RSX not fast enough yet

[woundingchaney]

They might not want to end it, but if a bunch of us worked together to show our clout as consumers perhaps they would listen to us.

I don't care what they want. There is a peaceful and lawful, but also very intense "war" between consumes and producers. We as gaming consumers need to unite together and be more demanding of the producers. We cannot let them manipulate us forever.
They arent manipulating us whatsoever, yeah they are misleading, yeah they over promising, yes they are vague but these companies are producing what many of us not only consider to be a hobby but a passion. If the specs of the console arent what you want them to be would that stop you from buying the console and enjoying the games, hopefully not. Even so just what do the specs mean to you personally?? Surely there has to be a few games that have peaked your interests regardless of what the final specs will be.

There is never going to be a large enough petition amongst gamers for what your suggesting, just lay back and enjoy the show (after all thats what all this is - entertainment).

[Nameless]

They arent manipulating us whatsoever, yeah they are misleading, yeah they over promising, yes they are vague but these companies are producing what many of us not only consider to be a hobby but a passion. If the specs of the console arent what you want them to be would that stop you from buying the console and enjoying the games, hopefully not. Even so just what do the specs mean to you personally?? Surely there has to be a few games that have peaked your interests regardless of what the final specs will be.

There is never going to be a large enough petition amongst gamers for what your suggesting, just lay back and enjoy the show (after all thats what all this is - entertainment).
Well said...

[makeitlookreal]

If the gaming community would put aside all their doubts and skepticism they could accomplish quite a lot. That's the problem with 95% of most efforts. Almost everyone is so scared of not accomplishing anything that they don't at least try in the first place. However, if everyone would put their doubts away and make the effort there is at least a chance of something positive happening.

I don't know if I will purchase the PS3 or not. I am impressed with MGS4. That is the only game so far that *really* makes me want to purchase the PS3. It is the only game that in my opinion comes close to the quality that I desire from a gaming system. If I could know for a fact that the PS3 had the specifications and features to make it both worth $599 and stay competitive with PC's for at least a few years then I would purchase it.

Quite frankly, more than any other issue, Sony's vague hints, strange comments, and fooling around with the specifications and not being honest with us is making more very angry. If I don't purchase a PS3 the #1 reason will be because I don't want to do business with a company that is not honest and straitforward. Sony doesn't HAVE to tell us anything, but even if it does not make any difference at all on PRINCIPLE I may have to NOT buy the PS3 just for their lack of openness.

I am getting sick and tired of hearing rumors everyday. It is fun to discuss them, but at the same time it makes me want to vomit because there is no need for them! Sony could be honest with us and all the rumors (or at least most of them) would end.

The specifications DO matter to me and ARE important. First of all, if their specifications don't AT LEAST match what they announced at E3 05 then I will not purchase their system, period. Secondly, for $599 we deserve a top notch gaming system with the best specs possible. Third, I want a system that clearly blows away all the competition and can produce at least almost photorealistic games that are BEYOND THE UNCANNY VALLEY! If the PS3 does not have the specifications I want then I may not purchase it.....

MGS4 is an amazing game, but it is the only thing so far that makes me want to purchase the PS3.

[makeitlookreal]

Also, I can't enjoy the show when Sony is manipulating us like dogs hanging steaks above us. The bad thing is we don't have to be manipulated by them. We can PEACEFULLY and LAWFULLY with our TEETH (dollars) let our owners know we can HUNT for food (another console more powerful than the PS3 or 360) ourselves if they don't come clean with us and be stratforward.

[xbdestroya]

MILR no one here is going to buy a PS3 based on whether Sony came out and gave us RSX numbers or not - they're going to buy (or not buy) the system based on games, other functionality, price, etc...

The internal specs are a means to an end, not the end itself.

I personally am definitely not being manipulated by Sony - this I know as fact. I'd like to think few people here are.

Let's try to keep the thread technical such that we might all learn from each other though. You may feel that simply knowing the final specs are why a lot of us are into these discussions; but it's not. It's the process of detective work, architectural discussion, weighing certain pros against certain cons, understanding why certain decisions were made... it's the whole journey from a to b that makes it 'fun.'

[makeitlookreal]

I am going to set an example. Sony can keep their PS3. Their not manipulating me anymore. They don't want to be honest with me then I don't want to do business with them.

I guess there is no more need for me to post on these boards.

However, if the game making group ever gets going I will help them out.

[Nameless]

MILR no one here is going to buy a PS3 based on whether Sony came out and gave us RSX numbers or not - they're going to buy (or not buy) the system based on games, other functionality, price, etc...

The internal specs are a means to an end, not the end itself.

I personally am definitely not being manipulated by Sony - this I know as fact. I'd like to think few people here are.

Let's try to keep the thread technical such that we might all learn from each other though. You may feel that simply knowing the final specs are why a lot of us are into these discussions; but it's not. It's the process of detective work, architectural discussion, weighing certain pros against certain cons, understanding why certain decisions were made... it's the whole journey from a to b that makes it 'fun.'
Well said...
I'm sure MILR will be posting something tomorrow... LOL
If the final stats are this important to MILR's sanity, he should take a break from the forums until September at that time we should have final PS3 specs. :birthday:

[version]

sony will delay ps3's launch to next spring :D

VAPORWARE:D

[masonite]

i have a feeling that if a news report comes out in the morning stating that a sniper has been picking off sony execs as they walk out of SCEA headquarters, i know exactly who the sniper would be.....

MILR, the wait is just part of the game....if you think you need to start a revolution to make sony tell you all of their innermost secrets, then do so, i'll even start calling you "MILR Guevara" if you so desire, but you have to realise that the lead up to a launch is like a game of chess between the companies....one person moves.....then the other waits, contemplates their strategy, decides on their strategy....creates backup plans......Sony have carefully decided and analysed every move that they have made so far to ensure that every person is as compelled as possible to buy their product - as with both nintendo and microsoft. you should thank sony (and microsoft and nintendo) for making the lead up to launch as exciting as possible for you, the consumer.

remember, absence makes the heart grow fonder.... ;)

[Red_Eyes]

Dude man, information will come out when the time is right. The reason the wait seems so long is because you're sitting there and refreshing every 5 seconds. Stop. Get up. Go outside. And get a life for at least 3 months. Then TGS will come and information will come out.

[Crossbar]

Dude man, information will come out when the time is right. The reason the wait seems so long is because you're sitting there and refreshing every 5 seconds. Stop. Get up. Go outside. And get a life for at least 3 months. Then TGS will come and information will come out.
Couldn't agree more, I went offline during a week of vacation. And guess what, the world didn't stop neither did it go into high speed.

I must admit I was a bit curious when going online again, but that was about it. Pull out your internet connection for a couple of days and see if you can handle it. If you don't, you should try get some advisory about how to get rid off you internet addiction.

I also agree with overclocked.

[GUNDAMSEED]

Man even if I don't know about ps3 specs , i will bet getting on just because of FF13 and DMC4 . I spend over 300hours on DMC3. Plus i have a HDTV and i am buying a new one, i need some HD content for it .

PS3 for 600US here i come.

[xbdestroya]

Ok I pruned this thread some, because just as MILR's own posts can clutter, too many responses to him do the same.

RSX. Let's get back to it. :smoke:

[Heinrich4]

I hate to put this up, but the most recent indicator came from the Inq:



Anyway though not many notable folk have come out in support of this document (in fact none), at the same time none have come out against it either. In these circles, sometimes that can be the same as acknowledgement.

Who knows though...

@MILR: Southbridge functionality on RSX won't do anything to make games look better.

PS - Obviously Charlie's interpretations are lacking if he is emphasizing that this is what makes RSX different - but the doc if real is interesting should it indeed be implying integrated S/B functionality.


If RSX is South bridge(like "nforce way of life"...), maybe have some possibility to include or I/O chip(like cpu psone in ps2) a "combo" EE+GS of ps2?

[CARTIER90]

will the PS3 handle as 'much' physics as a PC with an ageia card (retailing for £215 in uk)....?

[version]

agea a lot of shit

[venomv]

will the PS3 handle as 'much' physics as a PC with an ageia card (retailing for £215 in uk)....?

According to Ageia the only systems capible of utilizing everything in there latest release is a PC with a Physics card and the PS3. (or something along those lines anyway)

[woundingchaney]

agea a lot of shit
I as well am severly unimpressed by Ageia attempts at not only software physics but the support and performance of their physics card.

[Z]

well, PS3 devkits come bundled with a complete version of both Agiea and Havok. so, no game should really have an excuse for having bad physics. only bad or limited funding.

[overclocked]

I as well am severly unimpressed by Ageia attempts at not only software physics but the support and performance of their physics card.

Yeah the question should imo how good is the "middleware" software from Ageia at first for PS3? However i would expect it will not take that long to find out.
In those few PC titles that has "support" i think its rather a question of the game-software instead of the hardware imo.

[cpiasminc]

well, PS3 devkits come bundled with a complete version of both Agiea and Havok. so, no game should really have an excuse for having bad physics. only bad or limited funding.
Number one... what makes you think Ageia and Havok are "good" physics?
Number two... what makes you think having physics middleware means everybody will use it (even if it really were free with PS3 dev registration)?
Number three... If you seriously believe that having and using a commercial physics engine automatically solves everything, then you will READ THIS (http://psinext.e-mpire.com/index.php?categoryid=4&m_articles_articleid=271) NOW. If you like, I can go into more technical details which I deliberately avoided there.

[PhYmon]

I agree with Cpi the only give to developers the middleware to see what the devs kits are capable of, showing some things that can get done with it and how can they make it, but most of the developers dont even try them in order to make their own, thats why they get frustrated but in the end they made a wonderful piece of work.

[Z]

someone said Agiea is better for environmental , large scale?, physics whyle Havok is better and character and object interactions. I don't know what are the differences between the two apps.

Cpi, all I'm saying is that having two complete versions of physics engines by default is much better than having nothng at all. how many games use Havok? well, now eveybody can use it if they don't have their own engines or want to purchase something else. I simply see it better to have something than nothng at all.

be gentle O_o'

[cpiasminc]

Especially since, Havok for instance gets anywhere between $50k and $100k, and that was the rate about 3 years back. And the thing is that even if they give it away for free with a PS3 license, what happens if you're a multi-platform dev? You still have to license a 360 or PC version and pay for multi-platform support service...

someone said Agiea is better for environmental , large scale?, physics whyle Havok is better and character and object interactions. I don't know what are the differences between the two apps.
NovodeX solves a lot of interactions by allowing penetration and using pmaps to solve for the maximum depth of penetration. Making it a lot more accurate, but also very sensitive to floating point issues. ODE also does this, but uses algebraic solutions to solve for depth of penetration. Havok does NOT allow penetration and instead uses the difference between the length of the step and the point of first collision to estimate energy of collision for the complementarity problem (lots and lots of GJK steps). Havok is more stable by nature not just because of this, but also because of a lot of protection against corner cases and so on. This makes Havok VERY stable even at 15 Hz (whereas Ageia isn't stable at all until you get up to 50 Hz), but also very erroneous and very slow.

They both also use collision libraries ultimately derived from very old implementations which, btw, are complete and utter crap. And I mean this in the sense that the algebra itself is just plain wrong.

Cpi, all I'm saying is that having two complete versions of physics engines by default is much better than having nothng at all.
What makes you think they just *have* it? Sony simply acquired the rights to sublicense PS3 versions of middleware ranging from Havok/NovodeX to UE3 and in time, probably Gamebryo as well... Devs still have to buy those licenses if they want to use it.

be gentle O_o'
Sorry, but I've heard all too many of these "don't worry be happy" posts about middleware provisions and it rubs me a little raw. I feel the need to wring someone's neck.

[Arnaud_M]

I as well am severly unimpressed by Ageia attempts at not only software physics but the support and performance of their physics card.

Wounding, have you seen the CellFactor demo, using Ageia middleware ? It is very, very impressive (and while the name is CellFactor, it has nothing to do with the Cell processor).

Arnaud

[cpiasminc]

The main thing about speed of physics simulations is that a high number of active elements is not a huge speed drain -- it's when the number of active elements in one single compound collision is high that performance really dies.

If I take 500 boxes but no collisions involve more than 2 boxes at a time, it's still pretty fast and the LCM is still sparse. If I've got 120 all touching each other (and none of them are at equilibrium, meaning we can't cull them), then you've got troubles.

Wounding, have you seen the CellFactor demo, using Ageia middleware ? It is very, very impressive
That one is mainly a hardware demo. Nice how they like to pretend they support cylinder collision (they're extruded octagons). The whole demo is simply about high element counts more so than any good reactions.

[rog27]

My opinion on middleware engines is that they are generally crap. If devs are serious about making epic games with some serious production value (like blockbuster movies), they are probably willing to invest the time, money, and effort to make a proprietary engine which emphasizes and enables the visceral effects necessary for that particular scenario and which fits with the art style.

[Crossbar]

My opinion on middleware engines is that they are generally crap. If devs are serious about making epic games with some serious production value (like blockbuster movies), they are probably willing to invest the time, money, and effort to make a proprietary engine which emphasizes and enables the visceral effects necessary for that particular scenario and which fits with the art style.
Remember it's always about return on investment. The larger sums you invest in your own development the larger risk (budget overdraw, missed deadlines etc.) you are taking compared to investing into some middleware which may guarantee a certain level of quality, given you have competent people using it.

I have no practical experience from game engines, but I am certainly in favor of middleware as concept in any business. If game engines generally have bad quality, then it's bad, but that may be because they are still pretty imature. Middleware done right certainly helps quality and develpment time. I certainly don't want skip all the commercial libraries I am using.

[Angeljuice]

Stop cutting threads up, this one had a good sideline. If you insist on posting MILR's stupid questions that do nothing but repeat themselves in thread after thread, you should keep all the responses, OR STOP POSTING HIS CRAP IN THE FIRST PLACE.

5..4..3..2..1..removed.

[Angeljuice]

Blah, I'm not playing anymore....OVER AND OUT...

[Twilight Prince]

^double post

[BillCosby]

That one is mainly a hardware demo. Nice how they like to pretend they support cylinder collision (they're extruded octagons). The whole demo is simply about high element counts more so than any good reactions.

Wow i didnt know that. I was impressed by the demo until u said that. They did a good job on the texture mapping of the barrels in that case, cause I never would have thought they weren't cylinder's. Oh and a question. Why would cylinders be more difficult to program collision detection for? Oh and now a comment. With the support in the PS3 for more advanced physics, and the overall power of RSX/Cell combo. It would be neat to see an updated form of GEOMOD. U know the thing in Red Faction that made it possible to blow a hole through a whole level. U could go through the whole game and never use a door.

[cpiasminc]

They did a good job on the texture mapping of the barrels in that case, cause I never would have thought they weren't cylinder's.
Well render geometry and collision geometry are two different things (other than terrain, the two almost never match). The render geometry is probably more detailed than simply octagons... just the collision geometry is octagons. Currently, there is no physics engine anywhere that supports real cylinders.

Well, to be correct about that, we were working on just such a thing at my last job (and I'm sure many others have attempted it), but it was never completely working, and everybody who was working on it has since left that company.

One of the theories I was wondering about was the use of superquadrics, but since this leads to high-order fractional polynomials, it becomes a pain in the neck to solve (though part of the idea was to try and determine the limit as n tended to zero and use that to our advantage in the actual algebra) -- but even 4th order polynomials are virtually impossible inside of single precision floating point. I've seen cases where values close to epsilon^2 are not small enough to be considered zero. Errors as small as 10^-14 meters resulted in solution errors as large 5 meters.

And it will be even worse on the SPEs because they're SIMD, and the single precision floats don't get promoted to doubles internally like they would on a scalar IEEE FPU.

Why would cylinders be more difficult to program collision detection for?
The main problems are the cylinder rims and the caps. While some engines have had limited success doing tubes, a cylinder... Well, you've got a problem with your axis-separation tests (http://gpwiki.org/index.php/Polygon_Collision). Now the side of the cylinder is relatively simple because you can do it like a sphere sweep, and if that was all she wrote, life would be fine. Unfortunately, axis separation is not so obvious with a cylinder because there are 2*(infinity + 1) axes of separation and while this infinite aspect exists with a lot of other shapes, the axes to check are more trivial. While you can try to do it as a capsule + 2 boxes (and many games do this), the result is problematic because the cylinders can't be stacked arbitrarily, and you also get weird "climbing" in a rim or cap collision. An ellipsoid sweep is also doable, but the resulting cylinder is not very stable standing.

Now at least with an octagon, you only have 10 primary axes of separation (8 sides + 2 cap faces). Meaning in a worst case collision between two octagons, you can't have more than 100 tests -- and in any case, you cull away the non-facing ones, so it ends up being more like 9 tests.

The worst part of it is if you're running the type of simulation that needs to determine penetration depth. As it just so happens, there's no good solution for a cylinder. An approximation of a cylinder can be estimated by face, even though it's inherently not perfect. And there's the further problem of sweeping a cylinder... what the hell is the sweep of a cylinder? In an ideal world, all cylinders would move on their axes, so the sweep of a cylinder would be nothing more than a longer cylinder. With an octagon, you've got a convex hull shape, and the sweep of a convex hull is still a convex hull.

[BillCosby]

Oh okay I see. Its not that the barrels themselves arent cyllinders its in the engine u have to designate a boundary around the object as the area where collision occurs, and they didnt use cylinders they used octagons. Gotcha.

[rog27]

Remember it's always about return on investment. The larger sums you invest in your own development the larger risk (budget overdraw, missed deadlines etc.) you are taking compared to investing into some middleware which may guarantee a certain level of quality, given you have competent people using it.

I have no practical experience from game engines, but I am certainly in favor of middleware as concept in any business. If game engines generally have bad quality, then it's bad, but that may be because they are still pretty imature. Middleware done right certainly helps quality and develpment time. I certainly don't want skip all the commercial libraries I am using.

I wasn't advocating dropping all use of libraries or even complete middleware engines themselves. I was just saying that most big budget games benefit from having their own proprietary engine. In a way, Gears of War is an example of this, quite ironically, because Epic understands the nuances and trivialities of its own engines and probably had the game in mind before the engine was even built. One could say that the engine was built specifically for GOW and Unreal Tournament 2007.

[Crossbar]

I wasn't advocating dropping all use of libraries or even complete middleware engines themselves. I was just saying that most big budget games benefit from having their own proprietary engine. In a way, Gears of War is an example of this, quite ironically, because Epic understands the nuances and trivialities of its own engines and probably had the game in mind before the engine was even built. One could say that the engine was built specifically for GOW and Unreal Tournament 2007.
OK, I get your point.

However I recomend this podcast:
http://gamasutra.com/features/200606...dcast_01.shtml

From that I learnt Midway moved all their production to the UE3, that sounds pretty significant IMO.

[cpiasminc]

From that I learnt Midway moved all their production to the UE3, that sounds pretty significant IMO.
For a company as large as Midway with so many studios scattered over several areas, it becomes rather important to standardize not only to be able to share technologies, but also to weed out talent. It will soon become easier to find artists or scripters or programmers who have UE3 experience, and that's part of the value for them.

[Crossbar]

For a company as large as Midway with so many studios scattered over several areas, it becomes rather important to standardize not only to be able to share technologies, but also to weed out talent. It will soon become easier to find artists or scripters or programmers who have UE3 experience, and that's part of the value for them.
They probably got a pretty good deal from Epic as well, given the size of the deal.

The more developers that learn UE3, the larger momentum the UE3 will gain. At a certain point the developers will start propel the UE3 demand and further development. Which engine do you think is the largest competitor of UE3? Is that the Crytek engine?

[Smokey]

actually can you tell us cpi which do you think is the best engine if not ignore me :)

[cpiasminc]

The more developers that learn UE3, the larger momentum the UE3 will gain. At a certain point the developers will start propel the UE3 demand and further development. Which engine do you think is the largest competitor of UE3? Is that the Crytek engine?
UE3 has competitors? Unfortunately, the volume margin (or at least apparent volume margin) between Unreal and other engines is pretty large. In all fairness, an engine like Crytek could actually have more licensees than UE3 does, but many of them are small studios who'll never publish a successful game in a million years (Crytek is about 90% cheaper than UE3, last I checked). Croteam's engine is a similar story -- actually quite solid and cheap enough for small developers, so it might actually have a very high volume.

actually can you tell us cpi which do you think is the best engine if not ignore me
Well, I'm partial to an RTGI engine I worked on for my senior thesis -- ran smooth on a 128-node Origin cluster... Not that I'm biased or anything. :angelgrin

Seriously though, that's like asking which car is the best on the road. If you seriously want an answer, I'll go with this one (http://www.bath.ac.uk/~ccsshb/12cyl/). That's the best I can tell you.

[Nameless]

Now that's an engine CPI!

[yoshaw]

Well, I'm partial to an RTGI engine I worked on for my senior thesis -- ran smooth on a 128-node Origin cluster... Not that I'm biased or anything. :angelgrin

CPI, Please correct me if I'm wrong. When you say RTGI engine, do you mean the RTE1080 engine being used for the WarDevil game?

[cpiasminc]

No, I don't mean that at all. Like I said, I did it for my senior thesis, and it ran on an SGI Origin cluster (NCSA)... that was 5 years ago.

RTGI - Realtime Global Illumination. It was sort of a job queue approach that stepped rays through an octree to sample diffuse-diffuse interactions at each vertex. Fine for LD+, and L(S*)|(D*), but not LS+D.

[Smokey]

http://www.bath.ac.uk/~ccsshb/12cyl/

lol cpi, ill go HOLDEN mate :)

[cpiasminc]

Whaaat... you don't relish the idea of being able to clean your cylinder walls out by jumping right into them?

[Smokey]

Whaaat... you don't relish the idea of being able to clean your cylinder walls out by jumping right into them?
it could be fun i spose...cut down on mechanics & such :smoke:

[Crossbar]

I hate to put this up, but the most recent indicator came from the Inq:
...Documents shown to us on the trip showed that there were two versions, labeled 'CEB' and 'DEH'. The earlier CEB has a chip called CBE (yes, I know) connected to main memory and to a south bridge. The SB is then connected to a chip labled 'NV4x' over a PCIe link. The upcoming version DEH has the CBE and main memory, but the NV4x/SB combo is replaced by the RSX. If you wanted to know what makes RSX more than a GPU, it looks like it is SB functionality...

OK, now we can put the integrated South Bridge in RSX rumour to rest, it's been debunked by a trustworthy PS3-developer at B3D:
http://www.beyond3d.com/forum/showpost.php?p=778734&postcount=30

The inquirer has once provided us with a load of BULLSHIT, which is just business as usual. When will we stop paying attention to them? I suggest from now and on. :)

[overclocked]

OK, now we can put the integrated South Bridge in RSX rumour to rest, it's been debunked by a trustworthy PS3-developer at B3D:
http://www.beyond3d.com/forum/showpost.php?p=778734&postcount=30

The inquirer has once provided us with a load of BULLSHIT, which is just business as usual. When will we stop paying attention to them? I suggest from now and on. :)

I think he meant it in another way atleast if you are refering to the SB thingy.
All the rest from shitquirer is shit as we all know. :angel:

[cpiasminc]

He did mean it another way. The point he was making was not that the southbridge would or would not be integrated... it's that even if it is on the same die or the same package, doesn't mean it's actually interacting with the GPU or VRAM or anything else in any way. It's basically a separate component unto itself.

The only thing it would really share on the same die with RSX would be a common FlexIO controller. Which is certainly easier on the design considering you can have all 8 FlexIO lanes connecting to only two points, and the distribution of lanes is on-chip.

[xbdestroya]

Man I had no idea that thread was getting so interesting.

Ultimately, I really would like some sense one way or the other as to whether it's on-die or not. It's definitely true that just because the abstract of the bandwidth distribution on the chart shows a discrete SB, doesn't actually mean that it's on a physically seperate chip. But knowing for transistor count purposes would at least allow for a more accurate allocation mentally. :)

[Crossbar]

@overclocked & cpiasminc: It seems like you guys know a little bit more than the public information.
Anything more you want to share with the rest of us? :)

I guess in that case Mr Wibble just worded his post very poorly:

Lets assume for one moment the inquier's rumor that RSX is acting as southbridge is correct,....... The southbridge acts as a southbridge and the graphics chip does the graphics. ...
His choice of words sounded like two discrete components to me.

[cpiasminc]

Ultimately, I really would like some sense one way or the other as to whether it's on-die or not. It's definitely true that just because the abstract of the bandwidth distribution on the chart shows a discrete SB, doesn't actually mean that it's on a physically seperate chip.
Lest we forget that the original leaked block diagram of Xbox 360 showed a discrete Northbridge rather than one integrated into Xenos. It also suggested that the CPU was an MCM of sorts, even though logic would have clearly said that was silly from a cost standpoint. The main thing is that it is supposed to be a bandwidth layout, so things are separated for clarity.

The diagram with a separate southbridge, for that matter, is one of Goto's little speculations like almost everything he puts up. He's Japan's "version" more or less (okay maybe not that bad, but along those lines).

[Crossbar]

Thanks for the clarification. I can see you worded it carefully. :)

MrWibble also clarified himself somewhat:
http://www.beyond3d.com/forum/showpost.php?p=779152&postcount=40