsubscribe now. | Email delivery powered by Google |
Google, 1600 Amphitheatre Parkway, Mountain View, CA 94043, United States |
No comments:
Post a Comment