FSX 3rd Party Issues:Early Warning

[Coral Sea 0]

Hi Phil, thanks for answering me.I'm still trying to figure correctly what you've posted on some phrases as English is not my native language.As an additional consideration, you may not know about this yet, but I would assure you for sure that a huge large number of developers (like me) fear the approaching release date of a new FS as a terrible nightmare to be faced on :-) without knowing if all things we've done so hard will be still useable or become completely useless.Think about my own experience last year (don't believe it has been very diferent to other guys): by early/mid September we started to be bombed by hundreds of customers mails and forum messages asking wether our (their) sceneries would be or not compatible with FSX, if we would make the changes required, if we would charge for them or not, and so on. This alone is already stressing enough and let our atention away from our work, even more that we really didn't know the exact answers to give them.By the time of the FS2002/FS9 transition affair, things were smoother and we could manage to update (and enhance) all sceneries to the new sim and let them available for all users for free. We had a hope we would do this again but the reality hit us hard and we had to explain that some way to the disapointed customers. The amount of work/time demanded to re-author everything requires that we just can't afford doing this update for free anymore, as every single scenery has became a truly NEW product comparing to the FS9 version done before. The only way to keep customers happy was offering nice reductions on prices for ancient users of original versions, as we are doing.The worst part of such a problematic transition is that, for the ones that depend on this work to support and mantain a family, like me, is terrible to watch the falling downhill movement of orders since September.By this time, FSX was the talk of the moment, and users were haulting further orders as they thought that new native FSX products were right around the corner, easy to be done and released or just updated, and it was just a matter of some weeks after FSX appearence to start ordering such adds.That is: since October very few users were placing orders to further FS9 adds, as nobody wanted to waste his money on old-fashioned FS9 products, waiting or the brand new FSX adds. By mid November/early December, there were still few orders for FS9 adds, and FSX was shocking simmers with the issues, so a very small number of them were ordering FSX products fearing their impact on frames, but our families were still to be maintained and our stress level were the only thing growing up :-) side-by-side with our charge of extra work.We were again lucky to decide on making every single native FSX scenery compatible with FS9 too as a new version, and let two independent installers avaliable free of charge to our customers. This is what has saved and allowed us to stay on market, but at what a price of tension and distress.Now things are going back to normality, at least for us, but I can't say the same to a large number of good and nice developers that are still working on the same and suffering as we had. FSX has really turned the market upside down like an earthquake, and what we all fear is that the next one has already been predicted and is scheduled, his name is FS11 :-)I really like and use FSX, our sceneries look great on it, in fact even the impact of them over the default ones was noticeable smaller comparing with the same impact FS9 versions had over their defaults, don't get it the wrong way. What we fear is the effects such Sim releases causes. Notice for a fact: this 2 years life-cicle for a sim is a very small one. When things are just becoming good and nice for eerybody, the sim is tamed, developed to all personal tastes and preferences, all adds are enhanced, updated and enjoyable to their limits, everything perfect... then comes a new Sim and causes all this on us.Look at FS9. For staying alone for 3 years instead of 2, their adds were so nice, well developed and so beloved for users. This is why there is so complaint about FSX too: simmers like them, want to keep using them, miss them on the new Sim.That's why (besides FSX SP1, of course) people is just to conected with FS9 for a long time being.Excuse me for the long post. Just trying to present you with one of the other faces os this FS universe.It's a sad feeling that takes us all, when developing a new project that requires time, dedication and effort, knowing that in just two years it will become useless.Does the fact that FS9 success for staying alone for 3 years, having the chance to be enhanced so much, making simmers so happy, may be a chance to think about not releasing FS11 in just 1 1/2 years from now? Or am I being inocent on this question?My best regards,Carlos Pereira.www.tropicalsim.com

[Mace 1,896]

>Hi Rhett,>>I don't mean classic flattens here. What I mean is deform the>height mesh using a flatning (averaging) algorithm. You only>do this for the mesh that is 'under' the vector data.Arggh! You mean POLYGON DATA...not vector data! :)Streams, roads, railroads, are vector data...the big rivers are water polys.> With the>right averaging algorithm you get beautiful sloping rivers and>you loose the spikes caused by bridges and dams.>That's exactly what I said.The mesh would be "deformed" as you put it, by the water polygon that is placed over it.Taxiways and Runways already do a similar thing in FSX. They have a built-in flatten property. I think the same could be done for water polys, except that it would flatten to a varying elevation as determined by the surrounding terrain elevation data.I think that is potentially an elegant solution.RhettAMD 3700+ (@2310 mhz), eVGA 7800GT 256 (Guru3D 93.71), ASUS A8N-E, PC Power 510 SLI, 2 GB Corsair XMS 2.5-3-3-8 (1T), WD 250 gig 7200 rpm SATA2, CoolerMaster Praetorian case

[Guest Ozzie]

Even IF English is not your native language - BRILLIANTY putI would love to see an "official reply" You should also consider incorporating some of this into your advertising

[PhilTaylor 45]

Some things, like new features and changes, are just a part of the business we are in.However, our goal is to be helping the 3rd party community get ahead of the power curve with the beta program. Were you in the beta program? Didn't that provide the heads-up as to the toolset and engine changes? Or was the SDK not in a good enough state to provide that head-start? Is there something we can do better there? That is one of the purposes of the beta, to enable the 3rd party community to get a head-start. So feedback here is valuable to us.I can categorically state FS11 will not be out in 1.5 years from now.So you will have the opportunity to leverage your work similarly to FS9.

[Guest JeanLuc_]

Hi Phil,thank you for taking the time to answer in these forums in such an open minded way.About backward compatibility, it was expressed a while back (the poll has been running here at avsim some time ago - should be easy to find it back with search) that performance was one of the main concern, more than backward compatibility. Up to the point of having many expressing that if dropping backward compatibility for performance was inevitable, then it would be the better choice (forgive me if I misinterpret the results of this "survey").As a matter of fact, there are no payware add-on I know of that runs directly from FS9 to FSX. They almost all (if not all) require rewritting of some sort (when dealing with gauges, aircraft models and flight dynamics).In the case of sceneries, it is not an issue of backward compatibility, it is an issue of not having the same level of functionality in FSX than we had in FS9, inspite having numerous scenery vendors in the beta test team.In any case, thank you for the support after releasing FSX in trying to make it better.Jean-Luc

[boshar 6]

It would be even beter (cleaner) if they corrected the included mesh so you don't need to do it on the fly using waterpolys! :7

[Guest vas_yan]

Hi Phil and thanks for the communication.I'd just like to point to a very common tourist area which has received a terrible portion of coastline data. The Greek islands (at least those that appear)show up square.That definitely hurt local national pride and did nothing to promote the Aegean Sea as a hot flying area :)I guess no need to fix the place since there's already a freeware replacement. Just have a look if you find the time to see how wrong things can be in some places.Kind regards,Vassilis

[Guest captaindobbin]

Dear Phil,Someone posted a suggestion for running the traffic on multiple threads; other people no doubt have had the same thought about auotgen. Is such a thing even possible, to separate one aspect of the engine from the core to take advantge of newer cpu architecture? Also, I for one think that fs9 compatibility should have been dropped. As you say, much time has to be devoted and the results are a mixed bag anyway. Granted a lot of people, like myself, have paid heaps of money for the former's add ons, but even with the need to re-soup up the new version, the cost of the sim is nowhere near actual flight training. Well as you can see we are all enthused, albeit apprehensively. Keep of the great work and most importantly, take all the time you need.

[Mace 1,896]

>That set is determined by a mix of the size of the fix + the>"value" of the location ( national pride, common tourist area,>hot flying area, etc ) plus how bad the RTM bits really are.heh :) Well then, I doubt the locations *I* bugged made it into SP1. :)It was a remote place or two in Nicaragua, near MNSC and MNBL. I bug reported both of them in January or December. It's ok, though, I fixed them myself, and post-SP1, if they are still not fixed I will upload my fixes. :)RhettAMD 3700+ (@2310 mhz), eVGA 7800GT 256 (Guru3D 93.71), ASUS A8N-E, PC Power 510 SLI, 2 GB Corsair XMS 2.5-3-3-8 (1T), WD 250 gig 7200 rpm SATA2, CoolerMaster Praetorian case

[Guest DC-9]

Hi Phil,As I am sure it's on the lowest of low priority levels, do you think it would be possible to further enhance the airport lighting capabilities in the patch such as adding a feature to scale back the taxiway lights like the runway lighting (a lot of people I believe would like this as well), as well as changing the remaining 2,000' of runway lighting from that yellow to a more realistic amber/redish orange color? I don't think it would take much to do, but know it would make a lot of people happier as well as adding more realism.

[DougMatthews 0]

>> With the right averaging algorithm you get beautiful >> sloping rivers and you loose the spikes caused by >> bridges and dams.>>That's exactly what I said.>>The mesh would be "deformed" as you put it, by the water>polygon that is placed over it.>>Taxiways and Runways already do a similar thing in FSX. They>have a built-in flatten property. I think the same could be>done for water polys, except that it would flatten to a>varying elevation as determined by the surrounding terrain>elevation data.>>I think that is potentially an elegant solution.>That is essentially what we are doing with sloped water. Unfortunately, interactions between our water sloping tool and some problematic DEM resulted in some water polygons that were not properly sloped, which resulted in problems like those found in the Thames river. A number of the more egregious water problems (i.e. Thames, Danube) will be corrected by SP1.-Doug

[DougMatthews 0]

>It would be even beter (cleaner) if they corrected the>included mesh so you don't need to do it on the fly using>waterpolys! :7 Improving the mesh is definitely desirable, but would not eliminate the need to "flatten" in-game. Perfecting the mesh is simply not feasible on a global scale. -Doug

[Mace 1,896]

Then this will always be a problem, unless we get super-accurate DEM's, which I don't see happening anytime soon. In specific locations, the DEM will have to be hand-tweaked if someone wants great accuracy.At present, I do not have an "easy" way to do sloped water. That will change with v1.06 of FS_KML.I re-did an entire river in Nicaragua using 1m "stair-steps" for elevation changes. It is not noticeable in flight. The only way it's noticeable is if you are on the water in an FS-boat. :) A true sloping polygon will be the best solution.Stair-stepping is superior to the legacy water, which follows the terrain mesh. By definition that kind of terrain-following water will be susceptible to pits and bumps in the mesh. Sloping, though, is superior to stair-stepping water polys for rivers.RhettAMD 3700+ (@2310 mhz), eVGA 7800GT 256 (Guru3D 93.71), ASUS A8N-E, PC Power 510 SLI, 2 GB Corsair XMS 2.5-3-3-8 (1T), WD 250 gig 7200 rpm SATA2, CoolerMaster Praetorian case

[russdirks 0]

Hi Phil,I'm not entirely clear on point #1."Any 3rd party who builds code intended for FSX with the VS 2005 SP1 bits installed will take a dependency on the new DLLs in the redist unless they take special care and compile with a special #define. This code will fail to load in FSX RTM." Is this true even if the end user has VS 2005 SP1 redist installed on his machine? Are you referring to DLL's only, or .EXE's as well? Thanks for the info,Russ

[Guest SkyDrift]

>...It is a hard question and I am sure we will discuss it in great detail with the community as Trains2 moves forward and FS11 starts to take shape....Phil>Phil - Thanks for your responses. They are appreciated. My perspective on this is that you (MS) will never be able to eliminate this problem by yourself. The nature of your business is such that you could not hire enough testers and developers to review & fix every problem within every square Km on earth - and make a profit on the product. so you are taking the only practical approach - fix the most egregious errors when found.And (as I'm sure you have already figured out) Trains2 is only going to compound the problem. If we FSX simmers - who spend 90% of our time flying at least a few 100ft above the ground - are unhappy with the terrain/scenery, imagine the unhappiness of trainsimers in their engines who can never get above 10-20ft or so (unless in an external, high camera view)!So what to do in future? I see a number of paths you could take:Option 1. Do nothing, take the complaints with good grace, and perhaps increase payment to your helpdesk vendor to allow them to hire one or two additional people to handle the increased call volumes.Option 2. Purchase/license more accurate digital data, should it exist, to reduce the errors. OK in theory, & perhaps doable for the major population centers where data will get more accurate in time, but there will never be that perfect 30cm dataset for that remote river in the remote highlands of a user's homeland that is their favorite flying location. While this approach might work in part for terrain on a semi-global scale, it's not practical for say, other scenery like airport buildings (or railway stations & trackside accessories for that matter), world-wide.Option 3. Get more involved in letting the hobbyist development community fix the problems. This is effectively what we do today when we upload our files to this and other forums. Some is totally new work, but most are enhancements (i.e. fixes) to your RTM product. The problem is, even though the SDK tools improve each version, they are still a long ways from being universally usable. Only die-hard (or stubborn) hobbyists or professional developers are willing to figure out all the SDKs quirks & foibles, the missing bits in documentation, and then write additional tools to make up for the shortcomings of the SDK.So today, there are not enough of us to do all this work either, but significantly improve the tools and that could dramatically change. Instead of complaining in forums like this of all the problems they encounter, users could fix many of them themselves.So, I'd respectfully suggest that your team puts more emphasis (i.e. dev-weeks) in the Trains2+FS11 dev cycle on enhancing & developing robust and comprehensive modelling tools and documentation very, very early on. You know who in our communities are the hard core developers. Ask us - professional & amateur alike - what we need that will allow us to model what we need (although you should have a pretty good idea since you & your team are reading these forums).As Churchill said: "Give us the tools, and we'll finish the job!"Not only will it help us, but the tools should make your own subsequent development work far, far easier - and quicker.Since the FS11 & Trains2 architecture is built on FSX, you have a unique opportunity here to solve three problems:a) dissatisfaction with the current SDK tools:( dissatisfaction with the inability of terrain/scenery bugs to be fixed in a timely manner (performance bugs is a different story)c) reduction in terrain/scenery issues in FS11 & Trainsim at RTM.Release these improved/new tools very, very early in the dev cycle - not 2 months before RTM, much earlier than that, please - to allow bugs in the tools to be found AND FIXED :>), to get feedback on features that are missing and that should be added to the RTM toolset, and to allow new hobbyists to learn how to use them *before* the RTM of Trainsim2 & FS11.The FSX community can enhance our world, and you get us to dogfood the tools for you. Plus, FS11 & Trainsim2 users will have a ready-made set of "enhancements" that they can download at RTM (because FSX terrain/scenery will be, as I understand it, compatible)rather than years later. Heck you might even increase FSX sales with Trainsimmers who by it so they can begin designing stations & trackside objects that they know will work in the new Trains2 product.I hope this provided food for thought.This has been a long post. Thanks for taking the time to read it.- Skydrift>And I see the community divided over it, so there is really no correct answer. Or am I wrong about the community's feeling on back-compat being split?