No speculation required! Microsoft themselves said so...

+1

Properly optimized and pared down to the essentials for the current task, they could easily have shaved 5GBs off the total scenery and art assets.

The essentials of course would be Hawaii and the GA aircraft that Microsoft has shown ad nauseam.

Gerry, I was not referring to the code, but precisely the opposite, which would be all assets other than code, most especially those that were retained only for backwards compatibility.Properly optimized and pared down to the essentials for the current task, they could easily have shaved 5GBs off the total scenery and art assets.Folks should also bear in mind that even though the most of the scenery and art assets have been recycled from FSX, that has absolutely nothing to do with the basic engine, which may well have been largely rewritten and/or optimized.

Are you saying that there is duplicate data in the files for FSX, FS9, FS8....?

Are you saying that there is duplicate data in the files for FSX, FS9, FS8....?

To some extent there is, night maps have four channels (RGBA) the minimum needed is one channel.The same with spec maps (RGBA) but the most optimized option is also a single channel.Even having so many unique texture files is bloat...they cloud be atlased into fewer large files.These changes could add efficiency with only a little impact on quality.IMO all the focus on the "engine" is misplaced because content has more impact on performance.The problem is that it can be a difficult path for optimizations because it's a tremendous amount of manual work.

I repost the extract from the Flight site with the last sentence in boldGiven that ACES were responsible for developing Flight Simulator in the past, the last clause really means:, and ACES can’t claim to have done the best job of welcoming a wider audience in the past.

Without the full context, paraphrasing it this way seems to change the meaning.Clearly 'we' means Microsoft specifically.I'd say it wasn't meant to ever imply this was a rational for closing the studio...

IMO all the focus on the "engine" is misplaced because content has more impact on performance.

Wold you need a new engine to create a more realistic ground, like the Outerra engine? The flat blurry ground of the ESP engine isn't very good looking, unless you look straight down on it from a sattelite's perspective.

Wold you need a new engine to create a more realistic ground, like the Outerra engine? The flat blurry ground of the ESP engine isn't very good looking, unless you look straight down on it from a sattelite's perspective.

Sometimes I put engine in quotes because it means a lot of different things :)...call the game a collection of components called engines or you can call the entire collection of engines an engine.The ground mesh is a component and it can be replaced…...though understanding the full effects of that change would take a whole lot of inside knowledge.So I would put a lot of trust in comments outside of team members I’m not so familiar with Outerra, but I believe what it does is invent high detail as you zoom in.Its terrain data is at a much lower resolution (closer to a regular simulator) but it has algorithms to interpret what the ground should look like as you get close in. The upside is there is high-detail everywhere you look. The downside might be that the invented detail doesn’t match what’s there in real life. Everything so far in Outerra is Idaho outback. But in suburbs of Huston the effect might look out of place…unless the system has a new set of rules for each region. Drawing a realistic face takes skill, but to draw a portrait of anyone’s face takes outstanding skill. Outerra is awesome and I always look forward to seeing it progress.

Without the full context, paraphrasing it this way seems to change the meaning.Clearly 'we' means Microsoft specifically.I'd say it wasn't meant to ever imply this was a rational for closing the studio...

The context is surely clear - I quoted exactly what Microsoft wrote.If Microsoft felt that FS11/Next was going to be a success it wouldn't have dropped it and closed ACES. If it had been a success ACES would have claimed the credit. If it wasn't ACES must take the responsibility.As I recall, there were rumours even at the time that it happened because ACES wouldn't respond to what Microsoft wanted.

IIRC alot of folks at ACES were infact pilots and a couple even owned their own planes... so you can assume they were enthusiast like many of us. Ones who would much rather see a SIM like FSX with all the openness to allow a healthy add on market to thrive. I can believe that MS and ACES would disagree on the direction each wanted the sim to go. But ultimately MS will win any argument/debate over direction as was surly the case with FLIGHT. If MS had pics of anything outside of Hawaii and they were as visually appealing as some of the Hawaiian shots we've seen... they would post them.If they had a top notch Hong Kong skyline or an amazing shot of the Alps they would post them...Not posting them serves no logical nor business purpose and they best of all, know this...Personally I am to the point now I feel they have none to show us for a very good reason, just as I believe there is a very good reason why we now have so many developers who are openly upset with MS and the direction FLIGHT is heading...They know more than we do, but if they are upset then something is wrong....

If Microsoft felt that FS11/Next was going to be a success it wouldn't have dropped it and closed ACES. If it had been a success ACES would have claimed the credit. If it wasn't ACES must take the responsibility.

Strictly speaking, structural changes also happen in organizations to find savings...So it's also likely MS was streamlining its organization.What we can see is MS funded FSNext and canceled ESP and TrainSim.What we don't really know is the extent of the personnel changes...rumors there seem to point in two directions.Having some experiences with rumours in organizations like this I don't put any trust in them.Even folks we think should know, and believe in their hearts they know, get it all so wrong.It's just like politics and all that...I don't think we'll find the real story, I'm sure it's about halfway between :)

If I were running the team for an overhauled sim, I would get from A to B as fast as possible, spending all my time designing the pipeline. First, render the 3d world. Get the mesh details down, decode it, compress it, decompress it in realtime, load tiles. Now texture it based on landclass. Add the vector data and water data. Populate the world with autogen trees and buildings. Now, add the movement mechanics. Add the 3d modeling and systems for the aircraft into play. Add the weather system. Then the AI ATC system with AI aircraft maneuvering and path following.All this can be done with just a small portion of the world (say, Hawaii) to limit development of art asset time during the pipeline development. Then make an example aircraft (one that utilizes the new tech developed for this iteration of the engine).Once the pipeline is complete or past a stage where people can be relocated to develop the assets, now add the rest of the world, region by region and upgrade any aircraft from the previous generation to the new generation standards, making completely new aircraft where needed due to imcompatibilities.I'd say MS is somewhere still developing and optimizing the pipeline of the engine. Once they're done with that, they can have few people performing last-bit tuning of the engine while the rest develop the rest of the assets. This is why we haven't seen all of the aircraft or an airport full of AI aircraft or any airborne aircraft yet. They haven't been integrated to the new engine yet. Maybe the big update will be an expected release (generalized of course - e.g. Summer 2012 as opposed to 12 Jul 2012 - so they will have leeway in case some wrenches are found thrown in by some monkeys at some point) if the pipeline is near complete and now their focus will shift on assets. Once you have a framework in place, developing assets is a fairly simple process. Look at 3rd party addons. If you create one addon (let's say a C++ gauge), you can take the code you developed and cut out the stuff that doesn't make sense from the PFD to make the ND, but the rest is still usable. Now you can focus on the code that makes the ND different from the PFD. But the initial design of the PFD gauge will take longer than the reforging into the ND as you have a working product you can change to make work. Same for aircraft (except for the mesh) or Autogen. A framework makes the whole process faster. Instead of spending time at the beginning crafting all assets needed for the final product, spend time on one asset to demonstrate proof-of-concept. Once at the finish line per-se, make the rest of the assets. If you don't have time to make all those intended (maybe you wanted 20 aircraft, but decide to forego 6 aircraft to get the product out the door), you can always make a free expansion update a few months later and add-in what you didn't ship and make your customers feel like they get more than what they originally paid for. After all, they didn't know you already intended on giving them those extra aircraft.As they develop the aircraft for the new engine, they will end up setting standards and updating SDKs. We have had an SDK dating back to FS2004 (granted it was released well after the main title was released) but it was not an after-thought, just development of the SDK can't proceed the final product as things may change and the SDK will be outdated at release.

How many planes have you seen? Does that mean those are the only planes that will be included?

I'm glad you have said this

Are you saying that there is duplicate data in the files for FSX, FS9, FS8....?

Have you ever spent any time at all examining how the various navigation data are scattered amongst many hundreds of separate files? A lot of the data values are duplicated, since certain key value(s) are used to index these files into a mostly functional form. This is the inevitable outcome of having continuously added "new stuff" to an old system, mostly to maintain the questionable goal of "backwards compatibility."The above is just one example of the type of optimization that is likely going on. The so-called scenry database(s) could also loose quite a bit of excess size from better arrangement as well as more efficient compression algorithims.

I can't say I have looked deeply into the structure of the files. What types of data are duplicated and where?

As long as it is just the demo scenery, like Sint Marteen for FSX, I'm OK with Hawaii.

Oh my, imagine when the demo gets released and it is Hawaii-only (similar to the FSX demo like you said). The pessimists will have a field day