What Adrian Woods said speaks to the issue and about working on the sim to follow FSX.http://blogs.technet.com/b/torgo3000/archive/2009/04/13/a-change-of-direction-sort-of.aspx

The Adrian Woods post is a good, but it’s also an example of why most companies insist a PR department conduct all communication.His comment is so inside baseball it's hard to understand it.Worse - misinterpreted his comment might seem to justify some things he is absolutely not saying.He’s talking about the potential freedom to modify shaders without having to re-export the art.That is a pipeline -tools issue not an engine issue. Granted an engine might be modified to take advantage of this...but that would not increase performance.Being a tools issue it would only increase a designer's “iteration-time” – the time req’d to make and test a change.That would in itself lead to a more polished and optimized game.This is the kind of change that could make a big difference. But it’s not about rewriting the engine.Any unnecessary work might threaten implementing this kind of improvement to the shader pipeline.It’s best to let them make this call... :( We’re being a bit like that patient who reads stuff of the internets and then tells a doctor how to treat them. :( :(

Mathew, The clock speed of PCs have not gone much higher in that last 6 years. (where as before it would double every two years.) In reply to horrible performance of FSX, ACES, in early 2007, said that, had they known that Intel would go to multi-core direction (instead of higher clock speeds)they would have coded FSX differently. Considering that FSX can not utilize more cores and considering that the clock speeds are not going higher, how would you suggest MS brings a sellable game in 2011 using the same code that was already obsolete in 2006? In fact FSX still brings the i7s to their knees!! That is not acceptable nor logical after 5 years of new PC advancements. I seriously doubt that they would not upgrade a good chunk of their code to take advantage of multi Core/Multi GPU PCs. If they do not rewrite the code to allow efficient use of Multi cores, Flight will not have any head room for added features and will not work on most non enthusiasts' off the shelf PCs. (Read their repeated statements of how they want to attract a broader range of customers many of whom do not have custom made, over clocked PCs). Lastly please note that this cycle of release would be 5 Years after introduction FSX (instead of typical 2). As soon as ACES was disbanded, many of the programmers got rehired to work on Flight. So development was not put on hold for long at all. Perhaps this long development cycle is indicative of a rewrite of the code rather than recycling that of FSX's old, tired and obsolete code. I admit I am speculating, but I am doing so based on what makes sense.

I couldn't of said it any better myself. :)

That ship has already sailed - SP2.The question is how far Flight takes multi-core.I’d speculate it’ll be really well optimized for today’s machine- say maybe a quad core. At the end of Flight’s lifecycle how many cores will a hardcore users have? …16 -32? Who knows maybe 128 ;)Obviously how they parallelize Flight today will affect its potential advantage on a 128 core :)Mulit-core - sure! Massively mulit-core who knows?

The Adrian Woods post is a good, but it’s also an example of why most companies insist a PR department conduct all communication.His comment is so inside baseball it's hard to understand it.Worse - misinterpreted his comment might seem to justify some things he is absolutely not saying.He’s talking about the potential freedom to modify shaders without having to re-export the art.That is a pipeline -tools issue not an engine issue. Granted an engine might be modified to take advantage of this...but that would not increase performance.Being a tools issue it would only increase a designer's “iteration-time” – the time req’d to make and test a change.That would in itself lead to a more polished and optimized game.This is the kind of change that could make a big difference. But it’s not about rewriting the engine.Any unnecessary work might threaten implementing this kind of improvement to the shader pipeline.It’s best to let them make this call... :( We’re being a bit like that patient who reads stuff of the internets and then tells a doctor how to treat them. :( :(

Are you talking about the snake oil doctors that peddle thier big pharma pills ?? I would not let them treat me....

It seems that we are basically splitting hairs and "arguing" about semantics.Bottom line, flight will be a major rewrite.So let's just wait for filght to be released as people have chosen their corners and evidenly will not come out until it is releaed.

Bottom line, flight will be a major rewrite.

No I disagree. MS are just 'tweaking' the existing FSX code-base to avoid a major rewrite. None of the screens released so far indicate a major rewrite. All I've seen so far are 'tweaks'. IMO the only completely new stuff will be the Live/Marketplace integration.

Matthew, I guess you didn't read my reply to you earlier post above in this same page. I'll rehash it in a shorter paragraph. The fact that FSX was written to take advantage of pure clock speed and not for utilizing multi core pcs, made it a obsolete code. ACES admitted that they were wrong to think PC clock speeds would continue to increase rapidly and didn't anticipate a switch to multi core processors. They admitted that it was a BIG mistake on their part coding FSX that way. The fact that FSX can not still be fully run on todays off the shelf system 5 years after its introduction (and two service packs) is a testament to that mistake!!!They are not going to use the same code that they called a big mistake 4 years ago. If they did, Flight will not run any better 4 years from now than FSX did 2 years ago. They have to recode Flight even if it doesn't offer visually different results as you are seeking. Otherwise its doomed to be a failure.

Matthew, I guess you didn't read my reply to you earlier post above in this same page. I'll rehash it in a shorter paragraph. The fact that FSX was written to take advantage of pure clock speed and not for utilizing multi core pcs, made it a obsolete code. ACES admitted that they were wrong to think PC clock speeds would continue to increase rapidly and didn't anticipate a switch to multi core processors. They admitted that it was a BIG mistake on their part coding FSX that way. The fact that FSX can not still be fully run on todays off the shelf system 5 years after its introduction (and two service packs) is a testament to that mistake!!!They are not going to use the same code that they called a big mistake 4 years ago. If they did, Flight will not run any better 4 years from now than FSX did 2 years ago. They have to recode Flight even if it doesn't offer visually different results as you are seeking. Otherwise its doomed to be a failure.

They did add some multi-threading support in the service packs. To get the absolute best performance on todays multi-core hardware they should be rewriting the entire sim to be multi-threaded from the outset and that includes graphics/scenery/weather/ai/atc etc.It's highly unlikely IMO that they are doing that. We have no evidence in the screenshots that Flight is anything more than 'tweaking' the existing FSX code (except for the obviously new Live/Marketplace code).You're also assuming that ACES members were rehired to work on Flight, but there is no evidence. The 18 month old blog post from a former ACES employee said that he was rehired to work on a new project, who says that is Flight? Yes he admits FSX could have been much better but does that mean they are rewriting Flight from scratch, no.Sandy-bridge can be easily overclocked to 4.6ghz on air. I think 4 years from now we will be seeing entry level PCs at 5ghz with 8 cores.Yes MS will tweak FSX in all areas to make it better performing for Flight. But no MS will not start again from scratch, it makes no sense. FSX after all is not broken just in need of a 'revamp'. You can think of Flight as FS11.edit: Of course now is the perfect time for the Flight devs to chime in and give us some definitive answers. The news section on the Flight website is pretty vacuous and has no doubt passed through the PR filter.

The news section on the Flight website is pretty vacuous and has no doubt passed through the PR filter.

What a fabulous filter it is!! Remove all meaningful information and leave nothing but vague and ambiguous statements. Definitely worded by a lawyer or a salesperson unsure of their product, which considering how far the program is from being finished, sounds about right.Regards, Mike Mann

You guys are arguing about the same thing now. One person calls it tweaks, another calls it rewriting part of the code. It is possible to tweak code to the point where every line is dfferent from the original and it is effectively rewritten.These days it takes 4-5 years with a respectable sized and competent team to write an engine from scratch, and there is no gaurantee that the new one will be any better than the old one. Given that the Acceleration Pack was released end of 2007 and there was work going on on other projects until 2009 it doesn't seem likely there would be time for a complete rewrite. That said, screenshots only give you a glimpse into a subset of the graphics features of an engine, and absolutely nothing about any other part of an engine. It is impossible to derive the feature set of an engine from 10 screenshots. Plus, a lot of what makes a game look like a particular game is the content, and content is generally one of the last parts of a game to be completed.

You're also assuming that ACES members were rehired to work on Flight, but there is no evidence. The 18 month old blog post from a former ACES employee said that he was rehired to work on a new project, who says that is Flight? Yes he admits FSX could have been much better but does that mean they are rewriting Flight from scratch, no.

Just the FAQs man, just the FAQs!

"Microsoft Flight" is being developed by many of the same creative minds that shaped "Microsoft Flight Simulator," and we're excited about delivering a new take on this classic franchise.

As it happens, they used precisely the same phrasing in their answers to two separate questions.

You guys are arguing about the same thing now. One person calls it tweaks, another calls it rewriting part of the code. It is possible to tweak code to the point where every line is dfferent from the original and it is effectively rewritten.

Sure, but rewriting the sim by constantly tweaking it would be more expensive / time consuming than rebuilding the sim from 'scratch'. If they are not rebuilding from scratch then I somehow doubt they will be tweaking the existing code until its effectively a complete rebuild.The evidence so far (eg the 'stars brighter than sun' bug) suggests that they are using the existing FSX code base. Those that are expecting Flight to be a completely new sim with amazing new features and performance are going to be disappointed.IMO Flight is just an update to FSX. We can think of it as 'FS11', FSX with some tweaks and Live/Marketplace integration.

"Microsoft Flight" is being developed by many of the same creative minds that shaped "Microsoft Flight Simulator," and we're excited about delivering a new take on this classic franchise.As it happens, they used precisely the same phrasing in their answers to two separate questions.

Could mean anybody. Were the FSX/TS2 programmers rehired? IMO they are the most important part of the dev team. I don't think people creating the art assets, writing the manual/help or creating the tutorials/adventures have much relevance when it comes to fixing the issues inherent in the core engine. If the original programmers are back, with all their experience in FSX/TS2, then there is a good chance that we might see the performance problems in FSX actually fixed in Flight.

What I don't understand is why does it have to be a new engine ?If MS fixes all the bugs in FSX and want to call it Flight what is the problem, personnaly I don't care if FSX engine is reworked and fixed if it's "gonna" run like a brand new one. (so far I only see FSX)They (who ever was working on FSX) did not have the time to finish the DX10 mode before the released of FSX so that is why they called it DX10 preview, if I use FSX with DX10 enable my frame rate goes up a lot and the flight is smoother, so let say they are working on FSX to make DX10 work like it was suppose to work and at the same time fix all the bugs....popping trees, flickering runways, poor frame rate, better default scenery like we see in their released pictures +++ ....we would have a very good simulator right.....so why should we look for a new engine?

What I don't understand is why does it have to be a new engine ?If MS fixes all the bugs in FSX and want to call it Flight what is the problem, personnaly I don't care if FSX engine is reworked and fixed if it's "gonna" run like a brand new one. (so far I only see FSX)They (who ever was working on FSX) did not have the time to finish the DX10 mode before the released of FSX so that is why they called it DX10 preview, if I use FSX with DX10 enable my frame rate goes up a lot and the flight is smoother, so let say they are working on FSX to make DX10 work like it was suppose to work and at the same time fix all the bugs....popping trees, flickering runways, poor frame rate, better default scenery like we see in their released pictures +++ ....we would have a very good simulator right.....so why should we look for a new engine?

Like you say, I'd actually prefer they 'tweak' the FSX code base and fix the problems plus add some new features (eg, cloud shadows). This approach might also mean its easier for existing FSX devs (eg PMDG/LDS/REX/Orbx etc) to patch existing products for use with Flight. Some people seem to think though that Flight is a complete rewrite from scratch. I'm arguing that its not.