The Kokua developer team has just posted an announcement stating what they now plan to focus on. Here is the list they published...
1.Focus on the next generation Kokua client rather than Kokua & Imprudence
2.Continue to support Linux 64 bit
3.Keep it easy to explore alternative grids (ie not Second Life proper)
4.Mesh support in Second Life and OpenSim
5.Support older hardware
6.Experimental builds
I was, in fact, beginning to wonder if the Imp/Kokua team was on the point of breaking up so it was a pleasant surprise to read the latest news. To be fair they did ask for user comments on what they thought the team should concentrate on at the start of the month but I was very late seeing it which probably reflects how far down in my consciousness they had sunk. Well, I did have high hopes at one time that the Imprudence team would come up with a solution to a pet wish of mine - improved grid manager - and, to date, it has never happened despite some noises during the past two years that alluded to the possibility it was being worked on - maybe.
Regardless of all other options I think it is vital that the Imp/Kokua team maintains and even increases their support for the Opensim and Aurora-based grids. The open Metaverse is expanding steadily and hypergrid travel is on the increase so we need a viewer that gives maximum support. Second Life is well catered for with something like 10 TPVs and the V3 official viewer but the open Metaverse is reliant on just few usable viewers including Imprudence which is probably the most widely used for Opensim. Personally, I think we have to move on now and Kokua is perhaps the best way to go if, and ONLY if, it puts a renewed focus on Opensim apart from support for many of the things expected of a V3-based viewer such as Mesh, Mesh upload, Moap, multi attachments etc. Making it possible for users to configure the UI windows, buttons and menus goes without saying but the kind of support that the viewer can give the open Metaverse should, MUST, include a Grid Search function to replace grid manager where grid owners can add their own grid address, info and logo/image. Currently, all the grid owners have to explain to noobs on their web sites how to add an address and it is probably a serious stumbling block to recruitment.
Here I doctored an image of Astra viewer to include the Grid Hop page open at OSgrid. Imagine entering a search term or key word in the search box I added to the footer bar and pressing the search button brings up the grid manager window with a list of possible grids and sims very similar to in-world search. In this case though it is not a localized search of the grid but search of the whole Metaverse. It doesn't have to be Grid Hop in the window which I used here by way of an example. It can be a completely new search engine that could even collect usage and traffic statistics for display. The viewer could, in effect, bring the open Metaverse together for statistical analysis which it surely needs.
What I would like to see is an improved grid manager with a search function and method for grid owners to add their own grids and standalone sims so new users don't have to add grid addresses themselves which a lot wont have a clue about but rather they can simply enter some key words to find grids of their choice. You need a registered name to login of course but why not allow logins to hg-enabled grids by adding the hg info to the user name regardless? That is, for my name, Gaga Gracious@hg.osgrid:80 for example. The grid manager could do this and I would log into OSgrid at a Plaza without having my name actually registered there (technically, I realize there are problems to solve in this method like obtaining login authorization via a back end data server but I feel sure it can be done). I think this could be the single most helpful feature that would benefit Opensim users and grid owners as a whole. By this method the grid list would always be up to date and noobs would never need to enter an address. I believe grid owners would welcome this feature and promote the viewer since they would be able to direct their new sign-ups to it knowing their grid is already featured and set up to go. No more need for grid owners to get someone to tweak the grid list and release a cloned viewer that does nothing for the Metaverse as a whole but is doing exactly what I am saying - only just for their own recruits. The sort of thing that would suit a grid Like SpotON3D that has no interest in supporting the rest of the open Metaverse unless they are in overall charge and profiting by it.
For the rest of us who believe in a free and open Metaverse where there is room for commercial interests and those who prefer mutual sharing then we need a viewer that is built with features that support multiple grids. The problem is, of course, the developers have their own preferences and ideas about what makes a useful viewer and not every developer is capable of delivering workable solutions to the huge wish list out there anyway. Even worse is that all to often we find that some, not all, but some developers suffer bouts of egotism especially when dealing with one another and cause projects to go awry and even fail possibly leading to a break up of the team. Some have even used their coding skills to hack and do bad things - Emerald Viewer being a case in point. But, with those cautionary words said I do think the Imprudence team genuinely want to work on solutions their users really want.
I have been plugging away at the grid search idea for a long time. Two years ago I asked for it on the Imprudence forum and the response was positive yet nothing has happened to date. In the last year I became involved some with the Aurora development team and the lead developer, Revolution Smythe did actually agree with me and went so far as to recommend work be done on Astra viewer to explore the grid search idea. Rev thought Grid Hop could be used and actually did some work changing the grid manager window in preparation but adding Mesh support seems to have occupied the developers since that time and, while I still get promises that the grid search feature is on the work bench, still nothing has happened to date. So that is where we stand and I can only hope the Imp/Kokua team do actually revisit these ideas.
The open Metaverse needs it's own viewer. That's what I think and sure it can be compatible with SL but the focus needs to be on Opensim grids. Too many developers, and I think some members of the Opensim team are guilty of this too, slavishly follow developments in Second Life as if they were the benchmark for virtual worlds. The grid manager and search would benefit the whole community and be a highly visible statement that Opensim can stand on it's own and become the prefered platform of the free Metaverse.
So this is 2012 and, before I launch into looking back at what I got right and wrong with my 2011 predictions last January, I have to mention some good news for Mesh fanatics desperate to see their beautiful mesh clothing move with the same grace as their pixelated bodies. I wrote about a crowd funded project back in October which was my last post for several months due to heavy commitments elsewhere. Anyway, the project was organized by Second Life residents who where frustrated by Linden Labs tunnel vision and lack of action. They raised money to pay a bounty to Karl Stiefvater, better known as Qarl Linden from his days at the Lab, to develop a parametric deformer, which will cause mesh clothing to fit the avatar shape and move with it more realistically.
Well, Karl has finished some preliminary work and put up a video to demonstrate things. Karl emphasis this is a first look and far from complete. The "heavy lifting" part is done he says but "I’m giving it to you in this form now, so that you can give me feedback. There are decisions to be made which we should make together."
Here is the demo vid and the source code is here. Feedback can be given here.
THE PREDICTIONS OF 2011...
First blowout to mention was that Opensim would come out of alpha in the last quarter of 2011. Well, no one has said it is beta so there you go. Still Alpha but, to be fair. I did say "maybe" when I predicted it would. On the other hand, the developers finally made up their mind to do something about the six month rule for submitting patches which everyone agrees was holding the project back and stopping people working on the viewer to. They set up the Overte foundation to handle the legalities the developers were so worried about. The advantages of the foundation are many since they can now raise funds and impose rules to make it difficult for Patent trolls hijacking parts of the project and stifling development and innovation
My 2nd prediction fell way short of my wishful thoughts but the latest release of Hypergrid dose have the Outward bounds permission and more work is being directed at security. Nothing yet has got Avination HG enabled despite the owners well-meaning statement early on in 2011 which said AVN would enable HG as soon as possible.
My 3rd prediction was a rather bald statement that more developers would pull out of TPV development. Well, actually, a few have, notably Kirstens and Jacek of Imprudence. But, on the whole I would say more rather than less are working on TPVs since Opensim devs relaxed their rules for submissions but getting stuff like Mesh to work in V1 viewers has got coders busy again especially finding upload solutions for Opensim where Mesh can be handled free while Linden Lab decided to penny-pinch (or nickel n dime as American's say) as usual.
My 5th prediction was that LL would lose more market share but, though they lost over 500 sims the traffic is still holding around 60k peek and 30k off-peek but, when you consider LL gets in excess of 10,000 new signup's a day, then they really are running on the spot and going nowhere. And, in my view, its not going to improve in the coming year since LL continues the same old policy of cashing in on everything they do rather than looking at new developments as part of the service people are already paying a high price for through tiers and setup fees. In deed, it seems to me they are not content to let the residents make any profit on top of what they pay to the Lab. What I see is LL heading off in the Farmville direction with Linden Realms when surly the residents should be building these games? I mean, the Lindens cleared the malls, markets and stores in-world so they could profit from their web-based market place and introduced Linden homes that compete with the land owners paying tier so they could bring in more premium signup's. Second Life is/should be a platform on which "Your World, Your Imagination" can bare fruits for the hobbyist, game maker and content seller. Linden Labs just can’t stand seeing anyone else doing well out of what they made unless they get a huge slice of the cake. Thankfully, Opensim is getting better all the time and the recent survey on Hypergrid Business clearly shows the open Metaverse has grown a lot since the last survey while SL continues it’s slow decline.
That’s the beauty of Opensim, it can truly claim to be Our World, Our Imagination from the ground up.
My 6th prediction was an odd one really since I said we would see a breakthrough in browser based portals into virtual worlds. Well, there was actually a lot of progress and no one solution that really grabbed a new market yet other than, perhaps, Unity 3D. But we did have a serious patent scare courtesy of SPOTON3D. Their browser solution actually worked quite well but it was almost the same as work done previously by others. What hit the news was that the owners of SPOTON3D had the cheek to file a patent on it which effectively stops anyone else from using the code - which was open source and in the public domain!
My 7th prediction about Blue Mars having a bright future if the money holds out? OUCH!!!! the money ran out!
My 8th prediction was more than bang on target! Kinect did come to virtual worlds successfully but, as far as I know, it got mostly used to build animations for avatars and not a lot else so far.
MY PREDICTIONS FOR 2012...
Well, I am not going to get carried away this time but I will stick my neck out and predict Linden Labs will pull out all the stops to get Second Life growing again. Really, they must, yeah? Or will this be the year they call the beginning of the end!
Enough of that. I want to make a considered prediction about OSgrid which currently is the leading grid of the open Metaverse. Based on the Opensim platform OSgrid is, in my view, the HUB of the Hypergrid with many small standalones connecting to it via hypergates. I predict OSgrid will grow substantially in the coming year because refugees from SL will continue to cross the divide in ever greater numbers. LL will pull the plug on V1 viewers soon and, since they will remain usable in Opensim with MOAP and Mesh this will have an impact too and might even herald the beginnings of a break with some of the SL protocols leading to a more Opensim-focused viewer.
Here Lbsa Plaza on OSgrid is busy most days. On this day 24 people were sharing information and getting help from mentors
I would also predict we will finally see a viewer for Opensim that includes Grid Search in one form or another as an option on the menu bar of the opening splash screen. This has been on my wish list for several years and the one thing I think that would really set the Open Metaverse apart from Second Life. I know there is a lot in common especially in the area of content but LL is never going to allow content to be transferred to Opensim grids even while so much that is sold in SL is actually made in Opensim!
It's true, people with large SL inventories are held back from going to Opensim - not that they have to lose that inventory anyway. There is nothing stopping people from using both Linden Lab's grid and the Open Metaverse grids since the viewer is presently compatible with both. Call it dipping a toe in the water and finding the Crocodiles no long bite. In fact, stability is better than it ever was and for sims run on good hardware run easily as well as Second Life and, in many cases, better. Personally, I experience more lag and crashes in SL these days than I do in OSgrid, especially on my own sims which I know are well resource. Yes, the beauty is I control the virtual server and it costs me a fraction of what LL would charge.
Now for a bit of a downer. Sadly, I do actually think the walled garden grids like InWorldz and Avination will not fair that well. We have already seen a sharp decline in Avination who's renters have probably gone to Kitely and OSgrid, both of which have seen the biggest increase of sims. InWorldz declined sharply at the time Avination was growing back at the start of 2011 but, even though they have recovered a lot they still have not grown beyond what they had before Avination took off. On the other hand even though Avination lost many sims and users they may yet pull up a bit but my money is on grids like OSgrid simply because more people want low cost virtual land and some community more than they worry over content. And anyway, OSgrid is bristling with free and low cost quality content anyway. Not just that but the physics in OSgrid, though far from ideal, are actually reasonably good. InWorldz physics are presently none-existant by contrast and Avination still has only ODE the same as OSgrid although there has been talk of them getting Havoc while InWorldz developers have been talking about PhysX for over a year now. In fact, I think the residents of InWorldz were promised it as early as March last year and it still hasn't happened. Perhaps I should predict that for 2012!
Can I predict anything for Aurora sim?
Actually, No, not really since I have been left largely in the dark about what is happening in that team even though I was invited to report news for them. All I can say is that Revolution Smythe continues along with other team members to patch the code but most of the work done is bug fixing and nothing new and exciting to report. Work also continues on Astra viewer and I was told my ideas about the grid search are in the works. I feel sure something is going to happen soon though. Just a feeling.
Anyway, I think I will leave it to the grid builders to predict for themselves where they are going.
News recently blogged by Inara Pey here revealed that a new sister project to Aurora Sim has been initiated to develop a viewer that is being designed to specifically support the platform's features. The developers had not intended for this to become public knowledge presently due to it's highly experimental nature but Inara appears to have noticed it on the Aurora github and the fact it supported Mesh prompted her to try it out in Second Life and report her findings and offer some pictures on her blog.
As I am now involved with Aurora Sim and responsible for the weekly news reports I have posted an official notice to Aurora Sim News which makes it clear that the viewer is not intended for general use, is not a stable release and is available mainly to team developers and testers working with it. Of course, the fact it supports Mesh caused something of a stir on Twitter after Inara Pey blogged about it but all that really happened was Latif Khalif, who develops the Radegast text viewer and recently joined the viewer team, decided to make a quick Windows installer so people could check out Mesh and the Astra viewer dose handle it pretty well. However, this rather over shadows the real purpose of the viewer.
Aurora Sim team were already working with Imprudence/Kokua devs as sister projects in order to get support for Aurora Sim features in the viewer but Aurora has been developing rapidly and there has been some delay in being able to carry out tests. Moreover, given the present uncertainty since the announcement by the Imprudence/Kokua lead, Jacek that she is to retire from the project in September, a group of viewer developers who support the Aurora Sim project decided to launch their own version based on Astra.
There is not a lot I can say at this time about the new viewer but it is based off of the Singularity viewer from Siana Gearz with many features & functions brought in from Imprudence & other viewers. It supports Mesh but not MOAP (Media on a Prim) presently and the developers are working on adding download functions and other features. Revolution Smythe told me a while back that when they finish the generic properties module this will allow new properties to be added to objects very easily. The module will send the information about all new properties it has to the viewer and the viewer will have a new panel in the Build tools window, which will show all of the properties, so that they can be edited easily. "Its annoying to go add each and every property manually into the viewer." he told me, "This would allow an easy way to add things to the viewer." So here we get some idea what the viewer should eventually handle.
There is also some change to the grid manager in the present experimental version of Astra which is the beginning of support for an idea I proposed to Revolution Smythe that would make it easy for users to find grids and allow grid owners to add their grid to a central database which the viewer can call data from. However, there is a great deal of work to be done and things change quickly so I can do no more than speculate on the official launch release. The viewer may even have a new name by then but, in any event, the current viewer releases remain experimental and are not intended to be in general use.
SpotON3d has released a browser plug-in that enables you to enter fully functional Open Sim virtual worlds including the ability to build and create content. This puts it ahead of the competition by a long margin. Up until now it looked like Unity was going to dominate the browser plug-in market for the 3d web but Mesh-enabled Unity3d requires content to be created in advance and downloaded. No one has yet been able to enable the Second Life building functions. I took a look at the SpotON3d browser plug-in after reading about it on Hypergrid Business and I was truly amazed how well it performed. I could even tab to other web pages and back while the viewer continued to function just fine.
Gaga arrives at the Welcome sim at SpotON3d grid
I went to the SpotOn3d web site here, signed up with my avatar name, Gaga Gracious and selected a starter avatar. After downloading the plug-in I was into to the Welcome sim, politely greeted and given some Landmarks. I took off to the freebie store and did some shopping which, I discovered all happens on a web page which comes up inside the viewer window where I could click to get what I wanted. I did a skin change and dressed in a new outfit then set off to explore. The source of the viewer is Hippo which is pretty light but tacked onto the side like SL viewer 2 are a bunch of buttons that open a side window for various purposes including managing account, Land Store, Help and Web Sales. It was not my preferred way because I dislike Viewer 2 but the important thing was this all worked seamlessly and in a browser.
Dancing at Club54
So, I took a trip (I did crash a few times on the SpotON3D grid) and eventually ended up dancing out of boredom at Club54. I never came across another avatar on my travels after leaving the Welcome sim. However, this plug-in could change all that because, for the casual web surfer, this was really quite easy and, unlike Kitely where you still have to download a full viewer separately, with this you don't. The plug-in download is pretty small but bigger than the Kitely plug-in and, once installed, opens up a browser page where you can register a name and chose your avatar. Once you have clicked the verification link sent to your email address then you can simply return to the login page at the browser and the fully functional viewer is activated. You just login and away you go.
Gaga is a hopeless flirt but she dose find time to sail,
build, script and make stuff. This is all
on OSgrid and Gaga demonstrates the browser
can handle building just fine.
I was pretty amazed by it and, though it did seem a little slow, it really was not that bad. I was able to rez a prim to edit, change clothes and appearance fine. Everything basically worked. Finally, I logged out rather than quit the screen and was presented with a SpotON3d splash screen and, at the bottom, all the buttons expected on a Hippo viewer including the grid list. Of course, I entered the addresses for a number of grids including OSgrid and both my Open Sim and Aurora grids. OSgrid was fine and so was my own Open Sim grid. I could login to them and everything functioned just the same as if I had opened a full viewer. I was able to log into my Aurora grid too but, Aurora having advanced features like variable sim sizes met with problems. On a standard region all went well though so, presently, the browser plug-in is good for SpotON3d obviously, and other Open Sim grids but there would need to be a new version for Aurora that handles it. But that is true for most current viewers too because only Imprudence and Astra are designed to enable the features of Aurora. Anyway, I logged into Nova grid which is Aurora-based and everything was just peachy. See some of the pictures. I could even use all the top menus and even change environment settings through all states from sunrise to sunset and night.
Arriving on Aurora-based Nova grid
Environment settings
Sunset
I am not sure but it might not be possible for anyone else to develop this for others grids since SpotON3d appear to be seeking a patent. The developers are using open source code in the form of the Hippo viewer which is actually downloaded to your PC as part of the plug-in package and then the viewer is launched inside a web page using a browser plug-in built on the FireBreath open source toolkit for creating cross-browser plug-ins. I actually know very little about SpotON3d other than they have something going with the Phoenix viewer developers which I noted don't offer any other grid addresses in their grid list besides Second Life and SpotOn3D. What I do know is that they run a large proprietary grid and host other grids too. They have their own vendor system and deliveries can be made to any grid on their servers but not outside their network. They also offer a cloud service for running events where a heavy load is expected for short periods. What I get from SpotON3D is that they appear to be separating their grid services from the rest of the open Metaverse and I don't just mean a single proprietary grid like Avination or InWorldz. I mean they are the provider of multiple grids and asset services within a single all-embracing network which they claim to be a 3d web.
Arriving at Second Life
A row has already broken out about the patenting of the plug-in where it has been pointed out that the technology is not new and was invented by a small start-up company called InDuality backed by IBM around four years ago. Since then others have used plug-ins for browsers to launch other versions including Unity3d-based worlds. The CEO of Kitely Virtual, Ilan Tochner has been particularly vocal on the blog network and has asked for the patent number under which it has been filed but so-far that has not been given. Patents and copyright issues have always been seen as threat to the open Metaverse where any one company could effectively shut off the open source development projects if they patent something that no one is able to find away around. The owners of any such patent could end up monopolizing the 3d web and impose their terms for licensing that would, of course, prevent other companies and none-profit concerns from offering competing alternatives.
Gaga in another avatar enjoys a touch of eastern promise and sailing with pirates on the Barbary Coast
With all that said however, a plug-in for launching a viewer in a web browser wont of itself prevent continued development of Open Sim or Aurora or, in deed, the third party viewers since all that is being shut off without seeking a licence from SpotON3D is one particular means to deliver the virtual experience, albeit potentially a powerful one. Kitely is probably SpotON3d's nearest rival in all this since Kitely has built it's business model on delivering cloud-based Open sim worlds to the public and it stands to reason they might be in the process of coding a similar solution. But, in any event, I have no doubt that the ability to seamlessly experience virtual worlds in a browser has the potential to vastly increase adoption and it would be a bitter blow to all those coders dedicating their time and effort for free to bring about an open Metaverse only to have one company that, in this particular case, has contributed nothing back monopolize an important part of it. Kitely do, at least, contribute code back to Open Sim and are not seeking patents that might do damage. They also actively respond to feedback where SpotON3D, as far as I can see, appear to be entrenched in there own plans rather like Linden Labs of Second Life.
With friends in a UFO, Gaga enjoys Sci-Fi too!
Kitely have shown how virtual worlds can be delivered easily and cheaply to the netizens who enjoy virtual worlds for escapist pursuets in role play, gaming and social interaction. Creativity is at the heart of much of what the open Metaverse is about. SpotON3D has developed yet another useful tool for delivering the virtual experience effectively and easily to this growing market. There is room for everyone to profit by developing technology and services but it would be a rum deal to take what the community has given freely and patent something that would turn it into a monopoly for one company. Deva Canto, inventor of Hypergrid, warned about this over a year ago here and I picked up on it in my article first published on Chapter & Metaverse blog, "Free Metaverse vs Patented Monopoly" then re-printed it here. In this case there is no real threat to Open Sim development in itself but the implications are abundantly clear. There is a clear and serious threat to the development of a free and open Metaverse if a patent like the one sought by SpotON3D succeeds. Others are likely to follow.
When I first read about Aurora sim back towards the end of 2010 when the project was first initiated I was impressed by the long list of improvements over core Opensim that the development team, headed by Revolution Smythe (Rev to his friends), promised. I was not only impressed but more than that, I was quietly intrigued. However, I remained skeptical that such a small team could deliver on their substantial promises considering that core Opensim, on which Aurora is based, has been in development for four years, and while improved greatly in the last two releases, still lacks so much that is taken for granted in SecondLife. Aurora team basically set out to turn Opensim on it's head and reconstruct it from the ground up.
Coming from a background in role play as I do the visual experience and ability to immerse one's self into a virtual world are very important to me. I want a high degree of realism which is still by far and away lacking in SecondLife. I want the kind of quality graphics and responsiveness one gets from games running on PC and Xbox but without the actual game they are selling of course. In short, I want a virtual world that is broader in scope, has better scripting and features that support the themes gamers and role players try to build for themselves and their fellow players. What I came to realize as I got more into the work the Aurora team were doing was that they are game focused and this got my interest up. Revolutions Smythe explained to me that, in the fall of 2009 before he started Aurora, he built a game off of what was the base of Aurora today. The problem was the viewer software couldn't handle it, he said. Fortunately, the Imprudence viewer developers agreed to collaborate and, by changing the licence slightly from that which Opensim uses, Imprudence/Kokua and Aurora Sim became sister projects in October 2010.
Aurora's developers set out with a broad set of ideas, the aim was to complete all that was missing from Opensim and the desire to bring about change that was not lead by developments in SecondLife alone. They replaced the Robust server, and underlining communication framework altogether. They rebuilt the scripting engine adding support for more functions including osFunctions and aaFunctions, and they are adding support for more scripting languages too including C# and VB. But there is so much more. "We are working on some advanced features," Revolution Smythe told me, "When we finish the generic properties module this will allow new properties to be added to objects very easily, such as the Cone of Silence (see video below), which disallows the viewer from seeing things inside the 'object'. The module will send the information about all new properties it has to the viewer and the viewer will have a new panel in the Build tools window, which will show all of the properties, so that they can be edited easily. Its annoying to go add each and every property manually into the viewer." he went on, "This would allow an easy way to add things to the viewer, and in which any version of the viewer supporting the module would be able to see it. I fully intend to let the sim designer change settings so that the viewer sees exactly what they wanted to show like with setting parcels and altitudes for windlight (not LightShare(tm)). We have already added new settings like Region/Estate in Imprudence and we will add the ability to turn off the minimap and avatar name tags so that RPGs and things can run smoother (no cheating). You just have to lock the clients to Imprudence only which is easy with the viewer ban module (included)."
The unpredictable lady, Aurora Borealis dances in the night sky while Gaga day dreams about the shape of virtual worlds to come.
Aurora 1.0 was released early in 2011 and I reported my experience with it here at the time and, accepting it was a pre-alpha release, it really didn't do more than prove the concept and show that the team were in business. The next release a month or so later, Aurora 2.0, was a good general tidy up of the code with multiple bug fixes and something completely new, variable sized regions!
Looking towards a single 65k sq mt region from a massive var-region on an Aurora grid up to 256 times the area.
Opensim has mega-regions which are clusters of sims where one sim is the parent and the rest are attached as child sims thus each region is still consuming server resources separately while with variable sized regions now under development in Aurora sim, or var-regions for short, they simply expand a single region up to 256 times a standard SL region area. That means that just one region is consuming server resources and, where it would take a lot of sims to build a really large space using the mega-region protocol, var-regions can expand to create truly massive spaces that would take hundreds of Opensim regions in a mega-cluster to create the same area. Aurora is already up to four times faster than Opensim. Also, by taking advantage of HTTP Textures, objects like buildings come into view more quickly too so increased performance like this lends considerable support to rendering such vast spaces.
Terraforming my Aurora standalone test sim was fast and smooth, better than I ever experienced in SecondLife or even Opensim. I shaped equal to 3 standard sims on this 1024X1024 var-region in little over 5 minutes. The distant land is left flat to show the sheer scale of it. All that on single server too!
On May 21st Aurora 3.0 was released just seven months into the project which is both a testament to the concept and the sheer determination of the developers to make rapid progress. Windlight features have been enhanced and they surpass Opensims's Lightshare. You can see further than one region away. Groups, Profiles, Abuse reports, Search and so many of the things taken for granted in SecondLife and still absent or dysfunctional in Opensim are now working including more responsive land editing, an integrated backup system, an integrated combat system and true server side bots.
A Role Player's dream come true!
Imagine a city that builds itself with all the effects of traffic and the bustle of city life. Parts might be degrading while new parts are in construction, even city lights and neons flickering on and off, and transport hurtling along while everywhere bots appear and busy themselves as the anonimous populace. The nearest comparison is Sim City but this is something else. It is a living, breathing backdrop to a game world where role players can pursue their storylines in a realistic setting that surpasses anything you could ever expect from SecondLife or Opensim.
Imagine building the dark cityscape of Blade Runner on an Aurora sim where punks rub shoulders with Hara Krishas and replicants are on the lose. Above you see the Gorean city of Ar on Role Play Worlds grid which Gaga visited last year. Now do this build with the city builder module and see it come to life!
A plug-in like the city builder and bot engine will elevate virtual worlds closer to the experience envisaged in prophetic Sci-Fi novels like Neromancer and Snowcrash. One can imagine Blade Runner set in an Aurora sim world or truly astronomical spaces for role play games similar to Eve and other space opera genre. To that end Aurora has support for controlling the viewer, as mentioned above, to over ride settings so the visitor sees the world and interacts with it as the designer intended. The physics engine, ODE, though inferior to Havok, has received special attention and a lot of work giving more support for vehicles. Movement is smoother for avatars too including a smaller, lighter capsule that moves more realistically. Then there is the gravity effects that can be centred and have everything revolve round it. It will be a Star Trek fan's dream come true.
Improving Physics...
Demo of the bots on an Aurora sim
Improving the physics are a key objective of the Aurora devs. It's not enough to be able to walk, run and fly. Vehicles have to function well too and be responsive to the sort of environmental effects associated with the real world and even the distant reaches of space. In SecondLife they have Havok which Linden Labs can afford the licence for. Opensim is limited to ODE without the kind of improvements done by the Aurora devs. It works and that is the best that can be said for it. The Opensim devs take the view it is for those using Opensim to obtain the licenses and upgrade the physics. They just don't see it as important as other things like Mesh and Hypergrid. In deed, Opensim devs, for all their work and dedication, they just don't appear to see Opensim as anything different than SecondLife even though time and again I have heard it said, Opensim is different. No, it's not. It has the same look and feel, too many of the same quirks, and yet is lacking some very basic stuff that is actually useful, and, after four years in the making, Opensim still doesn't function as well as SecondLife. On the other hand Aurora devs do see Aurora as different but they still want to retain and improve what is good and useful. They want to do better than SecondLife. They want to take it beyond the familiar paradigm so that virtual worlds really capture the imagination and lead to deeper immersion. There is no denying that physics remain a stumbling block but the Aurora team do recognize that good physics are an essential component of the core server code and, not to be beaten by the power of corporate money, they are considering more options.
Aurora devs think very little of ODE even though they have done a lot of work to improve it, work that should have been done by Opensim devs had they looked at Opensim as something more than a social/educational platform. It is one of the founding aims of Aurora devs to improve the physics and, recently, I was in on a discussion in which Navidia PhysX was mentioned. From what I gathered, Revolution Smythe has already done work to integrate PhysX but there was still a lot to do and another dev that has been working with PhysX separately was able to point Rev to new code sources and testing facilities. So the outcome of the conversation gave me reason to believe Aurora will move to a new physics engine before too long or they will continue to improve ODE far beyond it's present capability.
Project Wonderland Capabilities from Nicole Yankelovich on Vimeo.
The Cone of Silence implemented on Aurora sim is basically the same as Wonderland seen in the video and I suggested to Rev the cone might also block the viewer from seeing inside the object too which would give privacy on mature sims and serve to increase realism in the virtual world as well. Rev agreed and said he thought it can be done but we will have to wait a bit longer for that one. Perhaps they will change to name to Cone of Privacy too.
Apart from the problem of physics there is also the question of compatibility with Opensim. On the original features list they did state their aim to remain compatible as far as possible but, at present, anything ported to Aurora can't be ported back and, where initially hypergrid teleports did seem possible, they now aren't. The devs decided to disable HG on the premis that until HG stops being incompatible with itself (there are several versions all mutually incompatible with one another) there is no way it can be made to work properly in Aurora. However, the devs have now introduced their own version of HG called Inter World Connector or IWC.
Inter World Connector...
IWC, they say, will be more secure than Hypergrid since security receives particular attention throughout the system. The details of IWC are still not clear at this time but what I gathered was that IWC generates a secure URL on the fly and there are varying levels of trust between grids that are both user and grid operator settable. These levels of trust range from full trust (level 4), where inventory can be sent to other grids with the avatar, down to nothing gets out (level 0). So, where HG has the Outward bounds setting in 7.1 it would appear Aurora has more options although I have no information yet what they might be. One assumes travelling avatars will keep their appearance in the same way that HG works where the skin, shape and clothes, etc, are called from the avatar's home grid so are not downloaded to the visited grid and can't be copied. In Opensim 7.1 grid owners can turn HG on but prevent items acquire on their grid from leaving it while the avatar can still travel with their same name and appearance. I can only imagine IWC will work in a similar fashion but with extra permissions. However, Rev has said it wont be compatible with HG.
Now, I have noticed that many people that are interested in Aurora are concerned about compatibility since they are, for the most part, familiar and probably comfortable with the way Hypergrid works even if they are less than happy with the function of Opensim itself. In deed, Hypergrid is not only an essential component of Opensim for many but without the means to travel via inter-grid teleports it's hard to see how any virtual worlds platform can ever lay claim to be the Apache of the emerging 3D web. Many users are running their own standalone sims attached to grids like OSgrid. Others run grids of their own too. In fact most grids run on core Opensim with a few exceptions like InWorldz and Openlife which have developed their own forked version so are probably too far removed to ever be able to participate in inter-grid travel anyway. If IWC is going to be incompatible with HG then, for many, that will be a serious drawback to adopting Aurora. However, when asked at a recent meeting on the Aurora test grid, the lead developer, Rev did say he could make a bridging module but he didn't say he would. In any event IWC has yet to be fully implemented so currently it remains an open question. But it is an important issue for many, me included.
Okay, so I want to show off the Trekie-style top again. I made it complete with my very own Aurora Sim logo specially for this photo shoot of Gaga on my standalone. Shame about the hair though but I don't have any full perms to download presently. Anyway, I want to make this a tribute to the brilliant coders at Aurora.
Aurora is different in so many ways and yet remains open source. The developers are not content to build a bare nuts & bolts platform. They believe Opensim should have been more complete by now. All the devs have a long history in virtual worlds and started like most of us in SecondLife. They wanted better and brought their coding skills to Opensim in order change things and help build a truly open metaverse. Many have joined the Aurora team but still contribute to Opensim so there is still plenty of collaboration between the projects regardless of a parting of the ways. Most important to me is that they do seem to be listening to the community and, thankfully, they don't have to please any profit-hungry corporate overlords. They have similar views to the community and suffered the same treatment meted out by the blinkered mandarins that run Linden Labs. They tried hard to work with Opensim (and some still do) but the narrow focus thwarted their creative efforts to advance the code. They are visionaries and want more from a virtual world and they believe there is more to gain. My only worry is that while striving for perfection they don't lose sight of the hopes and dreams of the growing community that have already embraced the free Metaverse. If they throw the baby out with the bath water then Aurora risks becoming just another games engine. The socioeconomic model built by SecondLife is still part of the big picture while the virtual experience can and must improve. Connectivity is important too if the free metaverse is to thrive, and that requires a degree of compatibility in key areas especially content and avatar mobility across worlds. I have no doubt though that the Aurora team are on the right track and I believe the project will go far. And do it fast!
UPDATE [Aug 8th 2011] Aurora Sim has moved on considerably since the article above was publish. Currently we are using 4.1 which includes many bug fixes, improvemnets, IWC in grid mode and a Hypergrid Bridge module to connect with Open Sim grids. See more recent articles...
With SecondLife in slow decline and the open metaverse growing it is hard to understand why the third party viewer developers are not working to help users find the grids that already exist. Moreover, we know new grids are in preparation and due to come online soon but unless they get the kind of advertising and exposure that Avination recently got they will be hard pushed to get any notice at all as more and more new grids come on line. What I have argued for in the past is for the third party viewer developers to take a lead and do more with the grid list to help users to find more grids and the operators to gain extra exposure right where it would get the most notice, in the viewer itself.
Linden Lab's own viewer, the much promoted and maligned Viewer 2, has no grid list for obvious reasons. LL are not going to help their competitors but the best known third party viewers do support the grid list. The problem is that, while they do often give support to new server features, they don't actually support the open metaverse by promoting it. Most offer a hand full of pre-configured grids on their list but it is left to the user to add their own favourites and, while it's not that hard to add to the list, many potential recruits to the open metaverse would stumble at this point or just connect to those grids the viewer developers have chosen simply because they don't know of any others. This, of course, leaves out the many grids not favoured and, from what I have been reading, popular viewers like Phoenix are being particularly selective. And worse still, their are viewers being made available by the commercial grids that naturally are biased in favour of their own grid and, perhaps, SecondLife. Once again, competitive interest has lead to a raft of viewers that appear to cold shoulder the rest of the open metaverse.
Now, there can be other reasons to release a viewer like the need to support home grown features on the grid server. However, all viewers are essentially based on Linden Lab's viewer code and will be using the grid list feature first released in the Hippo viewer to support Opensim so it's hard to fathom why viewers like Imprudence, which is targeted at a wider audience, has been so slow to upgrade in the very area that would be most supportive. In fact we now see commercial interests like Kitely bypassing the login process altogether by inviting their customers to download a small application that takes control of their viewer regardless of the make. Kitely makes it easy just to click a button to activate the user's viewer and log them in to their grid as if it were the only one that existed, SecondLife excluded too.
The old MeerKat grid manager
As far back as January 2010 I asked on the Imprudence forum if they might improve the gird list manager and make it possible for owners to add their grid to a database via a web page so that the viewer automatically populates the grid list. I thought that when users open their viewer they could use a search option either on the login screen or in the grid manager to scan the database in various categories. This, I argued, would keep the grid list up to date and save users having to find grid addresses to add themselves and, at the same time, allow owners to get exposure for their grids by adding them to the listing page on the web possibly with a description and even a logo. How great would that be to open your viewer and search for places to visit?
At the time I posted about this to the Imprudence forum I was told by one of the developers they would look into it later when other, more pressing, work had been done. One year on it hasn't happened but I did see a new post on the forum recently which indicated they were working on it. In response to a request to be added to their grid list a new grid operator, AvWorlds, was told their grid details had been added to a Kokua issue tracker list so they will be included on a new web site that was planned. From what I gathered the viewer will call on the database to keep the viewer list up to date just as I had originally suggested but this is still in the works and probably will have to wait until the Kokua (SL viewer 2 compatible) viewer is finally released. Even then, the grid list improvements probably will be way down the list of priorities and the list may still not actually be searchable.
From my point of view, and no doubt that of grid operators, a searchable grid list would be the single most helpful advance in third party viewers when it comes to supporting the open metaverse and yet catering to SecondLife users or individual commercial interests still seems to get priority treatment. Imprudence, and MeerKat from the Metaverse Foundation before it, did give support but in recent times Imprudence devs seem to be struggling with uncertain priorities and a degree of burnout. MeerKat, which discontinued, had unique features some of which found their way into the Imprudence viewer but some features got quietly ignored such as the method of teleporting from SecondLife to Opensim grids and back. Even now people will argue that the inter-grid teleport feature is superceded by hypergrid and so it probably is but hypergrid doesn't actually permit teleports to SecondLife. However, with the advancement of Aurora sim and the release of Astra viewer from the team who originally developed MeerKat we are seeing both teams giving some support to Aurora Sim features such as variable sized regions so perhaps Astra devs will look at the grid list again. A really useful feature of it was the ability to store different names and password to associate with different grids.
This is what I mean by mixed priorities though. All third party developers must follow Linden Labs policy on TPVs and, enevitably, they will give priority to new secondLife features in viewer 2 which, unfortunately, was developed by a team of egocentric clowns known as 80/20 whose idea of what a viewer should be was something akin to a web browser and a cluttered mess of popup menus with thumbnails, icons and all manner of pointless gibberish that would appeal to Facebook users - so they hoped. But read how 80/20 saved SecondLife for yourself. Fortunately, the third party viewer developers seem to have heard the universal disaproval for the UI display, even if Linden Labs ignores it, and are putting in great effort to disentangle the clutter from it. It seems to be getting priority treatment though while features needed by Opensim users are being pushed on the back burner.
I just hope Imprudence and Astra devs will look away from SecondLife for a moment and ask themselves what they can do for Opensim and the free metaverse that would be really helpful. A searchable grid list and stored identities would be something a lot of us would thank them for but I wouldn't hold my breath to see it any time soon.
