Sorry this is so looooong and rambling, but I have a lot to say and will explain it 13 different ways to make sure you get it.
I have a paid copy and I wouldn't recommend 3D Game Studio for the following reasons:
1. It takes a LOT of money to make it sorta kinda usable ... to get proper features you gotta buy the BIG bucks version.
2. The 3D modeler (MED) is basically useless.
3. It seems to be extremely difficult to get good looking results. This is not just me (blush) but from games others have produced with it. Mind, some have produced respectable games. But the feeling I get is that they are few and far between.
A better option, if you want to spend that much money would be to buy MS Visual Studio, and a really good commercial modeler.
But thats just me
01. I'd agree with point #1; however if you are in this to "make money" then that shouldn't be an issue if you're wanting to use a full realtime engine. I've been using 3DGS since '99, and I just finally got the Pro version this year when A7 was introduced. I got lucky and acquired an A6 Pro license, via resale, from someone on eBay. A license is allowed to be sold once and only once, and I got it for $300 which is waaaay cheaper than buying it direct. I had to pay the 1/2 rate price to get the upgrade, so while it wasn't a huge savings (plus I bought A4 Commercial and upgraded to both A5 and A6), it's better than nothing.
I'd highly recommend A7 Commercial instead of the Pro version. Really, the only advantages to Pro are unlimited multiplayer (which I could care less about) and the resource manager. You can obtain free; albeit less secure resource managers, though. The two really nice Pro features are full 4 bones per vertex FBX import and being able to use faster BSP scene management WITH CONCAVE geometrry (as opposed to 1 bone per vertex and ABT scene managment (adaptive binary tree)) 3DGS has always been a BSP engine, so moving it to Pro initially pissed me off big time, but it's not exactly the same BSP and jcl (the main developer) views it as a true Professional user feature; thus someone who is serious will want it given the speed advantage (plus you can use it with arbitrary geometry now and are no longer limited to convex only geometry or that's what we've been told...I've yet to test this). Even jcl is on record stating that Commercial is the way to go for most users who want just a little more than what the entry level/demo version offers.
02. I also agree with point 2 but it's main purpose has almost always been that of an import go between so you can get it into the MDL format. A7 allows MDL7 and soon FBX, so this is not as big of a deal anymore. Believe it or not, there are users who use it as their primary modeler and I suppose if you learn your craft on it it's not as bad as coming from something that costs hundreds or thousands of dollars (or even free like Blender and Wings).
03. Getting nice visual results shouldn't be too difficult if you are familiar with what it is capable of and how to set the proper settings to their best values for your given situation (this knowledge is acquired from reading the manual/forum and, of course, experience). It's a DirectX engine and uses the same DTX that all DTX engines use, so it comes down to implementation and optimization in the end (this is ultimately what you're paying for with those $500,000 engine). You don't necessarily
need an expensive commercial modeler. There is a 3rd party Blender exporter, and the engine can now import FBX (although there are still some issues that will eventually be fixed), and most modelers can export this format. Wings can and is free and well respected, although it has no animation capabilities.
One of the more "high profile" games and an adventure game at that is
Daemonica. It got fairly good reviews and a sequel is in the works. It's point and click but doesn't use the pathfinding program I discuss below. Here's Adventure Gamers review:
http://www.adventuregamers.com/article/id,619. I think one or two of the problems with it has actually been patched since this review (like the herb/map problem). Like any engine in this class, you aren't going to see a lot of "quality" games because a large portion of the users are hobbyists. They either don't finish games or aren't necessarily releasing content that is AA/AAA. I think Daemonica is a solid AA title. I think my own is shaping up, in its current form to be a solid AA.5 title.
It has nothing but time and will improve during that time. I have big plans.
One of the cool, adventures made with an older iteration of A6 is a mini-game called Adventure in Pocia. It was well received, and while it's direct control (classic style Resident Evil variety AND the mouse for certain tasks), it's a well put together experiment. Get the title here and click the EXE file for the playable version:
http://www.mediafire.com/?cojvuvzm2czShaders are the "new" way, and proper use of them is how you're going to tell one engine from another anymore but 3DGS can handle any shader you throw at it, and like anything else, if it's a sloppily coded shader it's going to run slow. It is safe to say it isn't going to run as fast as a $250-500,000 engine (both shader and the engine itself) but how many of us have that kind of money to spend on a 3D game engine?
There's a great 3rd party Shaders package called
Sphere 2.0. It's $40 and you get the source code. Sphere 3.0 is in the works. I'm assuming this will work in A7 (2.0 doesn't and won't be patched given the differences between the A6/A7 SDKs), plus it will probably have PS3.0 shaders and hopefully the existing 2.0 ones, too. These kinds of shaders are coming to A7 natively (see Forecast link below), but the quality/speed is unknown for now.
The problem with buying MS Visual Studio is that you have to know how to code. The beauty of something like 3DGS is that you don't necessarily have to know how to code anything, and even then it's a much easier to learn scripting language. Of course, we now have Lite-C which is a game/3DGS specific version of ANSI C, so the line has been blurred a bit. If you take the time to read the beta and forecast pages (see bottom of post) and the scripting parts of the manual, you'll see it's much more powerful than the previous language called C-script.
Anyway, I'm very happy with it and recommend it to anyone wanting a good 3D engine. Some nice adventure game specific features are making their way in and the 3rd party pathfinding tool called Intense Pathfinding just got a new version which featured a recompile to make sure it worked in A7, bug fixes, and added one new feature that I've been asking for forever that makes the player move to the closest legal spot on the walk plane if you click on a vertical surface (like in a professional P&C games). It simply ignored the click beforehand, but this new feature makes it very nice for P&C adv. games. Get the demo at the link below. It doesn't look too pretty, but it isn't supposed to be given it's a tech demo. Consult the manual for controls other than the typical point and click controls (like the keys used to move the camera around in the level). Click the EXE file to play with the tech demo:
http://www.intense-i.com/files/IntensePathfinding_v103.zipHere's most of the better looking screens from my title if you want to see what it is capable of. There are no shaders and my main player model is only seen in one shot (the dude with the hat and overcoat). If you scroll up 2-3 posts, you'll see another shot that uses an experimental style. You can read about that and previous posts where I discuss the "2.75D style" I'm now using. Think A Vampyre Story with it's extreme use of parallax scrolling. My game is still all realtime, but there will be lots of "forced perspective that mimics what AVS is doing with all 2D backgrounds. Take a look at the first half of that
Svengali clip from one of the earlier posts right before the screenshots one to get an idea of what this will look like (the layers effect):
http://www.coniserver.net/ubbthreads/showflat.php?Cat=0&Board=UBB20&Number=794524#Post794524Beta page (public and private beta version features):
http://www.conitec.net/beta/Forecast page (up and coming features):
http://www.conitec.net/forecast.htm