Lordfly Digeridoo

Warning: Serious TL;DR ahead! Skip to bottom of the page for the Reader's Digest version.

Perhaps it's my lack of sleep, or my serial abuse of wanting to start grand, large projects, or both...

But I'm suddenly consumed with the idea of coming up with the "ideal" OpenSim grid.

It's obvious LL is at the very, very least a full year away from becoming a "grown up" company under the new CEO. While their grid software is more feature-complete than opensim, OpenSim has an army of passionate people who fill the SVN with so many bug fixes, incremental improvements, and efficient ideas that it puts LL's Jira to shame.

In other words, it may be only a matter of time before OpenSim surpasses LL in terms of features and stability.

This is essential, in fact. While OpenSim is a great tool, and there are at least half a dozen companies using it as a cornerstone of their business, it is not yet ready for prime time. Currency is a bit iffy, as well as the security of content. Every time I've connected to an OSim grid it's been frightfully laggy.

Whether this is due to an improperly configured client, not-properly-executed protocols, the very nature of the beast, or whatever, I'm not sure. I'm sadly not a programmer, just a user and occasional content producer.

On the flip side the sheer potential of OpenSim blows almost everything else out of the water. I cannot see a VW project coming out in the next two years or so that has the perfect storm of potential:

1) Rabid users ready to create, consume, and make an economy. You can make an outrageously efficient client/server architecture , but maybe the art content sucks. Maybe the art direction is amazing but the technical bits leave a lot to be desired. Or, worse still imagine you have both of those nailed down but no one comes. Contrary to Field of Dreams, if you build it, sometimes they won't come.

2)SL has a built-in pile of rabid content creators and consumers. They're used to the freeform, madcap economy, the infinite potential of content (can you imagine Twinity or IMVU deciding by committee to make a raincoat in a VW that doesn't rain?) The economy works because it's peer-to-peer for the most part, with LL acting as mediator.


3) A stable, mature platform already proven that isn't hindered artificially (quick, name the activities you can do in IMVU, Lively, or Twinity. Chat, chat, dance, chat, try on shirts, cyber with people who don't speak English). There's nothing wrong with limited-use virtual worlds, but after a while they all seem to be doing the same thing. Something like a billion dollars was invested in VW tech last year. I suspect 90% of it is going to be wasted on light, browser-based "social worlds" with a closed content input, limited social interactions, and poor community building.

How many Twinities do we need running around? Honestly?

SL might be fucking broken sometimes, but it's running, it's live, and it's doing stuff nobody's ever fucking done before. OpenSim can look at SL's mistakes and implement stuff right the first time, rather than swapping out stuff while it's running. See: Havok4, Mono, the source code spaghetti apparently causing all the server bugs (that they've mostly gotten nailed down, at least compared to 2 years ago).


SL users are known, fairly or not, for being obsessive and passionate about their favorite hobby. Some have jumped ship to OpenSim, but to the "outside world" it's still SL, just a whacky version of it. This...

....

Okay, I'm fucking rambling. Sorry, I haven't slept much this weekend due to caffeine, beer, and parties.

And SLUniverse is complaining about my post length. continued next...

__________________
http://www.lordfly.com/

Lordfly Digeridoo

Part deux:

Let's try this again.

1) I really like the concept of OpenSim, but it's implementation by the various grids leaves something to be desired, due to either technical limitations or improper setup. Laggy, weird economy, insistence on amateurly-designed areas, no customization of the SL client.

2) I'd love to help with an OpenSim project but my free time isn't free anymore - if I ever got involved with an OS project, grid-wide, I'd need to be paid somehow, be it with phat lewtz or free sim hosting or something. Doesn't mean I'd be an expensive asshole, just saying.

3) I don't program. I'd love to learn, but my brain isn't really wired that way. Instead I'm a bit of a small ninja with the SL building tools, and I come equipped with a background in city/urban planning that can see spatial design from a top-down perspective, soemthing I think most of these grids are going to need, REALLY fast.

4) It's great that most grids are about as wide-open as possible, but the intial user experience tends to leave something to be desired. Every time I've ported to an OS grid it's been laggy, and the 500-sim dropdown menu is completely fucking mind-boggling. You NEED an intial user experience that isn't a bunch of shitty cubes slapped with red and green spotlights in the middle of a grass prairie.

Even LL is realizing this - notice how they're switching their own concept of LL mainland from "crazy shithole" to "LL-managed estate"? Yeah, they realize the #1 complaint of new users is "what the fuck is this ugly place and what do you do?"

So anyways, I want to see the following in an OpenSim grid, hopefully possible, and who knows, if someone gets a bug up their ass about it and starts implementing it i'd like to help. Add your own!


Controlled spatial expansion and a "master plan". LL does this, sparingly. They didn't used to at the beginning, but with the first few continents added in 2004 they added roads and whatnot. While admirable, it doesn't go far enough in my opinion. Mature and PG sims need to be clustered. More "God-owned" infrastructure needs to be about to guide development - roads, landmarks, cities. LL is doing this with the Moles (something I miss doing, alas).

An ideal OpenSim grid needs a paid stable of content creators ready to make the world pretty. SL can be gorgeous in the right hands, and people tend to be inspired by their surroundings if done right. You don't want to know how many times I've ported to Abyss or Tableau and just wandered around staring at the content, generating ideas.

I realize I'm basically making a case for "you guys need a city planner, hire me immediately", but in the broader view it makes sense. If you want to make money you need to retain users. You NEED a proper first-hour experience or you're just another pile of shady businessmen hosting a bunch of random computers (see: CentralGrid).

You NEED a smartly-designed webpage that leads you through the initial setup - avatars, how to move, how to download the client, etc. Nothing wrong with using off-the-shelf software here - Wikia, Wordpress, Joomla, whatever. But at least make an attempt to unify the design, guys. One of the grids, I forget which, had as it's main webpage a wordpress page embedded in a frame with a Wiki slammed into the side of it for good measure. Yes, a whole fucking wiki.

Design Moar.


You NEED a compelling and tested tutorial experience. If a user has no idea where to go when you drop them in, you're toast. Grid operators need to start getting past the mindset of tossing people into the map and assuming they're all early-adopter users who can put up with crazy shit. That might be true now, but wait until LL does something that sends a lot of people OpenSim's way. No, it isn't OpenSpace pricing, but it might be a price hike of estate sims, or a removal of Mature, or the integration of the teen grid.

When that happens you should be seeing dollar signs. You can ensure that by coming up with a controlled initial experience for new users - a single set of designed tutorial sims that acclimate people to your grid. A list of rules. Maybe an introduction to the roleplay elements of your grid if you have that. Maybe a note saying that your grid is specialized for "BDSM bunny fetishists". Whatever. Let the user know what it is you're doing, what your goals are, and how you can help the new user enjoy their experience.


You NEED a stable sim setup. I know, this is like hoping for a Ferarri to pop up in my yard, but care should be taken. Sims adjacent to each other spatially should be in the same co-lo, to minimize weird crossing problems. Sims that are up "sometimes" (like hobby sims or from another hosting facility) need to be sequestered away from your "mainland". Yes, I'm stealing LL's ideas again, but it's a good one, if done terribly.

If your grid is open-connect, you really need to state what those sims are in a logical way. Too many times Iv'e tried connecting to a sim or a grid and gotten a disconnection instead. That's not very inviting.

No, OpenSim isn't 100% stable yet, but it's getting there. Don't enable features that aren't tested or aren't completely finished (currency, maybe?). Prevent content loss with a solid backup plan. And so on. You know, common sense Sysop procedures.

You NEED a customized client. This isn't always necessary, but if you really want to draw people into your business you have to wow them immediately - this means branding a custom-rolled client, with a new UI and text designed for YOUR grid, not LL's. This doesn't mean ban any other client,but having your own to introduce people to your world is a great way to keep them there.

Bonus points if you go through usability testing or have a "newbie friendly" client with minimal building tools and easy to use camera controls and movement.

You NEED incentives to lure independent content creators to your grid. Why should I make prefabs on your grid? What's in it for me, especially if you don't have a proper currency economy? Will you integrate paypal instead? Will you showcase my work on the front page of your website? Will you pimp me on Twitter and Facebook? Do you offer better building tools in a custom-rolled client (megaprims, a rotation interface that isn't retarded, defined texture guidelines, and so on?)

I'm kind of running out of ideas, but finally you NEED to start small. Throwing open your grid to all comers, letting people hook up their hosted sims to yours, and having a big club-orgy party thing sounds great, until you have to manage it.

So, tl;dr, right? Summary!

Opensim is neat, but sloppy. Fix this. Let's brainstorm the ideal utopia grid.

1) Don't always assume you'll be catering to only early adopters. Plan for the new users.

You can do this by:
2a) Planning the initial user experience before you ever open your first sim for business.
2b) Customizing the user experience with smartly-designed webpages, a custom-rolled client, and helpful tutorials.
2c) Figuring out what your grid is going to be doing, sticking with it, and guiding (not forcing) user interactions to go that way.
2d) A properly designed mainland/core set of sims that are inviting, professionally done, and give the new user a sense of wonderment.
2e) Hiring a design team.
2f) providing incentives for indie content makers to make top-notch stuff on YOUR grid and not your competition or LL's.
2g) Sequestering self-hosted sims away from the mainland, your "user experience" setup.
2h) Usability testing.
2i) proper website design
2j) hiring Lordfly to do sim building and planning (only half kidding, but srsly get someone to plan out expansion and spatial growth)

SO WHAT ELSE GUYZ?!?!

Lucifer Baphomet

Good thoughts, Lordfly.

__________________

radarm

On my blog (radarmasukami.com) I'm blogging about my experiences checking out Openlife Grid, which seems to be by far the largest OpenSim grid out there.

I have to say, these guys are going to go through HUGE growing pains, things that LL's put behind them already, and anyone expecting them to suddenly blow by SL are gonna be in for a surprise, IMHO.

I think the OpenSim concept is great, but for anyone to maintain a large VW based on it is going to be challenging, to say the least.

Lordfly Digeridoo

Any business venture is going to be challenging, doubly so on not-yet-ready-for-primetime software. But I think that if done right, and carefully, and slowly, a compelling experience could grow.

Beebo Brink

Every day on my way to and from work, I spend that half hour geekily planning out various aspects of my own "ideal" grid project.

I agree that the user introduction and experience is key to retention and it's been woefully mishandled. So, here's notes from my own musings:

Tourest Membership Sign-up

Free registration and entrance to the grid on a Tourist Visa plan that doesn't cost you anything (but also doesn't allow your stuff you Inventory with freebies).

You have your choice of Tourist Packages:

Select the backpacking avatar for tours of wilderness and landscape areas of beauty throughout the grid. You can set home to a variety of hostels scattered about the grid, where you can ask questions and get mini-tutorials on how to function on the grid. A packet of LMs provides an itinerary of places to tour.

Select the festive Hawaiian shirted avatar for a tour of the beach resorts and night clubs. Or just sit on the deck of the hotel room where you're allowed to set home for a designated amount of time (say 2 weeks) and drink hurricanes. Attend the hosted mixers, go strolling on the beach, take the scuba diving tour.

Select the urban chic avatars for music events, art galleries, fashion events, and tours of major metropolis areas of the grid.

And so on. Heck in one afternoon I could draw up the rough outline of about a half-dozen tours that would capture the essence of the most popular inner fantasy life of the majority of visitors and hook them right up to their dream drug of choice. Tailor the entry point builds to showcase that interest, provide tutorials based on that theme (Want to learn how to buy something? Here, buy this cute hat!)

And for god's sakes take those STUPID kludgy images off the SL website home page and replace them any of the hundreds of stunning photographs that display avatars and landscapes at their best, not their worst. Hell, the Show Your Style thread on SLU has enough mouth-watering photos of fashion choices to have millions of women yearning to hit the grid and shop till they drop.

__________________

Humps

Excellent post Lordfly,

I agree that whilst SL is the obvious model for teh Open source worlds, if they are to have any impact in the near future they need to distance themselves from 'the mother ship' as it were and carve out thier own identity. After all what would be the point of a simple clone of SL that frankly is not as stable at this point in time or is not as populated.

I think now, at the start is the time for the different grids to start carving out their own identity. In world experience goes a long way in conveying your brand to the customer, first impressions count and all that, so I agree 100% with you that a sloppy welcome area, sloppy web infrastructure and sloppy planning in world don't exactly shout professionalism.

I think where the open grids beat SL/LL hands down, even at this embryonic stage, is in promoting and engendering a sense of community and inclusion. That above all is what has drawn me into the Open Grids. I think this is tremendously important and a huge asset. However that alone is not enough. Community is all well and good but it can get chaotic quite quickly. There needs to be cohesion and someone to prioritise. I think the successful grids will be the ones that can balance the two ideals, stamp their own brand onto their grids and offer the experience that people want. A decent affordable product, well designed and presented, and excellent community/customer support.

I think if you build that ...they will definitely come.

Lordfly Digeridoo

What's nice about OpenSim is the concept is inherently scalable... my perpetually unfinished Nu Athens project would be, at most, 9 sims. It would be designed for no more than a few dozen people online at once. And that's fine - it's a story-telling experience, not an MMO.

Although I admit I did spend a long time today at work staring into space wondering how much startup money I'd need to start my own grid...

Beryl Greenacre

My ideal grid is one where nobody calls me "ma'am."

Hi Lordfly.

Malachi

However the thing that makes it a virtual world is not simulated contiguity of terrain, but user data that spans the worldlettes. That is, the LL central services (presence, inventory, etc.).

I've not been following OpenSim much at all, but the distributed sims is a great idea. If they could manage to tack on a cloud-like central data repository. It could become a world.

For example, when I hit Gmail, I could be talking to one (or more) of thousands of servers all of which know "me" and have "my stuff". An archipelago of disconnected worldlettes would be much like the fragmented web of today, no site has any idea that I am the same persistent "me" across NYT to Ars Technica, etc.

If there is OpenSim work along these lines, pointers would be much appreciated.