|
I can't figure out what the C++ FQA has to do with making you not want to use Win32 since the entire Win32 API is C and the FQA author is trying to convince you to use C instead of C++ for system level work.
|
#
?
Apr 18, 2011 21:27
|
|
|
|
| # ? May 27, 2024 13:28 |
|
|
Paniolo posted:I can't figure out what the C++ FQA has to do with making you not want to use Win32 since the entire Win32 API is C and the FQA author is trying to convince you to use C instead of C++ for system level work. Yes, but I'm not doing system level work. Dealing with the Win32 API is something that I don't have to do with most other languages though, which have their own interfaces that deal with it for me, and one of the big issues the FQA has with C++ is how it tries to straddle C while at the same time doing a bunch of stuff that's directly contrary to how C works. Also it's not so much scaring me off from the Win32 API as C++ in general. He has a lot of negative things to say about the way C++ handles classes, and I don't know how accurate they are, hence my coming here for a more measured opinion on the language. Incidentally, I don't really mind working with C when it comes to making quick programs to deal with file management or other system stuff. I just don't think it's ideal for making games.
|
|
#
?
Apr 18, 2011 21:31
|
|
|
Well personally I think if reading the FQA turns you off C++ then you probably shouldn't be using C++ anyway, so take that as you will.
|
|
#
?
Apr 18, 2011 21:35
|
|
|
The Cheshire Cat posted:I guess this is partially a games industry question as well as a programming question, so if there are any industry people reading this thread, is C++ knowledge an absolute must if I wanted to go into game design, or is it more important that I can demonstrate a good familiarity with more general design concepts, and not worry about language specific knowledge? I have enough of a programming background that it's not as if I can't read C++ code period; I'm just not very familiar with a lot of the Win32 elements needed to make a game that's not just pure console input. That kind of stuff is all handled by simple, standard libraries in Java and VB. Ideal or not, C++ is pretty much the industry standard for game development, but unless you want to be a coder it will be of zero use to you. Most designers work with scripting languages such as Lua and Unreal Script, but if you're familiar with C++/Java then you'll have no problems at all picking them up. I'm a coder, so this is based on what I've been told by colleagues, but generally companies will hire designers based on their portfolios. A good portfolio will contain designs for levels, along with analysis of why the level has been built in the way that it has been. For example: - What have you tried to do in the level to make it stand out, and give variety to the game play? - What cues are you giving the player, to explain to them what they're supposed to do? Being able to provide playable versions of those levels, maybe built in UnrealEd or Hammer is definitely a big plus as well, since you're showing that you're capable of using provided tools to produce something. You may also want to do some research into the games you enjoy (or even better, the types of games the company you're applying for makes) so you can discuss and critique the design decisions that were made in them. Making your own game from scratch using C++ is something you'd be doing if you were thinking of applying for a job as an engine or game play programmer. Obviously, having a well designed game that you've written yourself will impress people and could land you a job, but it's the well designed part that's what people will care about. All the the time you've sunk into rendering, input handling, memory management and whatever else will pretty much be wasted effort.
|
|
#
?
Apr 18, 2011 21:51
|
|
|
not a dinosaur posted:I can't answer the rest of your questions but this is probably the dumbest thing I've ever read
|
|
#
?
Apr 18, 2011 21:51
|
|
|
I'm under the impression that the vast majority of designers are promotions from QA. Unless you mean world design in which case you should be making maps for an existing game instead.
|
|
#
?
Apr 18, 2011 22:21
|
|
|
It really depends on the company I guess, most of the designers at the place I work at were initially hired as interns who were doing degrees in game design, or they simply applied from outside to be trainees. I've worked in the industry for nearly 7 years and I don't think I've ever run into anyone who got promoted from QA into design. I can think of a couple of people who got promoted to QA leads, and then managed to get jobs as producers, but that's a pretty different kettle of fish.
|
|
#
?
Apr 18, 2011 22:35
|
|
|
Eh- some people come in to design through internships, some come from the community, some come from QA, some come straight out of school. It's pretty much all over the map. As for designers knowing how to code, YES PLEASE. God, it's so damned hard to find designers who have technical backgrounds. Given the choice between a non-technical person with stronger design skills and a technical person with weaker design chops, I'll go with the latter 90% of the time. I can teach someone how to design (hell, it's nearly all experience anyway), but teaching non-technical people how to deal with a terminal much less full of debugging... *shudder*.
|
|
#
?
Apr 18, 2011 22:56
|
|
|
Gerblyn posted:Thanks, this is the kind of feedback I was looking for. I don't think I really have the math chops to make for a good graphical or engine programmer (I never intended to apply for those kinds of positions anyway, though), so if most of the programming I'd have to deal with is more on a scripting language level, I wouldn't have too much trouble with that. I've already messed around with scripting languages before (just some minor Fallout 3 tweaks, nothing really showy), so if that's the kind of thing I'm going to want to show off then I think I would probably benefit more by using something like Java where I can jump right into gameplay design rather than having to fiddle with a lot of back end like memory management and rendering (I'm not planning on doing anything graphically complicated since I'm not an artist, so anything I would make on my own would not be demanding enough on that front to require any sort of highly efficient engine). I think I'd be more likely to actually see a project through if I can quickly get to the point where I've actually got something on the screen, hence my being discouraged with C++; if stuff goes wrong I'd really rather it be because of a mistake I made in one of my own algorithms, rather than because I didn't understand one of the more finicky aspects of the language. The former I can usually identify by just watching the thing play out in a debugger. The latter could take me all day just to figure out where the problem is, let alone trying to find a solution. Also by Java and VB.NET (which is what I meant) being cleaner, yeah, I meant syntactically. C++ is kind of inconsistent with some of its syntax (compare cout << "stuff" to C's printf("stuff")), and just has MORE special symbols, like how you have to use either -> or . to achieve basically the same result, based on whether or not you're dealing with a pointer. Though I suppose they're also "cleaner" in the sense of "Garbage collected" (well, Java at least. I'm not sure if VB.NET is or not), so not having to worry about memory leaks thanks to orphaned data is nice. Not that you don't have to manage memory usage in Java, but it's more along the lines of "remove references to stuff you aren't actually using anymore" instead of "remember to free the memory you've allocated before you remove a reference", which is easier to deal with. *edit* I only mention VB.NET since it's something I've used at my current job (it's not a games industry job, just a standard IT position). I've heard a lot about C# in relation to games but never actually tried it out. From what I can tell, it's basically "Java, by Microsoft" VVVVVVVVVVVVVVVVVVVVVVVVVVVV The Cheshire Cat fucked around with this message at 23:28 on Apr 18, 2011 |
|
#
?
Apr 18, 2011 23:06
|
|
|
The Cheshire Cat posted:Also by Java and VB.NET (which is what I meant) being cleaner, yeah, I meant syntactically. C++ is kind of inconsistent with some of its syntax (compare cout << "stuff" to C's printf("stuff")), and just has MORE special symbols, like how you have to use either -> or . to achieve basically the same result, based on whether or not you're dealing with a pointer. Though I suppose they're also "cleaner" in the sense of "Garbage collected" (well, Java at least. I'm not sure if VB.NET is or not), so not having to worry about memory leaks thanks to orphaned data is nice. Not that you don't have to manage memory usage in Java, but it's more along the lines of "remove references to stuff you aren't actually using anymore" instead of "remember to free the memory you've allocated before you remove a reference", which is easier to deal with. .NET has all the garbage collection and object orientation of Java, but you might want to refer to C# instead of VB if you want to avoid a witch hunt. Just saying. ninja edit for less persecution complex: XNA also has far better support for C# than VB, and just about every .NET/XNA studio out there uses C# as a result. So if you want relevant experience, prefer C# over VB. ninjeff fucked around with this message at 23:20 on Apr 18, 2011 |
|
#
?
Apr 18, 2011 23:18
|
|
|
Cheshire Cat, It's possible to build a good engine from the ground up in C++. However, I wouldn't recommend it unless you are expert-level with C++. There are many caveats to using C++ and they will stumble you before you reach the finishing line. That article is perhaps a little bit pessimistic, things aren't nearly as bad as they are trying to argue, but only if you know what you're doing. If you don't know what you're doing then yeah, C++ more than most other languages is pretty terrible. In fact, in my opinion you shouldn't build your own engine at all, no matter what language you use. If you're trying to get a job as an engine designer then that's one thing, but probably you're trying to be a gameplay programmer or graphics programmer or whatever. Or, if you're just trying to make a simple game. If you build your own engine you'll be spending 80% of your time building the technology when you really need to spend 80% of your time doing design and usability and making the game fun. In other words, use Unity or UDK or something.
|
|
#
?
Apr 18, 2011 23:20
|
|
|
Well, you have to bear in mind that what I'm going for right now is just a simple Roguelike to get myself into the groove of actually taking ideas and making them into something concrete. Anything I make now I might not even show anyone (though I might end up using it as a portfolio piece a few iterations down the line when there's more meat to it). So any sort of "engine" I might be talking about is really just something as basic as "take the grid of level data, and put the relevant png/bmp/jpg/whatever files on the screen in the same layout". If I was going to work with something 3D, I probably would use Unity or UDK because as I said, I really don't have the knowledge to build my own 3D engine from scratch. Honestly, I would probably just use GameMaker except that something about GML just bugs me, though I can't quite put my finger on what. I think there's just one too many layers of abstraction between the game design and the code; with something like Java, all my methods are right there in the class definition. GML you might have some methods defined in code, but you might also have some of them handled via the "events" screen, which doesn't have any code at all. I think I'm somewhat of a luddite in that respect; when it comes to designing behaviour, I'm happier with just a text editor and a compiler than I am with a big toolset; I only really want to use visual editors for things that are visual in nature, like map layout.
|
|
#
?
Apr 18, 2011 23:36
|
|
|
not a dinosaur posted:I can't answer the rest of your questions but this is probably the dumbest thing I've ever read I think C++ is uglier than Java  I know this is becuase I worked with Java for several years before touching C++, but I'm probably the only person on earth who thinks C++ is more verbose. That said, I know I still gotta work with it and get used to it. Also, whoever made C# is awesome. C# owns and so does XNA. Orzo posted:You may like C++ more but a lot of people would agree that Java (and VB.NET, if that's what he meant) is a lot cleaner looking than C++, including myself. Yeah this sums up my feelings well. I understand why I need a header file from the standpoint of how C++ works. I just think they are a terrible concept and redundant. The Glumslinger fucked around with this message at 01:23 on Apr 19, 2011 |
|
#
?
Apr 19, 2011 01:18
|
|
|
I mean in general the more high-level a language is the more simple (clean) it's going to look, so it's really not a surprise. Eliminating the entire aspect of memory management tends to reduce your code to a much cleaner and readable piece of work.
|
|
#
?
Apr 19, 2011 01:26
|
|
|
I'm having some problems with moving sprites in XNA. So far I'm drawing a single sprite which I can move around with the arrow keys. Very straightforward. The weird thing is sometimes the movement will be nice and smooth and other times it will be jerky and stuttery. Sometimes I'll run my game and it will be fine and then if I close it down and immediately start it again and the stuttery movement is back. Google seems to suggest updating the video card drivers, and this could be the case as I recently had to format my computer, but could anybody shed some more light on this?
|
|
#
?
Apr 19, 2011 02:00
|
|
|
This might be a silly question, but are you sure it's a framerate issue and not a game logic/stuttery input issue? Are you outputting frames per second?
|
|
#
?
Apr 19, 2011 02:18
|
|
|
Try monitoring the IsRunningSlowly property of the GameTime object that gets passed to the Update and Draw routines. It should be false if the game is locked at 60FPS but if it is true then the Game object is falling behind and is running multiple Update calls before doing a Draw call in an effort to catch up. For something as simple as moving one sprite around IsRunningSlowly should always be false. If it is set to true then you most likely have an external issue like wonky video drivers. If IsRunningSlowly is false and you still see stuttering I'd guess there is some kind of issue with how you are calculating the position of the sprite - like if you are running a physics model with an integrator it might be returning the wrong value, etc.
|
|
#
?
Apr 19, 2011 02:21
|
|
|
No, I have nothing going on like physics or even collisions. The sprite is literally the only thing in the program. I'm not at my home computer at the moment but I did check IsRunningSlowly and from memory it was coming back true. I tried to compensate for this by multiplying my sprite's movement speed by the elapsed game time but that didn't work either. I'll test it again when I get home and try updating my drivers and report back. Thanks for the help guys.
|
|
#
?
Apr 19, 2011 02:57
|
|
|
crazylakerfan posted:I think C++ is uglier than Java Ya I was really surprised by how much power you can still have with C# and how Java-ish it is and takes care of the no-brainer stuff like memory management and garbage collection. I only learned it because I worked on a project that used it and found that it is the best option between C++ and VB
|
|
#
?
Apr 19, 2011 04:40
|
|
|
The Cheshire Cat posted:Well, you have to bear in mind that what I'm going for right now is just a simple Roguelike to get myself into the groove of actually taking ideas and making them into something concrete.
|
|
#
?
Apr 19, 2011 12:53
|
|
|
Mustard Snobbery posted:No, I have nothing going on like physics or even collisions. The sprite is literally the only thing in the program. Don't move based on frame rate, do it based on delta time. In Panda3D using Python that looks something like this: code:This takes frame rate out of the equation. You get better overall movement; though, it doesn't explain why you're running slow to begin with. So I guess this is more of a 'quick tip' and doesn't solve the root cause of your problem. Edit: forgot to add that you should have some protection against skipping frames or moving too quickly or something. I would throw an early escape in there... code:Winkle-Daddy fucked around with this message at 16:06 on Apr 19, 2011 |
|
#
?
Apr 19, 2011 16:02
|
|
|
Mustach posted:Then use an engine like libtcod instead of writing one from scratch. This. libtcod is really nice, can be called from nearly any language you can think of and let's you concentrate on what you JUST said - taking your ideas and turning them into something concrete. You will hear it repeated ad nauseum in this thread: make a GAME, not an ENGINE, unless as already pointed out - you want to be an engine developer. I messed around with libtcod and had something up and running about literally an hour.
|
|
#
?
Apr 19, 2011 16:39
|
|
|
Newbie coder here, teaching myself python and mysql currently. I am working on a racing MMO much like goallineblitz and court rivals. A game where you create the player and level him up and compete against other players All the races will be simulated, so you just adjust the stats and other stuff before each race. I have the basic race engine done and am starting to learn php & mysql, as I thought it would be the idea thing for the web interface and to store all the users data. However I have never done anything like this before wanted to make sure I was not making a big mistake by using python, php and sql. Any suggestions or advise would be greatly appreciated.
|
|
#
?
Apr 19, 2011 21:55
|
|
|
Logic posted:Newbie coder here, teaching myself python and mysql currently. I am working on a racing MMO much like goallineblitz and court rivals. A game where you create the player and level him up and compete against other players All the races will be simulated, so you just adjust the stats and other stuff before each race. I have the basic race engine done and am starting to learn php & mysql, as I thought it would be the idea thing for the web interface and to store all the users data. However I have never done anything like this before wanted to make sure I was not making a big mistake by using python, php and sql. In my humble opinion, if you're just doing stat based racing using python/php/mysql I don't see any issue with that. If I understand what you're doing correctly choosing to use high level languages is the way to go for doing this quickly. If you're learning Python both Unity3D and Panda3D have python bindings and would be a good next step for game development for you. They also both have web players so you can get bitchin' 3D graphics using languages you're already getting familiar with.
|
|
#
?
Apr 19, 2011 22:26
|
|
|
Mustach posted:Then use an engine like libtcod instead of writing one from scratch. I've actually already taken a look at libtcod before, but from what I can tell, what I'd like to do can't be done with that engine; one of the key things I want to go for is the ability to generate maps in 3 dimensions instead of one level at a time, so that the game is displayed isometrically and can do things not seen in most roguelikes, like multi-level rooms where flyers are actually flying, pits that actually lead directly to a specific spot on a lower level (and could actually be climbed back up to reach your former position), etc. Dwarf fortress uses 3 dimensions, but I don't think the top-down display works very well to communicate information to the player. If there is an engine out there that could do this sort of thing I'd be happy for the recommendation; even if it was something more for X-com like tactical games, as long as it was open source I could modify it to suit my particular needs for this game.
|
|
#
?
Apr 19, 2011 22:33
|
|
|
Can anyone recommend some free particle system for C/C++? I'll be using OpenGL but I'd assume most libraries would try to be rendering API agnostic.
|
|
#
?
Apr 20, 2011 01:18
|
|
|
I'm having a weird issue i'm not entirely sure how to approach correctly, in my game you shoot space ships with a turret, anyhoo the problem i'm having is that it's really hard to place the crosshair correctly for shooting stuff at far and short distances. You can play it here: http://dl.dropbox.com/u/6564397/3dr/3DRenderer.html At the moment I get a point an arbitrary (far) distance from the direction of the cannons, then have the camera look at that point, then have the bullets converge on that point. As is the bullets converge on the point seemingly fine, but the centre of the camera is off by a bit (like 20 pixels out of 240) at max range, so I adjusted the crosshair graphic accordingly. This however makes it hard to aim at stuff reasonably close without it seeming completely off. Is there like an actual way of doing this in some genius clever way I can't think of (other than raycasting from the cannon to see if it hits anything and placing the cross hair there which i'll probably try but it seems a bit wonky)? I mean if I look at flight sims convergence is a pretty big deal and the cross hair is set to a specific range anyway so it's not a huge deal if it is off, but it feels like it could be tighter.
|
|
#
?
Apr 21, 2011 03:43
|
|
|
brian posted:I'm having a weird issue i'm not entirely sure how to approach correctly, in my game you shoot space ships with a turret, anyhoo the problem i'm having is that it's really hard to place the crosshair correctly for shooting stuff at far and short distances. I'm probably missing something obvious but how do you control this game? It doesn't seem to respond to any mouse or keyboard input. I'm using Chrome 10.0 so maybe it's just not compatible?
|
|
#
?
Apr 21, 2011 04:00
|
|
|
It's flash, so click anywhere in the box in the middle and then use the arrow keys to aim and space to shoot.
|
|
#
?
Apr 21, 2011 04:04
|
|
|
brian posted:It's flash, so click anywhere in the box in the middle and then use the arrow keys to aim and space to shoot. I'm just stupid and never thought to use the actual arrow keys instead of wasd.
|
|
#
?
Apr 21, 2011 04:13
|
|
|
You have a parallax problem. Alternate way to do it is to raycast from the camera instead of from the turret, then aim the turret at whatever the crosshair is looking at, or if it's looking at nothing then converge at a default distance. This way the crosshair should never move. Also the big problem with aiming is that the arrows are overly sensitive. It's hard to hit such a small target so far away. Where are you planning on going with this game design wise? It needs more to spruce it up.
|
|
#
?
Apr 21, 2011 08:09
|
|
|
Vino posted:You have a parallax problem. Alternate way to do it is to raycast from the camera instead of from the turret, then aim the turret at whatever the crosshair is looking at, or if it's looking at nothing then converge at a default distance. This way the crosshair should never move. Computer-assisted aiming might help.
|
|
#
?
Apr 21, 2011 14:30
|
|
|
Vino posted:You have a parallax problem. Alternate way to do it is to raycast from the camera instead of from the turret, then aim the turret at whatever the crosshair is looking at, or if it's looking at nothing then converge at a default distance. This way the crosshair should never move. Yeah I was already doing the whole bullets go to where the camera is looking thing, it's just the manual offset made it look wrong when it was working fine. I turned that off and it's all good. As for controls and game design, I literally started the renderer last week and got OBB collisions done yesterday so like not really looking for input about either at the moment. I mean the game itself is going to be a rather simple arcadey game where you defend a big ship from lots of little ships via turrets all over it that you upgrade and change weapons of and whatnot. Anyway thanks for the help.
|
|
#
?
Apr 21, 2011 14:42
|
|
|
XNA's garbage collection is driving me crazy. I've been having a slow memory leak the entire time in my game. In trying to narrow it down, I noticed that my memory leak is coming from SpriteBatch.DrawString. Everytime I draw something on the screen, it instantiates a new String object and then takes a little bit to release it. Well when I'm drawing debug data on the screen, that's a new String object for every single frame. That adds up. I've tried switching to StringBuilder, but it still has to .ToString() it when it draws it on the screen so it's the same problem. Anyone have any tips on how to deal with this? If I comment out DrawString, my memory leak completely goes away. Only thing I can think of is to create a graphic 'font' and draw those little sprite letters onto the screen whenever I want to write something, but that just seems wrong for some reason.
|
|
#
?
Apr 21, 2011 15:01
|
|
|
Out of curiosity, how did you manage to narrow it down to a new String being the reason for the leak? If you are sure it's a string causing your memory leak? Is it possibly this issue? http://www.riemers.net/Forum/index.php?var=426&var2=0 Are you drawing a different string every single frame? Because strings are reference types and there exists only one copy of a given string in an application at the same time, which makes me doubt that the memory leak is actually the string itself. Maybe you could post some parts of your code?
|
|
#
?
Apr 21, 2011 15:09
|
|
|
The part that I comment out to get rid of the leak is: spriteBatch.DrawString(font, debugString, new Vector2(0,0), Color.white); One thing I thought of, in typing that out, is that I'm creating a new Vector2 every single frame too. I'll try removing that as well. I'm using the same font for all of my on-screen text so I don't think it's the same issue as the one you posted. The reason I believe it's the String is because when I watch the app using a memory profiler, I'm watching thousands of strings being created and released when the GC is ran. When I comment that single line out, those strings aren't being created at all.
|
|
#
?
Apr 21, 2011 15:19
|
|
|
Okay, based on that, the problem is probably not strings or Vector2 (Vector2 is a really simple struct and I'm pretty sure they can't cause memory leaks). If you could post more code that might be useful. Edit: Going to need to Google more, you should too, some people are having very similar problems and it might be strings after all. But those were very old (1.0, 2.0) versions of XNA and an XNA dev says they were fixed. Orzo fucked around with this message at 15:34 on Apr 21, 2011 |
|
#
?
Apr 21, 2011 15:27
|
|
|
When I comment that out, the memory leak stops and all the strings aren't being constantly created and destroyed. This is the exact issue I'm having: http://forums.create.msdn.com/forums/p/2589/12844.aspx All my code is at home so I can post more later tonight. edit: Ya I've googled this and found the same results. I wasn't sure if anyone else had found a fix that I haven't found yet.
|
|
#
?
Apr 21, 2011 15:42
|
|
|
That is the thread I was reading. But like I said, it's for a very old version of XNA and you shouldn't be having the exact same problem. Strings being constantly created and destroyed does not indicate a leak, it just indicates that strings are being used. If they are being created/destroyed at the same rate, you shouldn't be having problems. Do you have some numbers in this regard about their rates?
|
|
#
?
Apr 21, 2011 15:50
|
|
|
|
| # ? May 27, 2024 13:28 |
|
|
I'll post up more detailed information when I get home tonight. Sorry, I'm at work right now so I don't have access to all my stuff.
|
|
#
?
Apr 21, 2011 15:52
|
|