Trying to decide between C4 engine or XNA

So what languages are you familiar, have experience with, have made text games with, etc?

What is your experience with C++? What is your experience with C#?

Well I have about equal experience in both C# and C++.

I have used C++ with SDL in the past, and I have used XNA a little bit as well. I've made some simple games in both and have a couple books on each. If I choose C4 I'm going to pick up the books for that as well.

I'm not too worried about the learning curve for fully adapting to what I am going to use since I am about equally experienced, my main concern is with my development once I fully adapt to the engine or framework.

Quote:

Original post by spyridonz My only worry is development time. I have actually seen some people saying that developing a full game in XNA would be quicker then in C4? How is this possible?

You cannot compare C4 to XNA directly. That would be like comparing C4 to DirectX or OpenGL. The XNA Framework is just a thin layer of framework on top of the DirectX 9 API's. It provides a means to create portable managed code that will run on Windows, XBox 360, and the Zune platforms. A better comparision would be between C4 and Torque X. Torque X is a commerical (for hobbyists) engine built on top of the XNA Framework.

I do not have experience with the C4 engine. However, I do have alot of experience with XNA and other 3D engines. Glancing over the C4 website, it appears that the engine would provide significantly more functionality then the XNA Framework alone. If your goal is to build a 3D shooter game, something like C4 could be the way to go.

Without spending much time looking into the C4 engines license, it does not appear to be freely available(?). Have you considered looking into any of the freely available engines/libraries such as Panda 3D, Horde 3D, or Ogre (to name a few)? They may be a cheaper solution if your just starting off.

Quote:

Original post by spyridonz Anyone have any advice on how long development time would be

Honestly this is a very difficult question to answer. It really depends on a variety of factors. For instance, what type of game are you making? Is it 2D or 3D? How complex will the gameplay be? What technology will you be using? Will you be using a 3rd party engine, or will you be building one from scratch. Where will you be aquiring artwork, music, and sound effects from? All of these decisions must be taken into account in order to determine a time frame.

If your looking to build something quickly, you'll get the biggest bang for your buck using an 3rd party engine or library.

Hope this helps a bit.

Quote:

Original post by Billr17 Quote: Original post by spyridonz My only worry is development time. I have actually seen some people saying that developing a full game in XNA would be quicker then in C4? How is this possible? You cannot compare C4 to XNA directly. That would be like comparing C4 to DirectX or OpenGL. The XNA Framework is just a thin layer of framework on top of the DirectX 9 API's. It provides a means to create portable managed code that will run on Windows, XBox 360, and the Zune platforms. A better comparision would be between C4 and Torque X. Torque X is a commerical (for hobbyists) engine built on top of the XNA Framework. I do not have experience with the C4 engine. However, I do have alot of experience with XNA and other 3D engines. Glancing over the C4 website, it appears that the engine would provide significantly more functionality then the XNA Framework alone. If your goal is to build a 3D shooter game, something like C4 could be the way to go. Without spending much time looking into the C4 engines license, it does not appear to be freely available(?). Have you considered looking into any of the freely available engines/libraries such as Panda 3D, Horde 3D, or Ogre (to name a few)? They may be a cheaper solution if your just starting off. Quote: Original post by spyridonz Anyone have any advice on how long development time would be Honestly this is a very difficult question to answer. It really depends on a variety of factors. For instance, what type of game are you making? Is it 2D or 3D? How complex will the gameplay be? What technology will you be using? Will you be using a 3rd party engine, or will you be building one from scratch. Where will you be aquiring artwork, music, and sound effects from? All of these decisions must be taken into account in order to determine a time frame. If your looking to build something quickly, you'll get the biggest bang for your buck using an 3rd party engine or library. Hope this helps a bit.

Yep... I know they cant be directly compared =( which is alot of my problem lol.

C4 is not free, but upon liscencing you get free updates for life, which is one good plus about it. When it comes to engines in a reasonable price range, I think that is the one I would pick.

When I say "Advice on how long development time would be", I mean only time spent coding - not doing the artwork etc.

The reason I'm debating going for an engine is because it would save me all the time of making my own, so I could focus more on developing the actual game.

I was pretty sure this was the right decision for me, until I seen others say developing a game in C4 would take more time then it would to make a game in XNA....

Which is why I'm asking about all this... I dont know how this could be true. As you said, they can not be compared because C4 is a full fledged engine and XNA is just framework. Wouldnt it take a very long time to code all the things that C4 does in to XNA? I know C# is faster to code in then C++, but I dont see how it could be so much faster that it would make up for all the time it would take to create a full fledged engine...

I just dont want to spend the money for the C4 liscence, and then be dissapointed because I found out I would have been better off using XNA...

Quote:

Original post by spyridonz I was pretty sure this was the right decision for me, until I seen others say developing a game in C4 would take more time then it would to make a game in XNA...

Did you try asking the people in question why XNA would be faster than C4?

A few potential reasons:

• For a user who is more experienced with C# than C++ (and additionally may already have some XNA experience) C4 would obviously have more of a learning curve and take longer to produce results.


• For someone coming in as a near-absolute beginner C#/XNA have less potential pitfalls and "gotchas" to deal with than C++/C4; this type of person would also therefore probably get faster results from XNA.


• For someone who is making a game that isn't particularly well suited to the features C4 provides (I believe it's best for first or third person games and shooters in particular but havn't used or looked into it in too much detail personally) XNA may be a faster development platform because they'd be working around or against some of the additional features provided by C4 if they'd chosen that route instead.


• They might just be full of crap, or may be misinformed and havn't actually tried developing with both platforms. If they have tried both platforms they may have developed radically different projects with each (simple 2d game with XNA, complex shooter with C4) and are using that as a somewhat invalid basis for thier comparison. They have have tried C4 as a beginner and then XNA after gaining more experience thereby getting faster results.

As you can see, there are some cases where XNA may well be faster than C4, but there are also a whole bunch of cases (my last dot-point) that seem to suggest this but aren't really valid. It's really a person and project specific thing and can't be generalised to apply to everyone or neccesarily even to all projects by the same person.



If you're happy to licence C4 and have determined that the features it provides are suitable for the game you want to make then I'd personally say ignore input from others unless they provide some compelling reason to do otherwise and go ahead as you were originally intending - either option will get you there eventually, but only if you pick one and get to work with it.

Does that help?

Quote:

Original post by jbadams Did you try asking the people in question why XNA would be faster than C4?

Ahh.. well I seen the people say it on the forums, which is why I chose here as the place to ask it =)

Quote:

Original post by jbadams A few potential reasons: For a user who is more experienced with C# than C++ (and additionally may already have some XNA experience) C4 would obviously have more of a learning curve and take longer to produce results. For someone coming in as a near-absolute beginner C#/XNA have less potential pitfalls and "gotchas" to deal with than C++/C4; this type of person would also therefore probably get faster results from XNA. For someone who is making a game that isn't particularly well suited to the features C4 provides (I believe it's best for first or third person games and shooters in particular but havn't used or looked into it in too much detail personally) XNA may be a faster development platform because they'd be working around or against some of the additional features provided by C4 if they'd chosen that route instead. They might just be full of crap, or may be misinformed and havn't actually tried developing with both platforms. If they have tried both platforms they may have developed radically different projects with each (simple 2d game with XNA, complex shooter with C4) and are using that as a somewhat invalid basis for thier comparison. They have have tried C4 as a beginner and then XNA after gaining more experience thereby getting faster results. As you can see, there are some cases where XNA may well be faster than C4, but there are also a whole bunch of cases (my last dot-point) that seem to suggest this but aren't really valid. It's really a person and project specific thing and can't be generalised to apply to everyone or neccesarily even to all projects by the same person. If you're happy to licence C4 and have determined that the features it provides are suitable for the game you want to make then I'd personally say ignore input from others unless they provide some compelling reason to do otherwise and go ahead as you were originally intending - either option will get you there eventually, but only if you pick one and get to work with it. Does that help?

Actually, it does.

I just wanted to be sure that there was not something I was "missing" about XNA as I have not used it for about a year now.

It seems by the responses that things are much how I thought they were prior to reading the XNA comments.

I still have a few weeks to decide, but I'm pretty sure I'm still going for the C4 liscence. I liked what I read about it, with the only thing worrying me being development time. But if the majority of development is development on my actual game with much of the engine work done, then I do not mind much =)

I'm learning c4 at the moment and I say it has a steepish learning curve, not because it's in c++, but because the design is a bit quirky and there is insufficent documentation. On the plus side c4 has entire source code available and the source code is easy to understand.

Quote:

Original post by captainfreedom I'm learning c4 at the moment and I say it has a steepish learning curve, not because it's in c++, but because the design is a bit quirky and there is insufficent documentation. On the plus side c4 has entire source code available and the source code is easy to understand.

Have you tried the guide on the website as well? heres a link - http://terathon.com/store/index.php?main_page=product_info&products_id=22&zenid=84d37b254916d9c8f8157e1915a96b80

I plan on picking that up if I get the engine for sure, should at least help on all the basics up to your PhysX implementation.

Quote:

Original post by spyridonz Have you tried the guide on the website as well? heres a link - http://terathon.com/store/index.php?main_page=product_info&products_id=22&zenid=84d37b254916d9c8f8157e1915a96b80 I plan on picking that up if I get the engine for sure, should at least help on all the basics up to your PhysX implementation.

Yes, I have it. It's ok, but it's mainly step-by-step examples, and doesn't explain enough the architecture concepts

Quote:

Yep... I know they cant be directly compared =( which is alot of my problem lol.

You should of compared it with a XNA engine, like Torque X. But you didnt, so will go along with the engine vs framework lol.

Quote:

C4 is not free, but upon liscencing you get free updates for life, which is one good plus about it. When it comes to engines in a reasonable price range, I think that is the one I would pick.

What about updates forever(lil M$ joke there >.>) and its free? Thats XNA right there.

Quote:

The reason I'm debating going for an engine is because it would save me all the time of making my own, so I could focus more on developing the actual game.

But what if implementing those features arent that hard at all?

Quote:

I was pretty sure this was the right decision for me, until I seen others say developing a game in C4 would take more time then it would to make a game in XNA....

C++ coding,debugging takes alot of time and much more if you arent really experienced with it. C# is overall simpler to debug and code. Same with every other higher-leveled language, I belive.

Quote:

Which is why I'm asking about all this... I dont know how this could be true. As you said, they can not be compared because C4 is a full fledged engine and XNA is just framework.

I dont know what to say here, it depends on the engine and frameworks built-in features/code/whatever. But ask yourself, what has been done with the tools(C4 and XNA)? Make a few searches,see what others have done.

Quote:

Wouldnt it take a very long time to code all the things that C4 does in to XNA? I know C# is faster to code in then C++, but I dont see how it could be so much faster that it would make up for all the time it would take to create a full fledged engine...

See you hit something there, you dont want to make an engine, you want to make games. Well thast what you said. I bet C4 and other engines will have lots of features you wont even use or need for your game. And like I said above the language used will determine a bit of how long the dev would be, the simpler the faster usually. Something I noticed was that C4 has no built-in script language support, that means the game must be coded in C/C++(very bad), it does says visual script editor but I dont know much about that. You should check that out.

Quote:

I just dont want to spend the money for the C4 liscence, and then be dissapointed because I found out I would have been better off using XNA... I still have a few weeks to decide, but I'm pretty sure I'm still going for the C4 liscence. I liked what I read about it, with the only thing worrying me being development time. But if the majority of development is development on my actual game with much of the engine work done, then I do not mind much =)

You still have a few weeks to decide? Go grab XNA 3.0 and play with it for 2 weeks, see what you can do with it and how it goes. Try free products first give a trial to the comercial ones. Panda3D seems easy to use and pretty good and proven, should check that out. Dont go for the hype of what features you read about it, see what has been done and what others have to say about the tool before dishing out the cash.

Wish you good luck.

Quote:

Original post by 2DExtremeOwner Quote: Yep... I know they cant be directly compared =( which is alot of my problem lol. You should of compared it with a XNA engine, like Torque X. But you didnt, so will go along with the engine vs framework lol.

I was asking if development time to get XNA to do everything that C4 can do would really be faster then developing a game in C4... not comparing them =p

Quote:

Original post by 2DExtremeOwner Quote: The reason I'm debating going for an engine is because it would save me all the time of making my own, so I could focus more on developing the actual game. But what if implementing those features arent that hard at all?

That is exactly why I'm asking here... would it really take so little work to get all that implemented?

Quote:

Original post by 2DExtremeOwner Quote: Which is why I'm asking about all this... I dont know how this could be true. As you said, they can not be compared because C4 is a full fledged engine and XNA is just framework. I dont know what to say here, it depends on the engine and frameworks built-in features/code/whatever. But ask yourself, what has been done with the tools(C4 and XNA)? Make a few searches,see what others have done.

I did, but with so many unfinished projects, neither really stands out from the other very much. So sometimes I just have to go by the features pretty much.

Quote:

Original post by 2DExtremeOwner Quote: Wouldnt it take a very long time to code all the things that C4 does in to XNA? I know C# is faster to code in then C++, but I dont see how it could be so much faster that it would make up for all the time it would take to create a full fledged engine... See you hit something there, you dont want to make an engine, you want to make games. Well thast what you said. I bet C4 and other engines will have lots of features you wont even use or need for your game. And like I said above the language used will determine a bit of how long the dev would be, the simpler the faster usually. Something I noticed was that C4 has no built-in script language support, that means the game must be coded in C/C++(very bad), it does says visual script editor but I dont know much about that. You should check that out.

Ahh, yes well, somewhat... see I'm leaning towards NOT wanting to make an engine, but my primary reasoning behind that is because it would reduce my overall development time.

So it is kind of a lose-lose situation for me at the moment... If I go with C4, I have much of the engine finished and some nifty tools to work with, and I'm guaranteed new tools/features in the near future... but my development time will lack behind. If I go with XNA, I will have longer development time initially, but once the engine is in, my development time will be faster.

C4 has a node-based scripting system, but I dont know the full capabilities of it. I will be checking it out as I believe it is implemented in the demo. I read about projects implementing other, more traditional, scripting languages, but I dont believe they have any significant progress yet.

And I could go with other free engines of course, but I will get to that in a sec...

Quote:

Original post by 2DExtremeOwner Quote: I just dont want to spend the money for the C4 liscence, and then be dissapointed because I found out I would have been better off using XNA... I still have a few weeks to decide, but I'm pretty sure I'm still going for the C4 liscence. I liked what I read about it, with the only thing worrying me being development time. But if the majority of development is development on my actual game with much of the engine work done, then I do not mind much =) You still have a few weeks to decide? Go grab XNA 3.0 and play with it for 2 weeks, see what you can do with it and how it goes. Try free products first give a trial to the comercial ones. Panda3D seems easy to use and pretty good and proven, should check that out. Dont go for the hype of what features you read about it, see what has been done and what others have to say about the tool before dishing out the cash. Wish you good luck.

I have a lot to do these next few weeks lol. I'm currently making some basic models to do my testing with, then I'm going to look in to C4 a bit more, I have to see what updated in the 3.0 version of XNA, among trying the various free products.

About free engines, I have looked up Panda3D, but have not tried it yet either. The main thing slowing me down on this one was that I looked at what was made with the engine as you recommended. The commercial games (should) be pretty good representations of what the engine is capable of, and the ones made in Panda seem a bit dated or lacking to me. I dont doubt that it is a stable engine, and it is pretty impressive that it could handle MMO games, but I was never impressed by Toontown and PoTC lol. Although it is quite tempting due to the Python integration...

I'm trying not to go for the hype of the features, but as I said earlier, with so many unfinished products, sometimes I have to go by the features.

Seeing what others say about the tools - that's what I'm here for =)

Quote:

That is exactly why I'm asking here... would it really take so little work to get all that implemented?

You will only implement what you need to make your game. I dont know how C4 works but what if lets say everything is implemented from the get-go woudnt you have to unimplement some features? This needs to be anwered by more C4 experienced personnel as I am not qualified for this, im just giving a view to C4 and some engines. Maybe you should ask the C4 comunity also, they should have a forum and I have read that the owner is quick to answer/fix things. Emailing him would be a very good option.

Quote:

I did, but with so many unfinished projects, neither really stands out from the other very much. So sometimes I just have to go by the features pretty much.

http://en.wikipedia.org/wiki/Dead_Samurai Game in XNA I think made by 1 person or a small group in 4 months.

http://en.wikipedia.org/wiki/Quest_of_Persia:_Lotfali_Khan_Zand Game on C4.

Coudnt find much C4 games, XNA has quite a few but I usually find the names in game magazines so maybe a bit more googling around will get you more results.

Never go for features only, reviews and developments are more important. Devmaster does have alot of good reviews on C4. Check a few XNA based engines.

Visual3D.Net looks awesome:

http://www.devmaster.net/engines/engine_details.php?id=571
http://www.visual3d.net/game-engine

Torque X is also quite popular:

http://www.garagegames.com/products/torque/x/

For C# frameworks theres always XNA, theres Dark GDK.Net(which im personally waiting for and already reserved http://gdk.thegamecreators.com/?f=dgdk_net) and the SlimDX API.

Quote:

So it is kind of a lose-lose situation for me at the moment... If I go with C4, I have much of the engine finished and some nifty tools to work with, and I'm guaranteed new tools/features in the near future... but my development time will lack behind. If I go with XNA, I will have longer development time initially, but once the engine is in, my development time will be faster.

The only guarantee I see is that they will be free, nothing about new tools and features, if those happen then it will be free or should be by his claims. Also remember that its one developer, a good one but just one. He has much less to loose than lets say M$. Also with XNA you have Zune the 360 and Windows at your disposal with less problems.

Quote:

About free engines, I have looked up Panda3D, but have not tried it yet either. The main thing slowing me down on this one was that I looked at what was made with the engine as you recommended. The commercial games (should) be pretty good representations of what the engine is capable of, and the ones made in Panda seem a bit dated or lacking to me. I dont doubt that it is a stable engine, and it is pretty impressive that it could handle MMO games, but I was never impressed by Toontown and PoTC lol. Although it is quite tempting due to the Python integration...

We must be realistic here to make a good comercial game, overall, you must be either super skilled or have alot of money to spend an increadible amount of time and unbreakable motivation oh and almost forgot a team! Alone we aint gonna make the next Halo, the current Halo or the first Halo. The easier and the more abstract the tools we use the higher the chance we have of actually making and finishing a game and with less people to do it also. Time is money =P.

Edit: Had problems linking the sites, forgot how -.-...sorry.

[Edited by - 2DExtremeOwner on December 3, 2008 9:04:18 AM]

Quote:

Original post by 2DExtremeOwner Quote: That is exactly why I'm asking here... would it really take so little work to get all that implemented? You will only implement what you need to make your game. I dont know how C4 works but what if lets say everything is implemented from the get-go woudnt you have to unimplement some features? This needs to be anwered by more C4 experienced personnel as I am not qualified for this, im just giving a view to C4 and some engines. Maybe you should ask the C4 comunity also, they should have a forum and I have read that the owner is quick to answer/fix things. Emailing him would be a very good option.

For the most part, I think they would be alright. My preference is either 3rd person games (which I think would fit the engine quite well), or strategy games, and the book listed on the site comes with instructions on how to set the engine up for an RTS game. So this should not be too much of an issue.

Quote:

Quote: I did, but with so many unfinished projects, neither really stands out from the other very much. So sometimes I just have to go by the features pretty much. http://en.wikipedia.org/wiki/Dead_Samurai Game in XNA I think made by 1 person or a small group in 4 months. http://en.wikipedia.org/wiki/Quest_of_Persia:_Lotfali_Khan_Zand Game on C4. Coudnt find much C4 games, XNA has quite a few but I usually find the names in game magazines so maybe a bit more googling around will get you more results. Never go for features only, reviews and developments are more important. Devmaster does have alot of good reviews on C4. Check a few XNA based engines. Visual3D.Net looks awesome: http://www.devmaster.net/engines/engine_details.php?id=571 http://www.visual3d.net/game-engine Torque X is also quite popular: http://www.garagegames.com/products/torque/x/ For C# frameworks theres always XNA, theres Dark GDK.Net(which im personally waiting for and already reserved http://gdk.thegamecreators.com/?f=dgdk_net) and the SlimDX API.

I've played with Visual3D.Net a bit. Seems pretty cool, but I am waiting for a final release before giving it go again.

I've never used TorqueX but I have used TGE. I have always been worried about if they are going to support TX or TGEA in to the future, so that was one reason I was not sure if I should go for one of those.

One plus that C4 has over TX/TGE has been the updates for the engine, as those are typically not provided for the Torque engines. The only thing that I think I would miss from Torque in C4 is the networking code seems to be a bit superior.

I've never tried DarkGDK. I have used XNA a bit before so I'm a little experienced in it, but never used version 3.0

Quote:

Quote: So it is kind of a lose-lose situation for me at the moment... If I go with C4, I have much of the engine finished and some nifty tools to work with, and I'm guaranteed new tools/features in the near future... but my development time will lack behind. If I go with XNA, I will have longer development time initially, but once the engine is in, my development time will be faster. The only guarantee I see is that they will be free, nothing about new tools and features, if those happen then it will be free or should be by his claims. Also remember that its one developer, a good one but just one. He has much less to loose than lets say M$. Also with XNA you have Zune the 360 and Windows at your disposal with less problems.

According to the wiki, the next C4 update will include -

* A new system for large-scale voxel-based terrain rendering.
* A full set of new terrain editing tools.
* Visual shader graph editor.
* Dynamic shadow mapping for outdoor areas.
* New collision detection and rigid body physics support.
* Engine-level mover support (elevators, etc.).
* Dedicated server build option.

Which sounds nice for the engine, because Terrain tools were one of the main weaknesses of the engine from what I've read. Dedicated server build is nice as well, as networking was the major thing Torque had above C4 for me.

There are quite a few features planned for updates after that as well. Heres a list - http://www.terathon.com/wiki/index.php?title=Official_Roadmap

From what I've read, he provides quite a few updates like this a year as well.

Quote:

Quote: About free engines, I have looked up Panda3D, but have not tried it yet either. The main thing slowing me down on this one was that I looked at what was made with the engine as you recommended. The commercial games (should) be pretty good representations of what the engine is capable of, and the ones made in Panda seem a bit dated or lacking to me. I dont doubt that it is a stable engine, and it is pretty impressive that it could handle MMO games, but I was never impressed by Toontown and PoTC lol. Although it is quite tempting due to the Python integration... We must be realistic here to make a good comercial game, overall, you must be either super skilled or have alot of money to spend an increadible amount of time and unbreakable motivation oh and almost forgot a team! Alone we aint gonna make the next Halo, the current Halo or the first Halo. The easier and the more abstract the tools we use the higher the chance we have of actually making and finishing a game and with less people to do it also. Time is money =P. Edit: Had problems linking the sites, forgot how -.-...sorry.

lol I know... I'm not counting on making a good commercial game. If it happens that would be great, but I wont hold my breath. I mainly have a bunch of smaller ideas that I would like to toss together, and if one turns out to be interesting for me and my friends maybe we will form up as a team and work together on it. But as I said, not counting on it at this point.

I do like the idea of an engine with a pretty capable graphics engine, because although the games I make might not be long unless my friends decide to work with me, I want to try to make my games look good while they last, as I'm pretty good at art when I take the time to do it. 'Cause of this, I'm kind of hesitant on Panda3D.

I just mentioned those commercial games because they should be able to give you a good idea of what the engine is capable of, since they DO have the money and a team sufficient for it. I know the games I listed werent focused on a player base who had hardcore gaming PC's, but I wasnt too impressed nonetheless. Although I am still planning on taking Panda3D for a test drive to see what it is really capable of.

I've been having this "identity crisis" with choosing my language/framework/engine for a loooong time now lol. My first plan was to use C++ w/ Python scripting (for the nice mix of power and dev speed) and make an engine from the ground up - just for the know-how, but when it got to the point for me to learn DirectX, I decided to give XNA a go as it would be easier to learn... Which eventually led me to TorqueX but at the time it focused on 2d so I checked out TGE... which I read a book on but was reluctant to liscence, and then reading up on all the various engines out, here I am in my current confusion lol. I do like what I see in C4, its just that worry of added dev time... and Panda3D does seem enticing because it is along the lines of my original plans of using C++ with Python scripting. Hoping to get a real decision and pick up my liscence (if neccessary) this Xmas =) Maybe I'll just pick up a C4 liscence, and use Panda3D as well, depending on what project I am working on at the time...

Sounds like you made up your mind my friend, go with C4 and stay at it, if that fails and hopefully it wont go for Panda3D or other engines/frameworks and toolkits.

Good luck

late to the discussion :(

We use C4 at work. I've never used XNA but a colleague has.

Say you're writing a FPS and you want a real level. You get a bunch of content and import it into your XNA game. Your small little level runs at 5fps. Import that in C4 and it runs at 80 - because C4 is an engine designed with optimizations for entire levels of full high quality content in mind, and you would have to write/debug most of that (math intensive algorithmy stuff) yourself in XNA.

If you're writing a small 1-man-month game w/o any heavy game content & high end graphics, XNA will be just fine and might take less time (or it might not). If you're writing a bigger game (or want to learn the skills required to get hired by a studio) pick an engine.