Note: by "tools" i specifically don't mean game engines, plugins/components or tools that are not limited to game development (Blender, Soundforge, etc).

I'm specifically wondering if you're happy with the fact that you need to write tools yourself once games reach a certain complexity. Do you frown upon tool programming or do you actually like it? Do you rather spend time and money looking for and possibly purchasing a tool that gets the job done, or are you forced to do it yourself because the tool you need doesn't exist? If so, what unavailable tools/features do you desire the most? Which tools really ought to exist but don't or are in a barely usable state?

In general my question boils down to this: do you think there are enough game development tools (map/terrain/tilemap editors, visual User Interface Editors, converters, etc) available specifically for XNA or indie game development in general? If you've used some of them, are you happy with them and if not, why not?

Why i'm asking? Because as a professional game developer who has built and worked with numerous tools over my 10 year career i can tell you that the inhouse tools are typically leaving a lot to be desired but usually it's not that big a deal because a) if need be it can be fixed/improved and b) there's always help available to guide you through some rough tasks. I'm curious how Indie game developers view their tools since most of you either have to build them yourself or live with anything that's out there and in some cases those really useful tools haven't been updated in years.

Actually, one of my main problems is that I like making tools more than I like making games, so I keep going down the rat hole of making tools that have nice "ratchet" sounds when moving objects in snap mode, instead of banging on a game to get it out. I simply get more enjoyment out of the tool crafting, for some reason.

If you're an indie, you should not be making tools. You should be making games. It is very unlikely that, with an Indie game scope, you will recoup the time spent developing any kind of polished tool. If you can slap together something in WinForms in 3 hours that saves you a day, yes, go ahead, but anything more than that is just false optimization. Make the game!

I agree with Jon. Although I enjoy creating tools to a certain degree as well, there are enough of them out there that you should almost never have to make your own.

If you're going to release your mapmaker, I'd say build one. If not, it's not worth the effort.

Thanks for the input so far! Yes, i also often find myself solving problems rather than going forward with making a game. And in many cases that means writing a tool or at least some engine component. I enjoy my part as being an "enabler" for good game development practices and in most cases this means tools and workflows plus the frequent Q&A.

@jwatte: very well said: "If you're an indie, you should not be making tools." .... However as soon as you move forward to something more complex, for example building an RPG world or creating shoot'em up levels it's pretty hard and error-prone to do that without a good tool. I think in the end it boils down to scope, so if you plan to work on the game for 6 months and you get the basic mechanics done in 2 months with the rest of the time only tweaking and adding content, that's when good tools come in handy (and are in short supply i believe). Btw, did you release any of your tools or are they too specific for the games you worked on?

@to anyone: what's your favorite tool or which tool has helped you the most? Disregarding any of the obvious like the 3D modelling tool or XNA obviously. ;) Something game-specific. I still have a hard time finding any game-specific tool that's not an engine, or a game maker in itself. There's only a few tilemap editors (most of them old and quirky) and a few terrain/world editors and that's it.

I would like to have, and still may, create a map editor. But anything else for my game would probably have been a waste of my time.

I'm with jwatte, I love making tools :) At the moment I'm building myself a nice big inhouse physics engine that includes a polygon and level editor, and it's heaps of fun. I'm going to use the same physics engine for all my future games, I prefer to use my own engine than someone elses because it's easier to change things and I know the limitations of it.

Say, since we're on the topic of tools, I'm currently looking for one at the moment. I'm looking for something where you can input mass amount of PNGs and change specific colours in the PNGs to other colours, then output the new PNGs to somewhere else keeping the originals. Anyone know of a tool that can do that?

I like making tools too. I personally can't stand cooky-cutter tools. And that is what would be provided if there was anything provided. Most of the time tools are game specific, so it is necessary that the developers do it.

I loath making tools and put off creating them for a long as possible. When i do create them, they're function but only as much as absolutely needed. For instance, I didn't make any nice gui buttons for selecting which tiles to use in Nasty... instead you pressed T on the keyboard and then a second key to select which tile to use (which I referenced from a printout). Sure I could have programed something much slicker to do that, but something about devoting time to make tools that are never seen by the game player bothers me. I'm don't have a lot of free time, so I try and devote my time to making the game better as much as possible and spending times on polishing my tools takes time away from that.

If i were planning to release or include the editor as part of my game, I would definately make something better. But as is, I dislike making them so much that this will likely never happen.

For Pixel Man I just used Paint.NET for my level editor and it was the best decision ever. For my future platformers, I plan to continue using Paint.NET and simply having some intelligent loading code as I mentioned in this post: http://nickgravelyn.com/2009/09/pixel-man-post-mortem-2/. Things like "if I see 10 green tiles in a row" we use the left end piece, then 8 random grass tiles, then the right end piece. By utilizing something like Paint.NET, I get a robust editor to use for quickly making levels without the hassle of creating and maintaining a custom tool.

for example building an RPG world or creating shoot'em up levels it's pretty hard and error-prone to do that without a good tool

No. Honestly. For RPGs, Excel is all you need. Export a number of CSVs with your monsters and objects and spells and whatnot.
For shoot-em-ups, say with fancy path navigation (a la Galaga etc), you can probably just author all of them in XML. Or Excel, even :-)
I can see how you'd want to create a UI where you click and draw lines for the paths that your alien ships will follow, but honestly -- that's probably going to take longer than the time you "lose" by "having" to use XML to describe your flight paths.
Indie games just aren't big enough to warrant the level of tools that funded AAA games need.
The trick of being an indie is to learn how to not do the million-dollar, full-effort engineering approach. In some respects, anyone can do that, if they have the resources. The astute indie finds a way to accomplish his/her goals at 1/10th, or even 1/100th the cost. Often, that means re-designing the game to not require the same level of effort.

I have been a developer a long time. not a tool creator but an end product creator, I try and find or buy any tool I need and almost never create any.

I really hate tool creation and I usually postpone it until I catch myself doing some mundane task excessively often, or until it becomes clear that some repetitive task will not go away for the foreseeable future.

So for me, it goes like this: see if there's a free tool that does what I need, or almost what I need. If not, try to find a way to get by without a tool.

For example, for a maze game I created with XNA 1.0, instead of building a wacky in-house level editor, I skipped tool creation completely by writing a parser which would read mazes from text files. Then notepad became my level editor :)

With my current game, I'm still undecided on whether I should create a map editor. I believe creating one would allow me to create a higher-quality game experience (by making it easier for me to tweak maps). But on the other hand, creating maps in code (like world.AddBuilding(10, 12, new PowerPlantBuilding());) might work well enough. I'm still undecided and will probably postpone this decision until the game is working and the time comes to hire artists and create content (I'm not doing this in parallel to avoid having to let the artists change or re-do art if I change the game).

I don't typically make very polished tools, but the problem I have with other people's is when they are difficult to implement or don't do exactly what I need and I can't change them. I don't like having to craft my game around someone else's tool. So sometimes I'll build one but it will always be roughly slapped together. The only time I'll build a polished tool is if I'm actually including it with the game for the player to use, like an in-game level editor.

Murudai:

I'm with jwatte, I love making tools :) At the moment I'm building myself a nice big inhouse physics engine that includes a polygon and level editor, and it's heaps of fun. I'm going to use the same physics engine for all my future games, I prefer to use my own engine than someone elses because it's easier to change things and I know the limitations of it.

Say, since we're on the topic of tools, I'm currently looking for one at the moment. I'm looking for something where you can input mass amount of PNGs and change specific colours in the PNGs to other colours, then output the new PNGs to somewhere else keeping the originals. Anyone know of a tool that can do that?

I think you could program a batch converting script in photshop to do something similar, but i've never done it before so please don't quote me on that.

So if i get your answers right, there are two factions: some dislike writing tools because they want to focus on their game or generally don't like writing tools and that's that. And then there's those who actually like writing tools so much, it takes their focus away from making games. The underlying suggestion being that Indies should be resourceful and make do with what they have: a paint program, excel, notepad or even hardcode.

Well ... it kind of reminds me of that time in the 90's when everyone was building their websites with Notepad and MS Paint for the graphics, blink & marquee tags and all. Anyone remember that? Anyone still doing that today? ;)

I've been at this crossroad with my hobby projects as well. And i've been in both factions. Sometimes i was working on a cool game and stopped when i realized i needed to write a proper tool to actually build the game (read: add content). At other times i've enjoyed working on reusable components, plugins or tools so much that the point of actually making a game became a second thought. Looking back at my life's work so far i must admit to myself: i'm a tool developer. A "game tool" developer. I'm not a game developer, i enjoy helping people, enabling other developers more than i enjoy creating entertainment for players. I always keep the entertainment part in mind though because that's what everyone i worked with is doing and i'm the means to the end. But at the end of the day, i do develop tools and sometimes, i even am a tool. I'm also a programmer and tools are created to solve problems. Often very concrete problems. Not so with games. If you have a problem with a game you often don't even know you have a problem, and once you finally realize it the solution is rarely obvious enough that it can be said with certainty that it will actually solve the problem without creating new ones.

Uhm ... just some thoughts off the top of my head why some (like me) actually prefer writing tools over games. I think it may even boil down to: are you a programmer and love solving concrete problems? Or are you the creative type who tries new things out and you care more about the end than the means? The former write the tools the latter could really use. That synergy of two people each in their respective areas working together ... it doesn't happend very often, does it?

steffenj:

Uhm ... just some thoughts off the top of my head why some (like me) actually prefer writing tools over games. I think it may even boil down to: are you a programmer and love solving concrete problems? Or are you the creative type who tries new things out and you care more about the end than the means? The former write the tools the latter could really use. That synergy of two people each in their respective areas working together ... it doesn't happend very often, does it?

I'm definitely doing this as a creative outlet. I get satisfaction from solving a tricky programming problem, but for me it's all about the process of creating something out of nothing and being able to interact with it. Or taking an existing idea and thinking, "I would do this differently or this differently"... and then actually doing it. So when I write a tool it's a means to an end, I don't get a lot out of that as a process in and of itself. My problem with other people's tools is if I have to structure my code around them, or if they are inflexible. Open source would be the best so I could make changes if it wasn't exactly what I needed.

steffenj:

In general my question boils down to this: do you think there are enough game development tools (map/terrain/tilemap editors, visual User Interface Editors, converters, etc) available specifically for XNA or indie game development in general? If you've used some of them, are you happy with them and if not, why not?

 

Case Study 1: The HadesSpaniel Map Editor. 80% complete infinite-layer 2D map editor, featuring full brush support, cut and paste, sort tiles by categories, XML export and import etc etc. Number of games even remotely completed using this tool: 0.

Case Study 2: The HadesSpaniel Platform Game. Loads levels created with Tiled editor, has moving platforms, teleports, currently 10 enemy types, collectables, powerups, switches, keys, all driven by a script engine for powerful set-pieces, a front-end with various options, game is around 65% complete.

Once I got over the mindset of "nothing will be good enough except my own level editor" and started thinking about how I could use existing tools to help make my game, it was easy. I'm much closer to my platform masterpiece that I ever would have been if I built my own editor.

I recently went through the process of developing a fairy sophisticated set of tools, including an IDE framework for adding tool-plugins. It was quite fun at first, and I certainly learn't a lot but my biggest lessons were these:

If you're working with artists who have been working for a while, they are familiar with a toolchain already, in our case, Maya. If you're going to develop tools for your game engine, and you want your artists to be efficient the interface is going to have be like the tools they use (in our case Maya). We got to the point where our tools were holding up our art team because they were not really production-ready, so we actually dropped our tools (having spent several hundred manhours on them), in favour of using Maya's ability to add custom metadata to any object, and export that as a Collada file. Instantly we were able to create pretty much everything we needed in Maya, export as Collada, and if any special interpretation of our assets were needed, we could just put that into our model interpretation code.

Now, granted, this requires your artists to know a little to much about the metadata and engine mechanincs, but if your main aim is to get a game out, it is probably a far better route to take.

I'm still weighing up, in my mind, what the scenarios are where building your own tools would be a good idea, and I guess, you could use this as a rule of thumb:

Developing software is expensive, buying software isn't. If you're duplicating somebody else's work, you're not going to do it as well as they have without spending as much money as they have. They aren't going to charge you everything they have spent. Buy their software. If you've developed a fantastic new technology that nobody else does, and it is impossible to mold any current tool to create the assets that your technology demands, then by all means, create a tool to accompany your technology, but consider making that tool a plugin for existing toolchains, rather than developing your own complete, end-to-end solution (see the logic that this paragraph started with). That's just good business sense.

@HadesSpaniel: great case studies! ;) Did you ever release the Map Editor to the public as XML import/export makes it sound like a flexible tool. Wouldn't there be use for something like that for other developers or is it (as so many tools are) too specific to your game or engine?

@DynamicSamurai: good points about the good business sense. I fully agree. If someone sells a good tool or plugin i am willing to buy it if i can try it first because it is usually cheaper than developing it yourself. The big problem with these things i see is that they are often too specialized or too far away from game development to be useful in general, or too specific to a certain game or engine to be useful more across multiple games in general.

All of this makes me wonder ... i haven't checked the license yet but maybe someone knows if it's allowed to actually make money selling XNA components? Or tools built with or for XNA?

I know that GarageGames sells their TorqueX addon but if i'm not mistaken they have a special deal, at least last time i checked (XNA 2) you couldn't just extend XNA and charge for it although it can be argued that it's more than extensions. And they even sell extensions to their tool, for example the Platformer Kit and others. I've been wondering to focus my efforts into something like this, building a game kit rather than a complete game. I just don't know how much one can actually earn by doing that, or by selling tools and components as many would expect it to be free (simply because most tools/components are free, and even open source in many cases).

steffenj:

All of this makes me wonder ... i haven't checked the license yet but maybe someone knows if it's allowed to actually make money selling XNA components? Or tools built with or for XNA?

On the PC you can sell games/apps you've built using XNA Game Studio as long as you meet any legal issues for stuff you didn't personally create that isn't part of XNA GS (3rd party code/resources).

I am in preparation to write code for my second game so I need to write a particle effect editor (even though somebody probably wrote one) I am writing one my self same with the map and animation editors :)

I would love nothing more than to use existing tools and focus my time on making the GAME, I just haven't found anything that suits my needs - a non-tile based 2d editor (like Gleed2d), but with support for game entities too so I can drop in items, enemies, etc.

Gleed2D? Never heard of that, thanks for mentioning it. It sounds and looks promising and i'm positively surprised to also see tutorial videos, very nice! Too bad it seems dead now after only 2 weeks of frequent updates, and the forum is full of spam and ***. :( ... (nice, the forum automatically added the *** for the "po*n" word *G*)

Personally i'd like something like Gleed but specifically for designing user interfaces. Which means a set of controls i can add in, align with each other, maybe even some automatic layouting (eg moving controls around will move the other controls, think iPhone rearranging of icons). If anyone happens to know such a tool please let me know! But don't say Visual Studio. ;)

Trying to bring this back up in case someone has made some headway, or wants to make an open source XNA World Editor and Engine.

I've been trying to aviod writing the tools but I have not had much luck. I tried FreeWorld3D, which I thought was nice, but I could not contact the author. Unity is offering an Indie version, but does not have an XNA engine. And I purchased Torque X 3D today for $250 but it does not function as advertised. Someone from Torque may help me, but that remains to be seen.

If I were writing this I'd like to see it integrated with VS2008 so that our modifications to the shaders or engine (and tools for that matter) were seen in realtime.

If I were writing this in C++, with DirectX (or OpenGL) I'd be 95% done. Unfortunately I really don't have a C# codebase to work from, and more importantly, developing the tools at this time was not what I had intended. I do enough of this at work.

However, given that viable alternatives do not exist, I'd be willing to work with someone else on developing something that all could use.

BTW, about the only thing I don't know how to do, and I mean within 1 or 2 days, is how to create caves and cliffs. Octree, Quadtree, Multitexturing, Skinning - no problem. I have all the GPU gems, and have done quite a few shaders. One of my favorite sites is GPGPU.

My knee jerk thoughts are that for caves and cliffs I would have to use CSG, with a BSP, then fuse the final assembly.