Beyond 2004

By Andrew Herd (8 July 2005)

Flight Simulator With Guns? By "Mad Max" Merlin While a “CFS/FS merge would be a big culture change for older simmers, for whom the idea that Flight Simulator would never have guns has been an article of faith,” I have been lobbying our friends at Microsoft Game Studios for “Flight Simulator with guns” for years. CFS3 was a lot of fun but it certainly falls well short of a realistic combat flightsim experience. I’m not going to rag on CFS3 but issue a call to the community to let the Microsofties know whether you share my desire to have realistic combat aircraft modeled in FS10. With all due respect to the FS non-combat purists, I’d like to point out that there is no law that says you must fly military aircraft, discharge ordinance just because the sim supports realistic modern air combat. If you want to fly airliners and peaceful aircraft, you should of course do that. I like flying for fun as well as blowing stuff up. But, for those of us who like to light up targets and want to see missiles come off the rails, combining the robust realism of FS10 with modern air combat ordinance and exploding targets would be fantastic. After all, I too am a Microsoftie. I’ve been a die-hard FS and CFS flightsimmer since it was invented way back in the “Trash” 80 days. And I still prefer Microsoft Flight Simulator to all the other flightsims that are on the market today. Although I must admit a grudging respect for IL-2 and LOMAC and enjoy a mission or two with those sims. I would prefer flying both civilian flights and combat missions with a single sim. Even if it cost more to buy it. After all, I spent $50 for FS2004, $30 for CFS3, $50 for IL-2 and LOMAC. So, that adds up to $180 just to fly peacefully and also indulge my “blood lust”. Even if Microsoft combined the two and sold FS10 for say $80, it still would leave $100 in my pocket instead of Ubisoft”s I am convinced that FS with guns would yank the market away from IL-2 and LOMAC. Imagine the excitement of realistically flying a modern air combat mission. You have a very busy time on the flight deck in modern air combat. After shooting your F-14 Tomcat, F/A-18 or JSF fighter-bomber off the deck of an aircraft carrier, you need to navigate to target and coordinate with AWACS and JSTARS en route before you can lock target and fire your Phoenix missile. This puts a whole new spin on combat ATC and it would be a challenge to model, but a delight to fly. FS with guns would also be a bonanza for developers, both payware and freeware. While we already have fantastic military freeware such as Kirk Olsson’s F-16 and even more fabulous payware such as the F/A-18 from Aerial Foundry, the F-104 Starfighter from Captain Sim and many others – the ordinance is all show, no go. I want to dodge triple-A threats and SAMs that are really coming up at me. If I’m flying an F-117 Night Hawk, I want to feel what it’s really like to evade threats behind a hole in the sky. When flying a Tomcat, I want to lock target, see the missile come off the rails when I call “Fox 1” and watch my target go down in flames. Imagine the artistry required to not only model the airplane and panel, but also to faithfully model the laser guided weapons, JDAMS, A2A, A2G, A2S missiles, JSTARS missions, AWACS ATC and all the rest. You can get a taste of it in LOMAC, but I think FS10 with guns could take us to a much higher level of realism both in the cockpit and on the ground. While flying historical missions over Europe and the Pacific re-creating World War II missions is always a thrill, I want to fly today’s air combat missions in today’s modern combat aircraft. And not just the jets. A truly realistic sim of Apache Longbow and Black Hawk helicopters would be sensational. So far, all attempts at doing that are honorable mentions but no one has truly nailed it. What’s more, if you have it all in FS10, the convenience alone is worth the price of admission. I would like to experience what it’s really like to “go downtown” over Baghdad or take out a nuke facility in Iran. Or, at least experience those missions as real as it gets with Flight Simulator. Flying a CFS3 B-17 through German flak so thick you can walk on it is always a serious sim challenge, but doing it without working panels makes it a spectator sport and flying air combat without using weapons is like kissing your sister. I would like to experience what it was like to go downtown in a Phantom or Thud over Hanoi and Haiphong during the Vietnam War. Or drive a BUFF over the North during Operation Linebacker. I want to fly the Bone (B-1B) on nape-of-the-earth low-level penetration missions and see if I can put 5000 pound bunkerbusters on target. In this veteran flightsimmer’s opinion, there is really only one team that could deliver that sort of extremely sophisticated flightsim realism. Modern air combat is complicated stuff and I cannot wait to see what our friends at Microsoft could do with all that if they choose to take this old flightsimmer’s advice. For the Microsofties who read this, I have lots of emails from avid Flight Simulator fans who think the same thing. And for those flightsimmers who read this, let the Microsoft team know if you agree that FS10 should be FS with guns. Just send me your comments and opinions and I’ll make certain the Microsoft teams gets them all. “Mad Max” Merlin maxmerlin@flightsim.com

FS2004 is an excellent simulation of flight, with a super weather system, good air traffic control, some fine planes, reasonably adequate ground scenery and an 'open' interface which has encouraged a vast addon market, a major reason for many of us buying the sim in the first place. So far, so good, but the way Flight Simulator has been developed over the years has led to it becoming a curate's egg of a program - which is to say that some parts have ended up much better than others. This has happened because resource constraints have meant that development has had to be incremental; few versions of FS have ever done more than build on the features of their predecessors.

The development team for the next version faces a difficult task. For one thing, there aren't enough of them, because although Flight Simulator sells steadily, it has never been a killer app. To make matters worse, sales have been seriously dented by rampant piracy that shows no sign of abating. Piracy is such a problem with FS that it could conceivably kill the game off in the end - and it is hard to see any solution to this beyond people who use bootleg copies being more honest with themselves about the consequences of their actions. If you and I didn't get paid for going to work, we wouldn't go, and neither Microsoft, nor any of the myriad third party developers whose businesses are linked to the game think any differently. People who use ripped copies make all kinds of excuses for why they are too mean to pay for their pleasure, but I doubt many of them would unquestioningly pay for the gas to let a stranger use their vehicle, let alone saying it was OK for the guy to lend it out to all his friends too - there is no Robin Hood principle at stake here, piracy is theft, pure and simple. If it stopped, Flight Simulator would be more profitable as an app, which would fund a bigger development team and we would have a better game, so if you have a ripped CD in your drive, you can do something positive by going out and buying a genuine copy today.

Another problem the team faces is that some of the stuff that really needs attention isn't that sexy, doesn't get noticed by the majority of users and would be danged awkward to change. The flight model is a perfect example: the .air file/cfg system is creaking with age and in dire need of a major overhaul, but doing so risks making every addon obsolete overnight and any voices championing the improvements would be drowned by howls of rage from thousands of users denied their weekly fix of [insert your favorite addon plane here]. But, sooner or later, flight modelling has to be addressed, because there is an increasing mismatch between the sophistication of the planes' panels and systems and the way they behave in the virtual skies, the priority area being the extremes of flight. Given more realistic stalls - particularly accelerated stalls - Flight Simulator could become a standard part of every student pilot's training. And then there are spins, which carry enough risk to be a minority pursuit in real life and can be endlessly practiced in the safe sandbox of a simulator; yet very few FS planes will spin at all and those that will recover the moment the controls are centralised (interestingly, Il2 has the reverse problem, many of the planes being far too easy to spin and almost impossible to get out again). If the base code behind flight modelling could be upgraded, then planes with great handling like the RealAir Spitfire XIV in the screenshot would be the rule, rather than exceptions.

Now I have got started on the subject, there is the bugged P-factor, which has needed fixing for several versions now; and the weird way that FS planes show an initial rate of climb several times higher than the one they are capable of sustaining, a characteristic many pilots have wished were true of real planes as the distance between the prop and the hedge shrinks while the bird decides whether to fly or not.

Bound up with FS aerodynamics is powerplant modelling. For many years, Flight Simulator offered a choice between piston engines and jet engines and that was that. Then, when the King Air made its debut, we got a turboprop. Third party developers hate the turboprop model because it is essentially a bastardised version of the jet, which makes turning out decent turboprop sims tricky because the engines can't easily be made to work the way they should. We badly need a fixed turboprop - all the basic powerplants could do with an update, but the turboprop is the priority.

FS sound could do with an update to bring us into the surround sound era. Regular readers eyebrows will have shot off the tops of their heads as they read this, given my relaxed views about sound sets, but the major reason for my indifference is that Flight Simulator handles sound in such a basic format that it dooesn't matter how good the original recording is, it always ends up being indistinguishable from any other plane. A 5.1 sound option would make this area a whole deal more interesting.

Panels are a huge issue. FS cockpits have sat in limbo for the last couple of versions, largely because virtual cockpit graphics and ergonomics haven't caught up with their 2D equivalents. The result is that almost every FS plane ends up with two sets of panels - a 2D and a 3D set, which is an unnecessary waste of system resources, as the CPU has to service the demands of large amounts of code that isn't being used. If Microsoft didn't have to cater for sub 2.0 Ghz CPUs and older graphics cards, the 2D cockpit could have been thrown out after FS2002, but the fact we still have it is bound up with backward compatibility and hardware specification issues. The next version of FS is probably the first one which in which Microsoft can assume users will have excess graphics power on tap and I would love to see the gloves taken off and high resolution true color screen display allowed for the first time. The prize is photoreal virtual cockpits (VCs) - which would at long last let the VC rival the quality standards of the best 2D panels we see from developers like Flight1, Captain Sim and DreamFleet.

The last few versions of Flight Simulator have brought new classes of instruments with them, or at least substantial upgrades of existing ones and the obvious candidate for FS2006 is a Flight Management Computer (FMC) for the Boeings. The default GPS has been developed to a high standard and I have lost count of the number of postings in the forums highlighting the need for a basic FMC, so what about it, Microsoft? The alternative, which I'll admit is tempting, would be to produce one of those high res modern displays like the Garmin G1000 glass that new Cessnas are fitted with, but unless the team have more time on their hands than I think they do, we can have one new type of gauge and the FMC is has the votes.

Every time I use FS2004's air traffic control I marvel at how good it is and think how nice it would be to hear some different people. Much though I have appreciated hearing the dulcet tones of the FS team over the past couple of years, it would be add so much to the atmosphere if time could be found to double the number of voices and introduce some European, Australasian, African, Russian and South American accents, so that the FS skies could be truly international. There are a few well known bugs in the ATC, which I assume will get fixed, but an enhancement that would make good sense would be a virtual co-pilot along the lines of the ones seen in recent third party addons. The idea of being able to rely on some help with the flaps and gear on the trickier IFR approaches is too tempting to resist.

Flight Simulator's sky is well on the way to being as real as it gets, but it still has some way to go. From a user's point of view, the problem is that the graphics are good enough that it isn't hard to imagine them being nearly perfect, which would mean additional cloud shading options and some additional cloud layers, if FS2006 is to match reality. Clouds are important because simmers have nothing else to look at out the windows most of the time and good though they are, FS2004's clouds don't quite do it for me. The shapes are good, but FS clouds don't show the degree of color variation that real ones do; take a look at any post-cold front sky and you will see a scene that FS2004 cannot render, because it can't display that amazing mix of dark and light cumuli in a single layer that the real world can. Lack of layers and cloud types means that another sky to which Flight Simulator cannot do justice is an approaching front, that beautiful, but ominous mix of cirrus and altocumulus which heralds lowering skies and rain. Fronts are such a feature of flying in most of the world that it would be great to see them simulated better - can you imagine starting up FS2006 and seeing a sky like the one above?

I know it is a really old chestnut, but can we have a fix for taxiing? Real airplanes need a certain amount of additional power to start moving, but they don't leap off like scalded cats and make the pilot yell "What the hey!" and chop the throttle so he kills the roll and has to go through the pain all over. If this was the case in reality, ground sickness would be more of a problem than air sickness - with all those realistic taxiways and ground ATC, I think our virtual passengers deserve a break.

Which leads me to the question of runways. Flight Simulator's trademark flat gray rectangles were state-of-the-art once upon a time, but that was when we all thought FS5 was as good as it was going to get. Gray and flat was passé in FS2002 and looks archaic in FS2004, especially when you can land almost anywhere in the FS world (okay, okay, I know we can have green rectangles too). Way back, the gray rectangles had a purpose - they were the only place you could land and get away with it - but now we only have them because we have them and it is time for them to go. I appreciate they have all kinds of special properties, but there is no reason why those properties can't be attached to ordinary terrain tiles, which would bring the default runways into line with photoreal scenery runways. If that was the case, we could at long last escape from having all the default runways absolutely straight and level - real runways often have slight but noticeable slopes from one end to the other and even from side to side. A few of the braver third party scenery developers have broken the mould, with releases of gray rectangle-less airports going as far back as FS2002 (try Courchevel if you want to see one of the more extreme examples), but the time has come to update the standard here.

Now another thing about runways is that they behave differently depending on the weather. If it rains, planes take longer to stop; if it snows, they stop too quick; if it ices, they don't stop at all. Grass is different - very different. Even well cut grass is bouncy compared to pavement, but wet grass is skiddy as well as bouncy and long grass can spoil your entire day. Now I am not suggesting that every default field has to have its own grass cutting cycle, but it would be a lot more real if the rain 'falling' from FS clouds wet the runways and altered the braking action and if we had icing simulation (thinking around that, we could have it on the planes too, but maybe that could wait for another version). The modern generation of graphics cards is definitely up to displaying this kind of stuff.

Next, the scenery tiles. Flight Simulator has come on a long way since the wireframe world days - we have incredibly real mesh, cool AutoGen and numerous photographic sceneries. The fly in the ointment is that scenery tiles display at a relatively low resolution and, much worse, they get such low priority in FS2004's order of doing things that they only shuffle into their final position at the very last moment. This bugette, feature, call it what you like, causes problems if you fly fast over mountainous areas in the default scenery, let alone with photosceneries, with periodic attacks of annoying blurring that persist for thirty seconds or even longer. I would just love to see the last minute shuffle fixed so that photosceneries didn't blur any more - an upgrade to the display engine that would allow tens of thousands of simmers serious enjoyment of flying over perfectly realistic photographic landscapes without fear of being let down by the display engine. More than that, it would make FS2006 dangerously attractive to flying schools in the many areas which are now covered by photoscenery packages - learning VFR nav using photographic scenery is a whole lot easier than doing it the hard way. I would also put in a vote for having higher definition ground tiles; at present FS2004 cannot display resolutions of greater than 4.75 meters per pixel, whereas aerial photography gives us resolutions down to 30 centimeters a pixel. Okay, so increasing the resolution by a factor of five would result in huge increases in file sizes, but modern processors and hard disk sizes could easily handle a doubling of ground tile resolution.

AutoGen works well and it looks great from above a couple of thousand feet, but it still has some rough edges. The first is that it tends to crowd inappropriately close up to the runways of the default airports, particularly in rural areas. Come to think of it, the same happens in urban areas too and depending on display settings, I have found houses a couple of hundred feet from the centerline of 6000 foot runways before now, a situation which never happens in real life. Well, hardly ever - I still treasure the memory of the lady waving from her bathroom window that day we made the low, slow exit from Leeds Bradford - I would have waved back if I wasn't so worried we were going to join her. The one other problem with AutoGen is lack of variety and the way the houses always look like they are perched in place. I don't know how much effort would be required to fix this and make buildings blend into their surroundings more, but it would seriously improve approaches. Desirable, but low priority.

Planes. Oh, yeah, those. It used to be that the FS set was the FS set and every edition brought a single new aircraft at most. FS2004 stood tradition on its head and gave us all those wonderful old-timers, but as far as I can tell, most of them lie gathering virtual dust on hard disks around the globe. It is hard to argue with the choice of the FS2004 oldies, but if development time is to be concentrated where it counts, the only ones I would argue any kind of case for keeping are the Cub and the DC3; much though I have enjoyed trying my hand at the Vega, the Comet and the Vimy. I have a soft spot for the Trimotor, but Flight Simulator's lack of period scenery and limited opportunities for VFR navigation puts it at the back of the queue every time. The Cessna and Boeing panels are in dire need of better graphics and if we are going to have a new plane, there is a world of choice out there - the GA set would be considerably enhanced by a modern low wing single like a Cirrus or a Diamond. The Cirrus is attractive because of its state of the art panel, but developing the gauges might deprive the hoped-for FMC project of resources, while a diesel engined DA40 would be definitely be an interesting addition to the default hangar. Then again, FS choppers are hardly sophisticated and this area has long been ripe for an upgrade, although coming up with a convincing way of simulating helicopter control systems has so far evaded the developers.

With the demise of Combat Flight Simulator, it is always possible that we might be in for some dramatic changes, with a fusion of the two engines and wholesale import of CFS features into FS. Apart from the interesting possibility of using AI traffic as target drones, a CFS/FS merge would be a big culture change for older simmers, for whom the idea that Flight Simulator would never have guns has been an article of faith.

All things change and it will be interesting to see what the future brings.

Take it to the forums?