 |
|
|
|
|
FS2004 came out in a blaze of publicity, with Microsoft hosting journalists at launch events around the globe. I remember being flown to the Paris Air Show to listen to a presentation about how the new version would change our lives - we were all given a chance to try and fly a Flyer hardware sim, which left me with an enduring respect for the Wrights and greatly relieved that wing warping never caught on. If wing warping didn't set the market alight, neither did the 'Century of Flight' theme, but in retrospect, FS2004 was one of the most stable and satisfactory versions of Flight Simulator ever released and it had the great virtue of running faster on the same hardware than its predecessor had. To old hands like myself, the idea of a new version of Flight Simulator being quicker than the old one was little short of astonishing, because the entire history of the program up to that point was that when the new version came out, you threw away your existing PC and spent a shed load of cash on a new one. FS2004 managed to confound the critics and lull us into a false sense of security at the same time, largely because it was around for so long -- previous versions of the sim had a lifespan of roughly two years before they were replaced, while FS2004 lasted three. This was bad news for simmers reared on a diet of constant change, but good news for the rest of us, particularly the third-party developers, because it meant that their products had a long sales life and it is no accident that FS2004 supports a thriving - and loyal - add-on market.
Two years is a long time to wait if you are an end user, but perilously short if you are a developer, because every time a new version of Flight Simulator appears, there are delays getting the software development kits out, which are vital if third party developers are to create reliable products. Herein lies a problem for us all, because Microsoft only makes money out of add-ons for Flight Simulator if the add-ons trigger more sales of FS and in the lack of any evidence that this happens to a significant extent, SDK development is a charitable act - although it generates a tremendous amount of goodwill, for where would we be without add-ons? Now it is traditional for everyone to whinge about the length of time it takes Microsoft to release the SDKs, but when you bear in mind that they don't have to release them, it is good that they appear at all and in general, the important ones are normally out within six months of the launch. Once they are armed with the SDKs, third-party developers can start work on creating bug-free products that are completely compatible with the new version of FS, which usually takes another six months, leading to something like a one year delay between the release of a new version of the sim and the arrival of a decent supply of add-ons specifically codede for it, which is where we are now with FSX. Sure, there is usually a trickle of patched add-ons before the year is up, but these are by guess and by God releases and in my experience the really good stuff seldom appears much earlier than a year after a new version of FS hits the shelves. You take your chance with the early releases, although simple add-ons usually don't cause too many problems.
 |
 |
 |
This one year delay superimposed on a two year release cycle leaves add-on developers with a remarkably short time to make a profit, bearing in mind that rumors about the new version of Flight Simulator are usually begin to circulate eighteen months after launch of the 'old' one and act as an increasing brake on sales. In the past, given all the challenges it faces, it has surprised me that an add-on market for FS exists at all, especially when you consider the short selling span a product has before it becomes time expired and also take into account the rampant piracy that plagues the hobby. Truly, it is an unenviable task making a living out of Flight Simulator add-ons and it never surprises me when developers become tired and emotional and go off to do something more logical, like teaching polar bears Tai Chi. FS2004 saved a lot of bears from having to learn the art of relaxation, because it gave developers long enough to upgrade their old products, develop new ones and generally make enough money to keep the wife and feed the kids.
If FS2004 was different, FSX consolidates on some of the stuff Microsoft learned with that release as well as taking a giant step back into the past. Dealing first with the step backwards, FSX returns us -- big time -- to the days when the release of a new version of Flight Simulator meant that you had to buy a new PC. No, come to think of it, this is worse than that, because Aces (who coded the game) make no secret of the fact that FSX has been written for PCs that don't even exist yet and that at the time of release there was no hardware available that would allow FSX to be run satisfactorily with all the display sliders pulled to the right. This step backwards has been compounded by FSX being developed to run at its best on a new operating system -- Windows Vista, which had achieved hardly any market penetration at the time of FSX's launch and is still something of an unknown quantity in some respects even now.
The consolidation has lain in Microsoft's learning that if it gets its strategy right, a two year development cycle isn't compulsory and that three, or maybe even four years might be possible. In other words, FSX has been so designed that it will only achieve peak performance on hardware that won't hit our desks until a year or two from now, which will effectively extend the life of the product by allowing users to 'rediscover' it every time they buy a new PC. You may or may not agree with this way of thinking, but that is the way it is and, if you read on, as a philosophy, it has its plus points.
Needless to say, there are problems. One, which we can get out of the way fairly quickly, is that (in my experience) FS2004 doesn't run any better on Vista than FSX does on XP. This means that, for the first time since FS95, the operating system you run makes a difference to how well Flight Simulator runs. The bad news is that while FS2004 is stable on XP, two service packs later, FSX isn't exactly content on Vista, which isn't a total surprise, given that Vista has troubles of its own. Not only that, but DirectX 10, which theoretically could remove at least some of the performance bottlenecks, isn't fully supported by FSX yet, although Aces included a 'preview' mode with the Acceleration pack.
But, of course, DirectX 10 won't run on Windows XP, which means that if you are running FSX on an old XP system, you not only need a new PC, but a new operating system to go with it, in order to run a graphics sub-system that itsn't supported yet. Ho, hum.
 |
|
 |
Oh, yeah, then there is multicore. When Aces began the development of FSX back in 2003-4, it wasn't clear that multicore processors were going to be the future and by the time it was, in 2005-6, so many major decisions about FSX had been taken that only limited accommodation of multicore could be made, but SP1 moves terrain loading, terrain relighting and Autogen batching onto multicore if is available and SP2 expands the Autogen batching further. This means that there are benefits to having a multicore processor, though second and higher core processor load will rarely achieve 100% and tends to average out around the 20% level, depending on what you are doing. In a nutshell, the improvements in SP1 and 2 come into their own as far as multicore systems are concerned with ground texture handling and Autogen, but there is still a need for raw processor clock speed to get the best out of the sim.
So FSX is a problem child. It came along at exactly the wrong time, facing major changes in the operating system, graphics sub-systems and processor architecture and I guess we should be thankful that it runs at all. Aces deserve their due, they have definitely persisted and we have two service packs more than most of the earlier versions of Flight Simulator ever got. The kicker is that FSX has been designed to have a long shelf life, which means that this suboptimal situation is not going to go away any time soon and I wouldn't bet the farm on seeing FS11 much earlier than late 2009, perhaps 2010 and maybe early 2011. By then, Vista should be stable, DirectX 10 will be well understood and for all I know, entry level systems will have six cores; but FS11 does not help us with FSX.
The jungle telegraph works well among flight simmers, thanks to the forums on sites like this one. When FS2004 was released, people rushed to buy it and the FS2002 add-on market collapsed virtually overnight; when FSX came out, users knew it was trouble and many stayed with the devil they knew. The result has been a situation unprecedented in the history of Flight Simulator, which is that well over a year after the release of a new version of the sim, we still have a healthy supply of new payware add-ons appearing for the old version of the sim and freeware add-ons are in relatively short supply for the new one. For months I have been watching the Pilot Shop with something approaching disbelief, as I can't ever recall anything like this happening before, unless you count the overhang which occurred with FS98, which happened because FS2000 was such a dog. You ready for this? Right now, sales of FS2004 add-ons are matching those of FSX add-ons, something like fifteen months after FSX was released.
The message is that simmers are staying away from FSX in droves and as a reviewer, I find it easy to understand why, because although there are always howls of anguish in the forums when a new version of Flight Simulator appears, on this occasion the howls have been deeper and longer than usual, because -- by definition -- there isn't anyone out there who can run FSX 'properly' and everyone who owns a copy is a potential complainant. The harsh fact is that it doesn't matter how much money you spend on a PC, FSX will not, at the time of writing, let you run it with all the sliders maxed, while you fly a complex add-on airliner into a complex add-on airport on a foul day with real weather enabled. No machine exists that can do that and we have the authority for this of none other than the developers.
 |
|
 |
So here I am, running FSX on a 2.66 Ghz dual-core with 4 gigs of RAM and a stacked video card and the mother takes forever to load and when it does, is sometimes disinclined to stay loaded. As for running AI traffic, well forget it, brother, because AI does worse things to FSX than the pulling the plug out the back of the PC. Despite this, I run FSX in preference to FS2004, which I have hardly touched since the new version of the sim was released. Why do I do this to myself when so many others have chosen to stay with FS2004? The short answer is that I am a reviewer and reviewers get what they get to review, which has mostly been FSX products and in the course of using them, I have developed a soft spot for FSX. But the real reason is more complicated than that -- FSX has spirit and on top of that, I have been using Flight Simulator for so long now that I have come to respect Microsoft's loyalty to it in a world where products come and go so fast that three year warranties are becoming meaningless. The first time I ever used Flight Simulator, I ran it on a Tandy TRS-80 -- that's a long, long time ago now. So from where I sit, FSX is just another chapter in an evolving story and like all long stories, some of the chapters have made better reading than others and the tale has had its good, bad and just plain average moments, but there has always been something to look forward to.
So, some examples of things to look forward to. The add-on that triggered this editorial was the Captain Sim C-130, which has a long story of its own, but runs really well on FSX, just as long as you remember not to overdo those sliders. The C-130 fits into FSX's ambience like a glove and flying it from the virtual cockpit makes me feel as if I am actually there, at the controls of a real Hercules. Another example of a great plane for FSX is Aerosoft's Twin Otter, which looks a million dollars, even if it needs some tweaking to make the VC skin quickly - buy that plane and fly it into the same company's German Airports 1 add-on scenery and FSX will give you the sort of experience that FS2004 would struggle to provide, given that the latter cannot display ground textures any finer than five meters a pixel. Much the same can be said about PC Aviator's MegaSceneries, the cream of the crop so far being their Southern California pack; again, this is an order of magnitude better as an experience than anything FS2004 has to offer. If vintage warbirds are your thing, then Shockwave's P-40 for FSX is the pick of the bunch with Flight Replicas' Bf 109F chasing it hard - again, both fit FSX like a hand in a glove and staying with oldies, take a look at Aerosoft's Dornier 27 for FSX. All of these add-ons are great because they take advantage of features that only FSX can offer - in addition to all the hard work that lies behind the code.
At the same time as all these great FSX add-ons have been being released, there has been competition, in the form of great FS2004 add-ons. If you can live within the confines of a flat earth with an atmosphere that ends at 99,999 feet and a resolution of no more than 5 meters per pixel, FS2004 is a great simulation, not least because you can run it flat out with all the sliders maxed on existing hardware. In the year since FSX was released, we have reviewed some fantastic FS2004-only add-ons, including the freeware Netherlands 2000 V3, the DreamFleet Piper Archer III, and the Aeroplane Heaven Avro Lancaster special ops pack. Needless to say, there is an increasing trend for developers to release add-ons which are compatible with both FS2004 and FSX, and FS2004-only add-ons are on the wane, but then we aren't seeing too many FSX-only add-ons either.
The good news is that if you like FS2004 - and let's face it, with all the add-ons that are available, you can customise it just about any way you like - there is no shortage of goodies on offer. FS2004 is dead? Long live FS2004, the customer says. But, for all the problems that we face with FSX now, the same thing is going to happen to this new version of Flight Simulator, when, a couple of years in the future, it becomes the old version. Like FS2004, FSX is going to be around for a long time, at least another two years, which is good news, because it will give add-on developers plenty of time to upgrade their existing products and to develop new and even more exciting ones. There have been some welcome improvements in multicore usage with the service packs, but on the whole it seems we will have to wait for Uncle Intel to cure the frame rate problem by building us more powerful processors, which we will go out and buy, so we can run this magnificent program. So some time in 2010, by which time I will have reviewed another 150 add-ons or so, to add to the 300 plus I have seen to date, I expect at long last to find myself able to fly an FSX approach into a complex airport, in a complex airliner add-on, using 100% AI, full Autogen and the highest display settings.
But I won't be happy, because by then, I will have FS11 and it will need an even more powerful processor... or not. We will just have to wait and see.
Andrew Herd
andy@flightsim.com
Discuss this
in the Outer Marker message forum
Copyright © 2008 by FlightSim.Com. All Rights Reserved.