We get constant requests for Duality ZF to be released on the PC.  Not everyone has an Xbox 360, but most people have PCs.  Since XNA compiles to the PC -- how cool is that? -- it raises the question:  Why are we not releasing our games to the PC?  However, I don't see anybody doing this.  The number of PC XNA games is small.  Have any of you considered distribution on the PC?  What road blocks are stopping you?  I'll jump in with my concerns after I hear what a few of you think.

I'd say top concerns for most hobbyists would be distribution channel (all from your website, attempt Steam, something else?), handling payment for the game (is PayPal good enough for this, what other alternatives do we look at?), and security issues (how far do I go to make my game secure against piracy?  where's the line between security and driving off my customers?).  These things just don't come up on the XBox, so it tends to be the path of least resistance.

I don't have a QA staff or machines to test on enough video cards/configurations/OS's to even begin to think I could release a PC game for money and have any hope of it running and not crashing on all the various setups. I release all of my samples for free for PC and if I made a game, I would definitely consider releasing a free version of it for the PC. But I could never charge for something I couldn't verify.

That's about it for me. Payments are easy with things like paypal, and I already have websites for me it really is just that I wouldn't feel like a good human being selling something I couldn't test.

I'm very much considering it for Nasty. I invented a lot of time and money into Nasty and the XBLIG service simply is not providing nearly enough downloads for me to make it a profitable endevor. In comparison to how long it took to create Nasty, the amount of effort to create a PC version is minor (expecially considering that I do 90% of my development/testing on the PC anyways). I think PC gamers aren't so jaded about pricing either and are willing to pay more for a good game, routinely paying $10 to $20 (and in some cases a lot more) for games.

The issues I see are:

Installer... the built in 'Click Once' type installer doesn't work well at all (and that's being kind), so it requires something third-party or custom written. I don't think this is a big issue though and suspect there are free/cheap installers you can find.

Input... the same type of input we use for our games only works with Xbox controllers. A PC game would require both keyboard support and gamepad support.

Compatibility... This is easily my biggest hangup. I've had so little luck getting Nasty to run on anyone else's PC. One big issue seems to be that many monitors don't appear to1 support 1280x720 screen resolutions so I would have to look into some way of using a more standard resolution and scaling the game to that. I don't have the testing resources to test the many different PC configurations either..> I can test a couple, but that's it... I would hate to have to deal with angry customers who find that my game doesn't run on their system and I don't want to end up going tech support for them.

Distribution... I'd sell Nasty on my own website, which involves setting up the website to take credit card orders. Because I have only one game, I don't need anything elaborate. Portals are something to consider too and could do a lot in helping get the word out about your game. I haven't looked into these much and I'm not sure if non-flash or XNA games are even allowed by many of these.

The reasons I'm developing only for the XBox at this point are:

1) It's easier to develop for a single platform than multiple platforms (and I'm not even done with the easy case yet ;)

2) Microsoft handles XBox distribution and sales on an existing channel.

3) "XBox Indie Games" has built-in name recognition.

4) No worries about piracy - the XBox distribution channel handles it.

5) PC sales / piracy might cannibalize XBox sales.

6) PC distribution requires a .NET runtime - is this widespread?  Used to be an issue.

7) XBox systems are more uniform than PC systems - fewer platform compatibility worries.

XBox-only is arguably the wrong decision since I actually do all builds and most testing on the PC before testing on XBox, but these are my reasons for it.

It will be interesting to see what other people think.

Well,

IMO there are no technical reasons (I'm going to distribute my game on the PC as well). The only real limitation is the usage of the LIVE service: if you don't qualify for GfW (which is still closed to indie developers) you will not be able to use any of the LIVE support and obviously you have to use a different distributor. Just remember that if there is any logic content you don't want to make public domain you should obfuscate your code.

That's it.

Cheers,
Pino

We're planning to release Square Off for PC. In fact it's been in the pipeline since the beginning because we really want to put it into the local (Australian) market... which sadly still isn't possible with XBLIG. We're still deciding if we attempt Steam or similar or just self distribute. My gut feel is that Steam would get you a LOT of exposure, although I've read that they drive a hard bargain.

Controls won't be a huge problem - it already supports PC gamepad and keyboard (although playing a dual stick shooter on the keyboard is an art in itself!).

As for compatibility, it is a problem, but I don't think it's insurmountable. Test as many different setups as possible, give minimum system requirements, offer a free trial version and recommend people try before they buy as there will be no refunds.

PinoEire:

IMO there are no technical reasons (I'm going to distribute my game on the PC as well). The only real limitation is the usage of the LIVE service: if you don't qualify for WLG (which is still closed to indie developers) you will not be able to use any of the LIVE support and obviously you have to use a different distributor. Just remember that if there is any logic content you don't want to make public domain you should obfuscate your code.

Your code is not public domain regardless of whether or not it's obfuscated, and obfuscation does not stop pirates/unscrupulous individuals from disassembling your binaries. Just FYI. (Of course, if you actually *release* your code into the public domain, then it's public domain :) )

As far as PC distribution goes, I would strongly advise anyone interested in getting the most out of their work to take a shot at it. It's more difficult than releasing on the XBLIG marketplace, for sure, but the potential benefit can easily outweigh the time investment - your potential customer base is an order of magnitude larger, and you have significantly more control over distribution, pricing, updates, etc. Downsides like not having access to things like the XNA Net APIs are a pain, but there are easily accessible alternatives out there that you can somewhat trivially use to get feature parity for a PC version of your game.

As mentioned earlier in the thread, QA presents a significant challenge, due to the variety of machines out there - but since you're targeting the PC, there's nothing stopping you from releasing your game's PC version quietly and slowly ramping up your customer base via word-of-mouth and advertising, so that you can spot compatibility issues a few customers at a time before you start heavily marketing and promoting your title. Not needing to go through Peer Review also means that once you spot a compatibility issue, you can basically ship a fix as soon as it's done if you want to, instead of making your customers wait indefinitely for your fix to make its way through PR.

Piracy is obviously a big problem on the PC, but there are lots of tools at your disposal if you want to try and address it. Distribution channels like Steam offer effective anti-piracy solutions that customers don't mind, and if you're feeling crazy you could even license some of those insane DRM packages like Starforce - it's up to you. Offering cool online features like multiplayer and leaderboards only to registered players can also be an effective piracy deterrent.

Releasing on the PC also gives you access to some very powerful incentives that you can't offer on XBLIG - for example, you're free to accept preorders, and offer customers bonuses in order to incentivize them to preorder, which means you can start bringing in revenue before your game has even shipped. You can also easily release your title in an episodic format if that's what you choose, shipping chunks of content out every few months, and then submit those episodes in larger portions on XBLIG once you've had time to act upon all the player feedback you've received.

George Clingerman:

I don't have a QA staff or machines to test on enough video cards/configurations/OS's to even begin to think I could release a PC game for money and have any hope of it running and not crashing on all the various setups...

Couldn't you just have a demo and tell people to run the demo before buying the game, to make sure it works with their rig.

Zenfar:

George Clingerman:

I don't have a QA staff or machines to test on enough video cards/configurations/OS's to even begin to think I could release a PC game for money and have any hope of it running and not crashing on all the various setups...

Couldn't you just have a demo and tell people to run the demo before buying the game, to make sure it works with their rig.

Telling someone to do something and having them actually do something are two entirely different things :) Years of tech support has left me with little faith for users being able to follow directions.

Kris Steele:

Installer... the built in 'Click Once' type installer doesn't work well at all (and that's being kind), so it requires something third-party or custom written. I don't think this is a big issue though and suspect there are free/cheap installers you can find.

I'd like to hear what problems you've had. I've tested ClickOnce out of curiosity for potentially using it myself later and had no issues with it.

Kris Steele:

Input... the same type of input we use for our games only works with Xbox controllers. A PC game would require both keyboard support and gamepad support.

That's a big factor depending on game. Some games just don't transfer well. For instance my game Bloc makes no sense unless you have an Xbox 360 game pad. Sure some PC users will have one, but I doubt many.

Kevin Gadd:

Piracy is obviously a big problem on the PC, but there are lots of tools at your disposal if you want to try and address it.

IMO time is much better spent not worrying about it. If pirates can strip out the DRM of a multi-million dollar project like Fallout 3 or Spore in days, odds are there is nothing you can do to stop them from pirating your game if they wanted to. And if nobody wants to pirate your game... well then obviously it's wasted work to prevent nobody from doing nothing. :p


Personally I want to tackle PC at some point. On PC you get true online leaderboards (none of this P2P hackery nonsense), support for mods and other user content, and have the ability to flexibly update and ship DLC for your game. There are downsides but it's something I'm thinking about for future games, especially as I start planning games that certainly have been done before with keyboard/mouse controls.

To me, the largest stopping block is exactly what George said. I don't have the ability to widely test my game. I have the advantage of using the SunBurn engine which may handle most of that under the hood (I don't know yet), but there are still a lot of factors involved. Sure a demo helps, but gamers like to see some sort of minimum/recommended specs to know whether it's even worth the download.

As far as the looming piracy issue, my games on PC will never ship with DRM or anti-piracy stuff. From my experience as a consumer, anti-piracy stuff is always ineffective if anyone really wants to crack the game and most of the time quite annoying to users who honestly purchased the game. I'd rather just spend time rewarding people who bought it than trying to stop people from cracking it.

Nick Gravelyn:

I'd like to hear what problems you've had. I've tested ClickOnce out of curiosity for potentially using it myself later and had no issues with it.

It works, but it's flawed. Here's a couple comments from people I tried to get to test a click once install for Nasty (these are normal, non-techie users and nasty is a sub-50 meg game)... I have more somewhere but they're all pretty similar to these...

I've been waiting for the ".NET Framework" thing to start installing for quite some time now. Known error?

I downloaded it and tried to install it. Installer took too long so I canceled and deleted it.

I ran "setup". It seemed to finish, although there was no confirmation-of-installation screen.

It does seem to install each time but it is so slow (likely because of the .Net framework) and it doesn't give any indication that it is progressing (causing people to believe that it has failed). The slowness is probably nothing we can fix, but there needs to be something displayed to prevent the user from thinking the install has failed/locked up.

In my own personal testing, I could not ever get it to create any icons for the newly installed program (using a Windows Vista 64 computer). Some users echoed this problem (not sure if others didn't have it or just didn't mention it). When the installer got done, it simply closed and gave no indication of a successful install or not. I had to browse on my computer to find the installed game and manually create my own icons for it. This is certainly not something I can expect the average user to do.

There is also no upfront compatibility testing done... users were NOT happy to spend all that time waiting for it to install only to find out their graphics card was not good enough. I suppose I could wrap the installer in something and do this upfront first, but it would be nice to have it all in one nice package.

Nick Gravelyn:

That's a big factor depending on game. Some games just don't transfer well. For instance my game Bloc makes no sense unless you have an Xbox 360 game pad. Sure some PC users will have one, but I doubt many.

You could state that the game requires an Xbox 360 game pad.  If there was a central location for XNA games on the PC, the players going there would have the required gear and graphics cards (if the target was X-Box compatible shaders).  It would be great to have a central marketplace for this.

As for piracy why worry about it?  If your game is already not selling that great on the XBox just price it for $1 on the PC side and see what happens.

Zenfar:

You could state that the game requires an Xbox 360 game pad.  If there was a central location for XNA games on the PC, the players going there would have the required gear and graphics cards (if the target was X-Box compatible shaders).  It would be great to have a central marketplace for this.

As for piracy why worry about it?  If your game is already not selling that great on the XBox just price it for $1 on the PC side and see what happens.

I'd love if we could expect that, but I think it's unrealistic... expecially if we're selling $5 to $20, you can't expect someone to spend $50 on a gamepad. Many gamers already have gamepads, they don't want to buy more.

I'm on the same boat about piracy... it's bad and any measures you take are only going to inconvinence legit users. People are going to pirate your game (a lot). Enter the market knowing that... also know that some of those people that pirate your game will actually enjoy it and either buy it themselves, spread the word, or get others to buy it.

Guys,

I really don't see the input device problem. An XNA game will work just fine with any controller compatible with the Microsoft Common Controller Driver model. In the PC game (#if !XBOX360 ... #endif) you just let the player choose the controller type if you don't want even bother to check by yourself. You need to use DirectInput to manage controllers other than the MCCD model compatible, that's it... nothing big actually.

Cheers,
Pino

George Clingerman:

I don't have a QA staff or machines to test on enough video cards/configurations/OS's to even begin to think I could release a PC game for money and have any hope of it running and not crashing on all the various setups.

This is my #1 reason for never doing a PC release for an XNA app.

My #2 reason would simply be that it's .NET, which experience has taught me never to trust outside of a controlled environment (like the Xbox or a corporate network).

I've done some PC game work, and for that I much prefer C++ in combination with compact libraries that will handle the boilerplate stuff and then get out of the way and let me do what I want.

Zenfar:

You could state that the game requires an Xbox 360 game pad.

Sure, but that's not really that many people. I'd rather spend time working on a new game that works better for keyboard/mouse than spend time setting up any distribution/sales mechanism for a game that most people won't play. I know some sales are better than none, but I'm just against releasing a game on PC and saying you have to have an extra accessory to play it. If it's on PC, I should be able to play with a keyboard and mouse, or a game pad.

As a side note, I really wish more retail PC games gave me the choice of a game pad. Last I tried both Half Life 2 and Battlefield 2 failed to recognize my analog triggers, and most other games don't have any support at all. It seems like such a simple thing to add (especially for companies like Valve and Dice who have done console verions) that it makes it very frustrating to me.

Zenfar:

If there was a central location for XNA games on the PC, the players going there would have the required gear and graphics cards (if the target was X-Box compatible shaders).  It would be great to have a central marketplace for this.

The problem is that there isn't just one required graphics card. For Bloc you'd only need shader model 2.0. For my next game, we're going to fully use 3.0. For some XNA games, you probably only need 1.1. Then there's various levels of CPU requirements and memory requirements. There is no single minimum bar that will be adequate (without excluding a lot of people) so every game needs to know its own minimum bar and communicate that to users. Finding this bar is challenging. Not to mention some features work differently from GPU to GPU making the whole thing that much harder to deal with.

PinoEire:

Guys, I really don't see the input device problem. An XNA game will work just fine with any controller compatible with the Microsoft Common Controller Driver model. In the PC game (#if !XBOX360 ... #endif) you just let the player choose the controller type if you don't want even bother to check by yourself. You need to use DirectInput to manage controllers other than the MCCD model compatible, that's it... nothing big actually.

Correct me if I'm wrong, but isn't the Xbox 360 controller the only controller currently available that is Microsoft Common Controller Driver compatible? I thought I heard that mentioned once.

I don't think this is a big deal, but it is another hoop to go through w/ PC development. With a PC, because you're dealing with multiple controller configurations, you probably want to allow the user to change their default configurations too, which is something most Xbox 360 games don't allow (I don't do it in any of my games right now).

PinoEire:

I really don't see the input device problem. An XNA game will work just fine with any controller compatible with the Microsoft Common Controller Driver model. In the PC game (#if !XBOX360 ... #endif) you just let the player choose the controller type if you don't want even bother to check by yourself. You need to use DirectInput to manage controllers other than the MCCD model compatible, that's it... nothing big actually.

The problem really is the other way around. Many of us make games targeting the Xbox 360 and assuming use of a game pad, whereas on PC most users will expect keyboard/mouse controls. No, it's not a big issue, but it is something to consider before trying to release a game on the platform.

Fallout 3, Champions, and Marvel Ultimate Alliance.  Champions and Marvel Ultimate Alliance include button graphics and excellent control with the gamepad as well.

I can see making a new game specifically for the PC instead of a port, beside new projects are always such fun! 

I say tell the end user the shader model required and post a demo app, perhaps you can have the only way to purchase the game is to run the demo app!

My concerns are:

1. PC piracy may stunt Xbox 360 sales, since why buy it if you already have it?
2. PC speeds differ and can affect online scoreboards due to people playing in slo-mo.
3. Hacking is easier on PC, so they can more easily ruin online scoreboards.
4. Support costs for pirates.  PC indie game devs say the worst thing about piracy is not the money lost from people who wouldn't have bought it anyway, but the support costs for pirates which are 95% of their 'customers'.  That's 20x the customer base of the size who actually paid for and deserve support.

Further, I wouldn't care about things the PC can do that the Xbox 360 cannot, such as real online scoreboards, since I want a single source for both PC/XBox 360 to minimize bugs and development time.

Nick Gravelyn:

The problem really is the other way around. Many of us make games targeting the Xbox 360 and assuming use of a game pad, whereas on PC most users will expect keyboard/mouse controls. No, it's not a big issue, but it is something to consider before trying to release a game on the platform.

I do agree, it's just a design issue.

Kris Steele:

Correct me if I'm wrong, but isn't the Xbox 360 controller the only controller currently available that is Microsoft Common Controller Driver compatible? I thought I heard that mentioned once.

I have no idea: I'm a software engineer not a gamer so my sole job is to know about the technologies that I have to support. The info I collected about input devices are quite straightforward as follows:

Xbox 360: we manage the devices selectively to properly handle our game. The devices are enumerated and all have the same standard. No Mouse exists on the console. Chatpad are enumerated along with the gamepads, the USB keyboard is polled separately.We cannot detect the remote control nor gain access to the camera.

Windows: within XNA there is no chatpad support so we only poll the USB keyboard. We have mouse support and also gamepads and other devices MCCD compliant. To allow other devices we use DirectInput.

Given the above my personal strategy has been simple: for the PC platform I bought one original Xbox 360 wired controller, then I went to the local hardware store and picked up at random a few others cheap and common game devices of different kind (my desktop is a real mess now!). So I'm making sure that the all the input devices I have here are properly managed by my InputManager. I just poll all the acceptable devices and if I find more than one I ask the player to hit Start or the Fire button to play. Quite straightfoeward.

Cheers,
Pino

JasonD:

PC piracy may stunt Xbox 360 sales, since why buy it if you already have it?

You could do a delayed PC launch. I'm planning something similar with Pixel Man so as to not have the PC plans affect the high point of my 360 sales (when the game is new). I'm figuring a month or two out I'll execute on the plans for PC.

JasonD:

Hacking is easier on PC, so they can more easily ruin online scoreboards.

True, but there's not much you can do besides not do online leaderboards. You can do a bit of encryption or other little tricks to hide data when transmitting, but ultimately it'd take little more than some patience, Reflector, and a packet sniffer to start cracking open the transmitted data.

JasonD:

Further, I wouldn't care about things the PC can do that the Xbox 360 cannot, such as real online scoreboards, since I want a single source for both PC/XBox 360 to minimize bugs and development time.

While a valid point, I feel like failing to do something different is a waste of the platform. For Xbox, P2P leaderboards is the best you can do. But to do P2P leaderboards on PC seems asinine considering how incredibly flawed the technique is on any platform, but especially when you are completely free to implement a real online leaderboard (which can easily be run on a basic $50/year web host using some ASP.NET or PHP web service).


Some other things for people to think about:

1. Xbox 360 scales output. So you create a 1280x720 backbuffer and you're good on any display. On PC it doesn't work that way at all. You're responsible for at least choosing the proper resolution and expected to let the users choose a preferred resolution, all the while having a game that scales and behaves appropriately.
2. Other nice things on PC, for games that have some heavier graphical effects, is to enable different levels of effects (think Off, Low, Medium, Full) such that people can tweak the game to get their own preferred balance between quality and performance.
3. User permissions are a big thing when it comes to saving files. Too many indie games I've played save files in the directory where the exe is sitting. That works great if your users enjoy placing the exes in their user folders, but if I were to put some of those games into Program Files, they would crash immediately since they don't have write access in there (unless I run them as administrator, something I will never do for an indie game).
   So in my opinion when you go to save files, always use the user's App Data folder (can be retrieved through the Environment class's GetFolderPath method). You can save to their Documents folder (as some games and the XNA Framework's storage APIs do) but I personally hate the heck out of any app or game that feels entitled to messy up my documents folder when there's a folder specifically for application data. Maybe AppData doesn't exist pre-Vista (something I really doubt, but since I never used XP it is a possibility), but AppData should be your first choice for saving your game data IMO.
4. Work in social networking as it applies. You're on a PC so you can access the web. Try to work with Twitter or Facebook. Not only does this add some fun for your customers, it can act as viral marketing to attract new customers.

Nick Gravelyn:

So in my opinion when you go to save files, always use the user's App Data folder (can be retrieved through the Environment class's GetFolderPath method). You can save to their Documents folder (as some games and the XNA Framework's storage APIs do) but I personally hate the heck out of any app or game that feels entitled to messy up my documents folder when there's a folder specifically for application data. Maybe AppData doesn't exist pre-Vista (something I really doubt, but since I never used XP it is a possibility), but AppData should be your first choice for saving your game data IMO.

Save Files and Temp files are 2 different things.

Game save files e.g. Where I am in a game should be saved in My Documents. They are specific to a user and need to be in backups just like they are word documents. Vista onwards even provides a special folder for game saves see http://en.wikipedia.org/wiki/Special_Folders#List_of_special_folders.

Temp files e.g. caches, settings etc should be saved in app data since they are application specific and should never ben seen by a normal user.

There's a whole bunch of tech articles which includes how to vehave well on windows here http://msdn.microsoft.com/en-us/library/bb173039(VS.85).aspx and http://msdn.microsoft.com/en-us/library/dd607263(VS.85).aspx (Game Explorer). If you really want to go the whole hog then there's the windows logo certifactin guidelines which are long but useful to read about where you should and should not read/write stuff http://www.microsoft.com/whdc/winlogo/default.mspx. Some of its pretty petty but it does mean that your app stands a good chance of surviving OS upgrades in the future. In addition there is the Games For Winodws standards http://en.wikipedia.org/wiki/Games_for_Windows (sorry I can't find the main doc - maybe you have to be n the program to see it).

The ZMan:

Game save files e.g. Where I am in a game should be saved in My Documents. They are specific to a user and need to be in backups just like they are word documents. Vista onwards even provides a special folder for game saves see http://en.wikipedia.org/wiki/Special_Folders#List_of_special_folders.

The Vista game saves folder isn't actually in My Documents, but inside a sibling folder in the user's directory. I'm OK with using that, but I still disagree that games should put files into My Documents. I see the point about automatic backups, but I still get annoyed when programs start filling my documents full of folders and files that I don't want there.

Temp files e.g. caches, settings etc should be saved in app data since they are application specific and should never ben seen by a normal user.

Game saves are also application specific and, IMO, should never be seen by a normal user. :) If it weren't for the argument of automated backups, I don't see any reason why games shouldn't utilize AppData for game saves.