Spot the differences!

Were looks alone the only pertinent item, perhaps, but you're really making an astoundingly bad point with all of a screen shot to back it up.

Let's agree Blue Mars looks great. I think that is a given. As this actual review (screen shots and words) http://virtualunderworld.net/wordpress/?p=867 points out they took an off the shelf engine, avatar, etc to get there. All Linden needs to do is rid themselves of the shackles of the legacy environment we happen to be using (and making them money on) so they can start over. See whoever is still waiting in a year.

Graphics aside, this review is spot on where it counts. This is a not a world where you can make content and go sell it. There may be an approval (read censorship) process and almost certainly fees for this service.

Not running Windows? Well the beta is. And only Windows.

You like to window shop? Without a currency this is all you can do.

Got an 8000x series card?

Like awkward avatar movement?

The review sums it up very well in the final words.

” The graphics are great, I admit, but as I previously mentioned in a past post, great graphics do not make a great game.

At this point, Second Life and There have nothing to worry about.”

I think it was six months or so back you had a poll on how much of the SL pie OpenSim would have. I should of taken 0.32% at the time but I'm just now getting your thing for going to the track and betting money on three legged horses.

Spot the difference: One world has crappy looking avatars that are all teenagers with paste for skin and the other one has advanced avatars with prim attachments and a full economy that is not part of what can be “seen” and in addition will not have content made by “pro game developers” who, in all honesty, make crap because they have to manage triangles so their hair is cloth, there is no intricate detail on anything, all “detail” is painted on. In one there is a complete user created scripting system that allows residents to create almost anything they want while in the other what is there? In one there isn't mesh import yet but there will be causing the playing field to be leveled a lot in that respect. In the other good grief look at the ground. it looks so fake it isn't funny. Again that is pro game developer style work.

Oh in one there is adult entertainment and just about anything else you can dream up. In the other? you can play golf i hear. And type with childish looking bubbles over your head.

The biggest differences are not in what you can see. the big differences are in what you cannot see that makes one a multi million dollar a year economy for resident businesses on it's way to billions. the other is crippled out of the gate and will never be a money maker for the players. Only for the company assuming they get all those corporate accounts that they used trademarks from in the promo vid.

… even corporates, to a degree, are not stupid enough to jump into a virtual world that has nobody there

Aye, well, I don't think that there is such an overwhelming difference in the resulting images from either platform: the real difference comes from having good content creators, not from the underlying graphics engine. And there are hundreds of thousands of good content creators working in SL right now.

I've heard over and over again how Blue Mars incorporates the latest and greatest of detailed rendering tricks available to all high-end graphics cards. That's all very nice to know, from a purely technological point of view — it's good for the extreme 3D techies out there. However, even the Avatar Reality 3D modellers use baked, static shadows on most of their buildings to save polygons — instead of fiddling with dynamic shadows and texture maps. They're just more computationally expensive. Even game designers know that, just because you have all those awesome tricks waiting to be activated on the engine, you should be careful on the amount of client-generated lag they create… for an end-result that is probably not substantially different. Yes, you can accurately model an angora sweater where each fibre moves with the wind in Blue Mars — but will you really wish your users to be subjected to that amount of intense computation?

So it's a complex issue to deal with. Granted, developing high-end games for Blue Mars will be much more fun than doing them in Second Life, I'm sure of it. SL is simply too unpredictable in its stability, and too slow in the scripting, to allow high-speed realistic games. There is certainly a niche market to explore in Blue Mars for that — and it would be more reasonable to assume that their competition is Multiverse or VastPark (future unknown), not SL.

That's a good point He might be able to afford it, yes…

In a conversation with Extropia DaSilva yesterday (who is much more of gamer than I am), she made the following remark: when a brand new product is released, it's inevitable that one compares it to the previous generation. However, a brand new product might have pretty graphics, but it's not explored to its uttermost potential. So the fair comparison would be to compare Blue Mars with Second Life in, say, late 2002. Not with a immersive, vast virtual world, with a teeming economy and millions of (mostly) amateur content creators (but with 100,000 professional ones). The comparison would never be fair for Blue Mars.

Also, although Second Life back then was also targeted to attract developers, like Blue Mars now. But old-time Second Life veterans are looking for a platform that might attract consumers, not developers. Blue Mars has not yet any compelling reason to attract consumers — simply because independent developers have not had time yet to develop complete games/social areas in it, so that the consumers might have a feeling of what Blue Mars can actually offer.

Imagine that Blizzard joins up the developer programme and creates an independent city in Blue Mars with their teams, creating a “demo” MMORPG that looks like WoW on steroids. I'm quite sure that the visual impact of Blue Mars (compared with WoW!), fully exploited by the best graphic designers and best game designers in the world, would be more than compelling for the mainstream users So we'll have to wait until they start to develop things seriously in Blue Mars and see what happens.

Right now, Blue Mars' open beta is simply at a too early stage (no content from independent developers yet) to be worth of any comparison…

Well, I did understand your point, Ari — “eye candy” is not everything, and, really, the eye candy in Blue Mars is not that impressive. Sure, it's marginally better than SL at it's highest settings (e.g. using Kirstens Viewer, which fully implements a lot of half-done, half-abandoned new LL-created lighting models, on Ultra settings with shadows on), but not that better. For instance, I was a bit disappointed with the avatars: from far away (on the default camera viewing settings), they feel better than SL's avatars (if you ignore for a moment their exaggerated anims), except for the skin; but when they're viewed closed-up, they're ugly plastic dolls with a sombre expression. Nevertheless, they seem to have a higher polygon count than LL's own 7,500. So there are technologically advanced “edges” here and there — the surf on the beach is really convincing, but I don't believe that a massive exodus will drive people to Blue Mars to watch the surf Overall, despite the low FPS on some cards/hardware combinations (I haven't figured out what makes Blue Mars run so fast on some PCs, while it lags behind on several others, sometimes with better graphic cards), the platform feels stable and more fluid than SL — but then again, I wouldn't have expected any less from a mostly-static-content environment, where the majority of scenes can be neatly pre-rendered and are cached on your hard disk anyway (thus the huge download for merely 5 cities, even though some of them might actually be downloaded before you enter them — that's why BM says that each city can have up to half a Gigabyte of content, since they assume that amount of content can be downloaded reasonably quickly to your hard disk in a relatively short time when using broadband, and they're probably right).

I'm really not impressed with the “eye candy” because I happen to know that it's unfair to compare a scene on the SL mainland on low settings (but high FPS) with professionally-created content in Blue Mars on the highest possible settings I can only suggest that people try other viewers, ramp up their settings in SL, upgrade graphic drivers, tweak their card's and PC's settings, spend some time in tweaking performance, and then jump to the thousands upon thousands of professionally-created areas in SL. The difference will be small and mostly small details; dissing LL's engine as “old and outdated” is actually quite unfair. LL's engine is almost up to the highest standards — there are just 2 or 3 missing things, currently under development (and I'm NDA'd not to reveal them ), but it's by no means “lagging behind the competition”.

They have just two problems. One is supporting legacy content, as I commented on the Blue Mars forums. This mostly means that we can't “suddenly” get 50,000-polygon-avatars with a wide variety of skeletons, not because LL's engine doesn't support it, but because it would totally and completely break millions of skins and billions of articles of clothing LL's not insane enough to do that. Sure, Blue Mars's hair looks way nicer; but once flexisculpties become a feature of the main viewer (it has been a feature on several alternative viewers for over half a year now…), you won't spot any differences between SL hair and BM hair.

Besides supporting legacy content, there is the whole aspect of the target. SL started also by being a “developer's environment”, but LL quickly understood that the mainstream users also wish to have fun developing their own content. Collaborative, real-time user-content generation without limitations or pre-approval, in a contiguous environment, requires quite different techniques for the rendering engine to deal with. I'm absolutely flabbergasted with the way SL calculates shadows from multiple sources on dynamically changing content so quickly! It seems a miracle to engineering to me — since as a matter of fact I haven't seen any 3D rendering engine that can handle that. And on top of that you can dynamically change the lighting and the way the light sources affect the shadows using the Windlight settings — again, in real time. If in my college days I'd describe what SL's engine is able to do on a relatively normal (as in 2-year-old) desktop computer to my teachers, they'd think I'd be insane and would label it as “science fiction” — specially when I'd explained to them that this is not ray-tracing

Blue Mars, by contrast, uses the much more safer and less resource-intensive (and also way more researched and deployed) method of assuming most content is static and pre-loaded, so you can pre-render a lot of things before the first avatar steps into a city. Cities are isolated environments in the sense that there is no visual contiguity; you don't need to worry if, for instance, on a neighbouring city someone has just uploaded a huge building and now you need to figure out what happens to the shadows and lights on your city Instead, you know in advance which objects are dynamic — avatars, 'bots, vehicles — and even limited-dynamic items (doors opening/closing — which in fact I haven't seen yet in Blue Mars, but I haven't fully explored it yet) can be pre-rendered effectively. This makes things so much easier for the BM developers — and the result is that for the users the experience is more fluid, more immersive, and the platform more stable.

No, “eye-candy” is not what will make BM a success, but the kind of developers they attract to create compelling content and games on the cities. Imagine Blizzard doing a small-scale WoW on a city — but with the stunning graphics of Blue Mars, as opposed to the low-polygon engine that drives WoW Now that will certainly attract people And I believe that's what Avatar Reality is really after: high-level, professional content developers, that get a hosting service with a nice licensing system, and includes an economy which will allow developers to not only charge for access, but for buying items in Blue Mars. So it's like Multiverse/VastPark with a slightly different revenue model. If you wish, you cold imagine access to Azeroth in Blue Mars where users would not be charged anything monthly, but would use Blu's (BM's currency) to buy the items they need I can imagine that this would be appealing to a vast crowd of gamers and consumers.

For us SL residents, the social stratification of Blue Mars might feel too constraining. Money flows from the bottom to the top only: colonists (the users in BM) go to shops and buy items with Blu's. Shop owners pay rent to use the shops, from shop designers. Shop designers lease land from Block owners. Block owners lease space from City owners. Since City owners require a special agreement with Avatar Reality, and will not be sold to “anyone” with a credit card, but only to whom AR wants, the hierarchy is solidified. There is no “upwards mobility”. Even if you start spending lots of Blu's and finally sign an agreement to become a developer, you'll need to rent a shop. You can't simply say, “oh, I'll buy my own city and create as many shops as I wish” — because access to whole cities is limited by AR. There is also a question of transferable content between cities, I have no idea how that works.

In SL, by contrast, we have an equalitarian society. Everyone, from the very instant they logged in, is automatically a consumer, an explorer, a designer, a content creator/artist, a land owner — there are no restrictions, no limitations, no “special agreements” to sign, and everybody can be whatever they wish, and change their minds every day. The result? The anarchic mainland But that's the price you have to pay if you give people freedom. Avatar Reality wants to avoid “bad content” on Blue Mars, so I understand their choices…

i'd like to add – to help dispel this myth that the SL renderer is poor – that any time we try to render our content in a commercial engine (like the cyteck engine) the framerate drops to single digits.

we don't use our renderer because we're infatuated with it. we use it because it's better.

That's a very interesting observation, qarl! It would be very nice to have some tests and comparisons to show, side-by-side, how SL's renderer fares compared to others, using the same content.

And by “same content” I would definitely like to see how well Cryengine (or any other engine) deals with 100 avatars with 7,500 polygons each (base) and perhaps 20-50,000 additional polygons for all attachments they have Dealing with “millions of polygons” per scene, 50+ FPS, is not a simple task. I remember the first time I played Spore with the default settings. Oh sure, it got 70 or so FPS easily — but it was clear that each scene had just a handful of polygons to render

So, yes, this would be a much fairer comparison: how well does CryEngine2 or Unity or whatever is felt to be the “best” engine out there do when they have scenes with 20-50 million polygons to render?