Is this good or bad news for SL?

IMHO the dedication of the host company to a virtual world's positive experience will be the main factor in how successful they are. Technology is wonderful, but less important than if people are having a good time.

LL isn't very impressive technically, and they've certainly made a lot of people pretty annoyed, but they do seem to keep a lot happy as well.

While I realize Blue Mars is still in Beta, I'm concerned that the AR team seems to not want to get involved in the overall user experience. They seem to be trying to push a lot off onto the City Developers rather than taking responsibility for the entire platform. I question if that's going to be viable.

Whatever they paid, it was too much. "There" is an un-googlable word. You can find the site itself easily enough, but you can't look for any third-party blogs, message boards, or news media stories about it. You can't really research it or do any kind of due diligence. It must surely be a handicap for marketing or business development purposes.

For what it's worth, this is my summation of the SL experience, from my point of view as essentially a high function luddite. SL is relatively easy to start using, I've always considered the "learning curve" a myth. You don't need an MIT degree to learn the basics functions, I am proof of that. It does offer a wide range of activities, it isn't focused in any ome area, which I think is it's greatest strength. It is driven by user generated content, and creating that content is possible for any user who wants to with the provided tools as a minimum.

Considering the scope of SL, the range of users and equipment utilized, it is remarkably stable, even if it does frustrate us. In creating SL, LL has gone where no one has gone before, and they are still alone, having maintained as other competitors have fallen by the wayside.

LL, as a business entity on the other hand......

I spent about an hour on the phone with him today. The topic was "What Second Life Users want to know about Blue Mars".

We were both taking notes as we talked, and he's got a copy of my notes for review, since I want to be accurate before I start quoting him. But what a refreshing thing to talk one to one, and even to have some of my suggestions taken on!

Some newsworthy items came out of the conversation:

- A major beta update is due the first week of December, with new features and revamp of the movement system.

- Pricing will be announced at the same time, along with plans for further features.

- Rather than a 1.34GB installer (application plus cities in one package), they will go to a smaller application installer, and separate downloads for cities.

We covered a lot of questions, and I plan to make my notes available once they are ready.

Still not streaming, then?

My thoughts exactly. I idly wonder how big your harddrive needs to be for, say, visiting 35 cities...

Current places are in the low 10's of megabytes. A lot of that must be down to the systematic use of textures - as compared with the wild west of a mixed-use SL sim.

Your 35 cities are likely to need 100's of Megabytes rather then 100's of Gigabytes.
Hard disk is dirt cheap anyway - and almost free compared to the sort of high-end graphics card needed to run SL or BM well.
It's also worth remembering that a MB 'city' is equivalent to a small continent of SL sims. One can move around in that continent without any texture lag or sim-crossing nonsense.
One can also see right out to the horizon.

Blue Mars is not a 'world' like SL. It is a platform for developers.

With esata Terabyte drives currently costing about $100 who cares? It's much more likely that big hard disks and fast processors will be cheap and available in the near future than multi-gigabyte broadband connections. Think of it as an essentially unlimited local cache which allows for much higher visual quality without the drawing lag rampant in SL as you move around.

I take this to be a comment on the whole notion of discrete downloads rather than a background continuous updating process.
If that's what you mean then I completely concur. As discussed above, local content storage in a giant cache seems like a good idea to me, but requiring explicit downloads even if they are only a hundred megabytes seems like a design out of the last century.

One city requires a 1.35 GB installer.

I'm so confused. Is Opensim an actual 3D site or is it a 3D Site builder?

Open Sim is an actual 3D site that you set up and host on your own computer.

http://opensimulator.org/wiki/Main_Page

Lol, I'm not computer smart to I won't even bother.

The bulk of a city's data will be pre-downloaded before you enter, as well as personally owned items (like your clothes for example). What will come in streaming mode are updates to the city data, and avatar info as you encounter other people.



I believe that a city data package maxes out at 500MB, The demo cities in the current client run 100-300MB, though they are somewhat "unfinished" (a small core area that's been built in detail, and a lot of unused room around that). So lets say 10GB to answer your question.

My Crysis Wars install runs 7GB or so for comparison. That uses the same graphics engine



Once installed, the breakdown is as follows:

Application folder 27MB
Shared textures, animations, etc 360MB
3 "full" cities & 3 minor demo areas ~1 GB

Opensim is software to host a close copy of a second life region on your own server. You can then build on your own, without logging in anywhere outside your PC, or invite people to log in to your region.

There are several grids using the opensim software to host regions either commercially or collaboratively.

This is the mantra that makes me concerned about their viability.

Avatar Research seems a bit too anxious to avoid any direct involvement in making Blue Mars an attractive destination for users. Their business plan seems to be based on attracting technically expert City Developer partners with deep pockets who will deal with all the messy work to make the platform a success while they work in splendid isolation up in their ivory tower. Note that currently there are no such entities.

I hope I'm misreading their attitude, but that's my current impression.

One way or another the data has to get to your PC to see it. In SL you start requesting the data at the time you TP, and then have to wait to fill in your draw distance. In Blue Mars, you get most of the data before you TP in.

In SL the cache is limited to 1GB, so if I travel around a lot, when I go back to someplace it has to be downloaded *again*. If Blue Mars puts in a good city manager system, where I can *choose* which places to save and how much total disk storage to use, that would both reduce their server overhead, and give me a better overall experience.

I was told there will be an auto-update feature, so any changes to a city since you last visited will get taken care of in the background via a subscriber model. In other words,
cities you select will be updated, others will be ignored.

All this sounds excellent to me, but it also seems like something that wouldn't be that difficult to add to SL. Allowing the local cache to grow to any size and adding a local manager for it without breaking everything else doesn't sound like an overwhelming technical task for a decent software team to accomplish .

My understanding as a Blue Mars city developer is that it is the same relative standing as an estate owner in SL. Estate owners parcel up their regions for other people to rent, and Blue Mars cities will be parceled up into blocks to rent. In both cases the top level owner pays the hosting company.

I've been messing around with my practice city to learn how their stuff works, and it has not seemed fundamentally harder than building in SL aside from two things: Lack of documentation, and not much pre-made content right now. I even made my own set of base prims:

http://picasaweb.google.com/danielravennest/BlueMarsTechnical#5398075329702776338


Which at some point I will package up and make available to others. So if you don't want to learn one of the 3D tools, you can just drag these from inventory and texture them.
On the SL side, I never learned to script or make animations or sculpties either, I just use stuff other people made, and that will be possible in Blue Mars also.

Aside from working on the application software, they *will* be involved in much the way Linden Labs is, as far as account management. That's all the messy stuff regarding payments, player abuses, etc. They put themselves in that position by having you register with them directly, rather than register per city to play.

That's not a limitation of the streaming model, that's a limitation of SL's "virtual file system" that the cache uses. It is fixable, they just have to choose to fix it.

That link is a blank page.

Are they as flexible as SL parametric prims, or are they like a collection of sculpts?

Something that has puzzled me. Why on earth would any company pay huge amounts for the new Enterprise product from LL, when presumably one could achieve the same thing using Open Sim?

And content from SL can be imported into Open Sim, can it not? Or has that changed?

IBM asked that question and repackaged Opensim as their new 3d spinoff for Notes.

why? presumably because LL offers support, upgrades patches, etc with no need to understand the technical end in depth. at least that's the idea.

as for impoting... yes and no... there are legal potholes when you get into things you didn't create yourself.

I don't know how "user friendly" Enterprise is, but surely anyone with the in-house IT to set up and run a $55,000 product could also set up an Open Sim. Or maybe not . . .

Yeah, I suppose the content creation thing is a problem now for imports, eh?

Because Opensim is still only in alpha, and if version 0.6.7 is any indicator then it is only around 2/3rds complete.

Enterprise is ready now, and for the money you also get the server hardware. Someone else has already mentioned the support.

Content can be imported into Opensim, and vice-versa, currently. One thing that may not import, however, are scripts, as Opensim scripting has already diverged from LSL. However, import/export might change as the Opensim-devs have made it quite clear that Opensim is not SL and never will be, and they reserve all rights to diverge completely from the SL architecture, which is a good thing.

Rock