Monday, December 8, 2008

TERAVUS OUSLEY FTW!!!!!!!!!11 uno

Teravus Ousley, among many other things, is the Open Dynamics Engine physics engine guru for the OpenSim project. Teravus, who is a proficient and prolific c# programmer, committed work on the evening of Dec 6 (or thereabouts) that fixed physical linkset and linkset collision issues.

This is a matter of some significance, and the importance of this work from a simulation science standpoint cannot be overstated.

Thanks Teravus, and Kudos!


Thursday, December 4, 2008

Building Rexx Viewer on Linux

First off, the source is BIG. Weighing in at 20MB, it's a hefty download.

Having just unzipped it, I am surprised to see no README or INSTALL in the archive root. This is not especially encouraging. What is there are four directories: avatargenerator, maxscriptexport, rexmeshtool, and rexviewer.

With the complete dearth of any sort of build instructions, I am left with no choices but educated guesses - I rather suspect that avatargenerator and maxscriptexporter are utilities, and so I will begin with what might be the one included dependency, rexmeshtool. Here goes Norton!

Ok, so the rexmeshtool directory would appear to contain some sort of microsoft project. My guess is, there is no nant build configuration in the .sln file. Let's go see :D

Well, that didnt go far LOL

No nant .build file - this is essentially a microsoft-only project, and without a bit more work from the it's devs, it simply isnt going anywhere on linux.

I'd love to yell long and loud about the many things I had to do to get this to build - but unfortunately, that isnt going to happen.

So what I will yell long and loud about is this: This project is in NO WAY ready for building on linux. It may be a few short steps from it - it may be light years from it.

But without at least a prebuild file to gen a nant config to bootstrap the build process, it just isnt going to happen.

I will be keeping my eye on this, and on the Rexx guys - but they have a bit more work to do if they truly want to see this run on a linux desktop. Keep an eye on this blog for news relating to progress along these lines.

Cheers for now :)
Hiro P

Physical Vehicles Update

**** Physical Vehicles Update ****

The short story is, we have a few now :D

Thanks to one Owen Oyen of SecondLife and OSGrid :)

One day last month, Owen showed up at my deltazone region to discuss physical vehicles. Owen makes some rather specialized content in SL - he is the one who brought us the pitching of boats under sail, tidal currents based on an analysis of sailing grounds, and a host of other good things related to virtual sailing in SL.

Owen's concern that day we met was why exactly there were no physical vehicle functions implemented yet in opensim. After a brief bit of discussion, we came to the conclusion that the llVehicle class of functions in LSL are really a dramatically simplified approximation of a linear motor with some twists thrown in - literally - for rotating a vehicle about it's axis of travel.

As proof of our sound thinking, Owen produced in collaboration with myself, a very satisfying and functional virtual sailboat :) recently, Owen and Bri Hasp have extended this further, producing an iceboat as well :D

Both are available at or near the south edge of Bri's 'Baya' sim on OSGrid. Unfortunately, Owen has chosen not to make the scripts open - leaving it to me to apply what I learned with him in the production of such scripts.

While I have not yet produced these scritps (I got a job shortly after Owen released his freebie-but-no-copy boat), I will do so in the future - and, I will happily share with you the basics of the implementation.

First, llSetForce and llApplyImpulse are used as motors, depending on the need and circumstance. Owen, I think, finally settled on the use of llSetForce exclusively, but for reasons that were a bit obscure to me. Second, to prevent the vehicle from accumulating energy on an exponential curve, llGetMass is used to calculate a bit of friction, which takes the form of a force applied in a direction opposite to the direction of motion, in the same timer-driven calculation that applies the force which keeps the boat moving.

Finally, a wind source and sail position/forces are calculated and applied. Coupled with some stabilization and correction of movement through the periodic re-inititalization of axes outside that along which movement occurs with constants (usually zero), we have a boat that is readily pushed about the sim on Owen's 'wizard wind', which he has scripted into the boat as well.

Maybe that reads like greek to you - if so, go pick up one of Owen's freebies at Baya and just go have fun with it. Be sure to read the included notecards, because this vehicle has peculiarities and quirks uniquely associated with opensim. If not, I heartily ecourage you to take up this ground-breaking work and see what sorts of vehicles you can produce.

Off to attempt to build Rexx viewer source on linux :D

Cheers for now :D

and see the following post, wherein I describe my rexx-viewer-on-linux-viewer-building experiences.


Friday, November 7, 2008

Scripted Physical Vehicles in OpenSim

First, to establish a bit of context on this report, I am referring to any vehicle that moves under the influence of the physics simulation. This specifically excludes llSetPos and like techniques.

While my experiments did not involve e.g., llSetVehicleType, they are still physical vehicles too - and dont worry, we'll examine the status of llSetVehicleType and friends while we're at it.

The first trouble I encountered in attempting to make a physical vehicle in OpenSim, is that there are no llVehicle* style functions implemented in the script engine. Or are there? a casual glance at the source code (if there is such a thing) would lead one to believe so. A closer investigation, though, will show that while the 'Not Implemented' stubs have been replaced with meaningful code that passes values up to the wooly innards of OpenSim, the afore-mentioned innards don't actually pass it on to the physics engine. But just about everything is in place to do so.

Why hasn't this last step been done? The answer to that is a bit complicated.

The short story is, the wooly innards are about to get rearranged, hopefully into a somewhat less wooly form. And as this is where all that physical script plumbing empties out, it's arrangement has considerable bearing on how llSetVehicle and friends are implemented.

Hold that thought while we take a stress breaker :)

What I *have* had at least some moderate success with is making a flying prim that is a physical vehicle. I know, but it's a start. How can this be done? Why, with llApplyImpulse and llTargetOmega.

Ok, it's an utter kludge LOL but it works.

Why llTargetOmega? it's the only LSL function currently implemented that can physically rotate a prim. llApplyImpulse provides both the lift and the forward thrust. I use the standard vehicle 'control' event to obtain the users inputs via 'llTakeControls', and I use llTargetOmega to stop and start a rotation of the prim when the appropriate keys are pressed to turn left and right.

Why dont I post that script here and now? It's just too ugly. When I spend a bit more time with it, I *might* actually have a reasonably manueverable flying prim - but as it is now, it's just barely proof-of-concept.

Now, stressbreaker is over!

Basically, the work on the OpenSim 'wooly innards' I referred to above is highly complex, represents a huge change to the core codebase, and must be worked over and discussed by at minimum 10 different software engineers. It's going to take a bit of time, I'm afraid. Be patient, it'll be worth the wait.

In the interim, I hope to properly implement llSetRotaionalImpulse and llSetRotationalVelocity, both of which bear enough resemblance to llTargetOmega that I should be able use it as a reference in the source code; if I am successfull, this should give us sufficient tools to implement some very useable vehicles based on llSetForce and llApplyImpulse.

Crude, I know. But better than a foot-race :)

Cheers for now!

Monday, September 8, 2008


Kudos to JHurliman of the libomv project and Teravus Ousley of the OpenSim project for their highly successful efforts this weekend!

JHurliman and friends have recently finished up some fairly major work on libomv, and Teravus worked closely with him on the one hand, and with the opensim/osgrid user community on the other, to bring this fresh new libomv into our OpenSim project.

The technical hurdles were numerous and sometimes quite high, but these fellows were definitely up to the task and it looks as if all the real show-stoppers and toe-stubbers have been addressed; things are cooking along quite nicely. :D

Cheers Guys! Great Job!

Friday, August 22, 2008

Concerning the future of the 'Grid' as a paradigm for VR

The Death of the Grid is Nigh!

Two of the most interesting OpenSim-related blog posts I have seen in recent weeks were penned by none other than Adam Frisby (his post is at, and Justin Clarke-Casey (who's post is at, pose the question, "does the grid continue to have relevance in the future of OpenSim, long-term?"

Whoa! Heresy! Blasephemy! Chaos! teh barbarians are at the gates!
Or are they?

The History and Nature of The Grid

First, lets reflect a bit by what we mean when we say ''Grid'' in the context of VR. VR has it's technical beginnings in the code of MMORPGs. If you don't know what an MMORPG is, you may wish to crawl back under your rock ;) SRSLY though, OpenSim inherits many of it's metaphors through a direct line of inheritance going all the way back to the text-only muds of yesteryear, and even more remotely to games like Infocom's ''Zork''.

For our purposes, a ''Grid'' is a means for grafting little slices of geographical metaphor together to present them as a contiguous aggregation of ''Land'', or ''RealEstate'', with all of the connotations and overtones that come along with that as baggage (or value, depending on one's perspective). That perspective largely stems from one's interests in implementing a virtual space.

This is a fairly traditional paradigm, and is largely taken for granted even in the literature which initially inspired the developement of current VR models; to such a degree that a potential debate begins to take form centered on the notion that grid vs. not-grid is an assertion in need of evaluation.

For myself and others, this is a question which is largely evocative of swift reaction, we having invested a great deal of time and energy in developing, supporting, and using grid applications of OpenSim and supporting software and practices. Upon reflection, however, I have arrived at the position that the question is largely moot. And I'll gladly tell you why in subsequent paragraphs, as we leave off the discussion of canonical grid architecture and examine that of OpenSim.

OpenSim Architecture

OpenSim is designed to address not just one but a potentially infinite set of use cases. Great care has been taken both in the forward-looking design and the backward-glancing refactorings to implement feature sets as modules. To say that OpenSim is very configurable is a gross understatement. OpenSim is literally three different complete software systems; the simplest of which is the 'Standalone' codebase - it is really the other two combined, but 'lite'. The remaining two software systems are the 'Grid Services', and the 'Region Server'.

As an aside, our first clue to trouble appears here, that is, even if there is no duplication in code, there is certainly duplication of purpose - the two variations (grid and standalone) are distinct in the interest of addressing concerns of scale. While scale is definitely a matter of concern if one wishes to embrace a large body of users, it's pretty well acepted that two distinct but similar applications is not the ideal way to address this concern.

The reason grid mode scales better than standalone is that the grid provides a centralized backstore for assets, user authentication, and user-ownership of assets. Or so the story goes. The truth is, these services do not, of a necesity, need to be delivered from a single source; in fact, there is every reason to believe that spreading them around a bit will - you guessed it - scale better.

To elaborate a little further on standalone (hopefully without putting too fine a point on it), the standalone codebase implements a specific use-case: the 'Walled Garden'. This walled garden, as the name suggests, is a slice of geographical metaphor that is unaffiliated in any other explicitly virtual way with any other seperate but similar slice of geographical metaphor.

A Potential Pheonix in the Making

Note that the primary reason for the existence of these two distinct but similar applications is more profoundly technical than social. Some might beg to differ, but the ability to limit incidental visits by anonymous or other uninvited users is more by consequence than by design. This is evidenced in the relative ease with which standalone has been patched in early OGP/InterOp testing.

What may rise from the ashes, or, where I'm actually headed with all this

Essentially, what needs to happen (and therefore, what will most likely happen) is that the current state of affairs will slowly be displaced by one in which a much more thoroughly distributed model will be preferred. Think standalone mode on steroids: configurable to use local or remote backstores; to participate (or not) in some third party authentication and/or asset providers; to appear (or not to appear) on agglomerative geographical representations, with concommitant metaphorical participation in-situ (see into/from 'adjacent' regions, participate in object/script crossings, physics/water table/terrain cooperations, etc.)

This architecture pushes relevant responsibilities for authentication, hosting, and permissions/ownership back onto the region operator, where they really should be. Does this really eliminate the 'grid'? potentially, perhaps, in any oraganizational sense. Grid operators will likely become the first operators of the afore-mentioned service level organizations. Indeed, they already are ;)

The Nature of the (new) Beast

- This 'idealized' OpenSim will be simpler to support. There will only be one of them, as we will no longer make a distinction between gridmode and standalone.
- It will be more homogenous, as all instances will be potentially reachable from all others, even where no metaphorical geographical relationship exists
- It will be more socially satisfying as all spatial agglomeration will be voluntary, deliberate, and by design.

Long Live the Grid, all hail the Grid

To bring this all to a comfortable close, I think the death of the grid is a bit overstated. Sure, it'll pass (especially in it's current incarnation), but it will represent the progress of many sure steps forward, and produce a far more usefull operational paradigm; one in which politics will become less important in the developement of the software, promoting easier support, smoother developement and testing, and a better, more satisfying VR experience for Operator and User alike.

So for my part, I say bring it on!

James G. Stallings II
aka daTwitch/Hiro Protagonist/Lazarus Longstaff

Friday, August 15, 2008


I haz created a more or less traditional book in OpenSim using LSL scripting.

It runs in a single cube prim. For your own copy of my first effort, paste this into a script and drop it in a cube prim:

vector opensize=<0.064,0.5,0.35>;
vector closedsize=<0.04,0.265,0.35>;
string book="closed"; // b6258725-335d-4e47-9676-39361010f5cc 34270b40-68a2-4341-aeb3-10e3b2f45214
list pageUUIDs=["ce3f600b-0028-40f0-a77b-10426b8ff975","07a9e95b-ef73-47ef-80f1-b6c168529142","e39d7c07-aa29-4776-8083-b6a055427995","589e5863-b3a5-4441-a2c8-fd960cf1997e","32ae5dd8-582c-410a-afc9-fef75fc4a6b4","4152fba5-02ef-4f1a-a7d2-9d4ebc4a8be1","f4427be8-651d-47d5-ab06-5597443b92d2","aeb9162e-a2e3-4942-ba58-1ed0cbb250ab","98888e86-b1d4-412f-b275-a98d04038dad","1840ea8f-ba8e-47b0-8182-cf3abc56bb8d","a00dd649-f617-4f39-a4e0-6b110199d439","a063425d-b31a-4309-8d4d-1886f1cd8530","afb3f53e-9ca6-425e-874a-f9bba796d7cf","09c65ace-d015-4a6c-9467-421f9f728f42","bb976ea4-f423-4706-b7f8-33fd4d856029","8b7f7d9e-57f2-4c17-a010-30972be1982c","707e44d8-e432-47f7-9535-b42e032f485f","f860fb9b-d1fb-486c-948f-7131bade9662","b065b77a-6e6c-4cfd-9079-50df43455595","006e3243-9817-49e2-b975-1c1c8e5b5222","2e86574c-21a1-46ef-a6cb-e076c6e94d27","f4df80e8-2e47-4287-abc8-9422d299995c","5470da29-83f5-4cab-8ee1-c0cd80102450","19ef268b-03e5-4678-aff8-d7db5d3b1ccf","fef6f8ed-554f-44e5-88ec-93338196981b","b61d1655-044f-4332-aa1b-558b042f2ea1","dbdeb221-5eb7-461e-acd1-a57b0e733c4f", "bf5b0c57-2d57-4bb4-82f3-a22ce8e9a1ac", "ccee6b1d-7201-43c1-bf31-e664a6f4c808",
"b6258725-335d-4e47-9676-39361010f5cc", // right here gets a repeat

integer pagesets=72;

integer pageset=0;

string ID=llList2String(pageUUIDs,pageset);
key textureID=(key) ID;
llSetPrimitiveParams([PRIM_TEXTURE, 5, textureID, <0.5,1,0>, <0.25,0,0>, 0.0]);
llSetPrimitiveParams([PRIM_TEXTURE, 6, textureID, <0.5,1,0>, <-0.25,0,0>, 0.0]);
// llSetTexture(textureID,5);
// llSetTexture(textureID,6);

llSetPrimitiveParams([PRIM_TYPE, PRIM_TYPE_BOX, 0, <0,1,0>, 0.0, <0,0,0>, <1,1,1>, <0,0,0>]);
llSetPrimitiveParams([PRIM_SIZE, closedsize]);


touch_start(integer num_times)
llSetPrimitiveParams([PRIM_TYPE, PRIM_TYPE_BOX, 0, <0,0.725,0>, 0.0, <0,0,0>, <1,1,1>, <0,0,0>]);
llSetPrimitiveParams([PRIM_SIZE, opensize]);
} else if(book=="open")
pageset=pageset + 1;
} else {

See if you can figure out why it repeats the 31st pageset, I cant - it may be a bug in OpenSim.

Oh, and a bit of an easter egg for those of you running linux that read this far:
get a debian pkg of qavimator binaries here:


In the Interim

How should I put this - I'm at no particular loss for words, but have little to say that's topical.

Sounds like a good application for 'Stream of Consciousness' writing:

Elation Shifting like Gold Sand with sun SHINING only dimly lit light lighting lightening but not too much wind of water contention strife simplistic overtones of disarmonious old music getting stale for friends new and old politics makes stranger bedfellows and still it's a chinese world after all

Monday, July 7, 2008

Concerning Open Content, DRM, and Second Inventory


This post is to address some concerns expressed by one of my colleagues with regard to my mention of Second Inventory in my last post.

Having come to grips with my reactionism in response to what felt a little bit like an attempt at censorship, I came also to a realization that to a degree he was right - I had made certain basic assumptions that not everyone is comfortable with.

As far as I am aware, Second Inventory does not allow it's users to copy content that doesn't belong to them; let me be perfectly clear for the record when I say I certainly hope it does not. I don't use it myself, being strictly a linux person.

I have every reason to believe that the tool was being used responsibly by the parties mentioned in this context in my prior post. As far as I know at the time of this writing, there are two viable ways for moving 3d content inter-grid - the copybot and Second Inventory. Both are tools; as such, they are meritorious only according to their usefulness. Like the hammer and the saw, they can be used in both constructive and destructive ways. The good or bad is in the hand that guides the tool.

I did not mean to either personally nor on behalf of anyone at,, or endorse content theft, content theft techniques, or any other sort of encroachment on anyone's rights to the legal use of their property. Nor do I anticipate changing my stance on that in the future.

That being said, however, I will continue to promote increased knowledge of tools and technologies useful for the ethical use and reuse of content in the most effective and efficient methods possible, and decline to take responsibility for the actions of the disaffected and despised who surely earn their infamy through their deeds.

With Kindest Regards,
James G. Stallings II
aka Lazarus Longstaff/Hiro Protagonist/daTwitch

Dateline 2008-06-07: The Week in Review

Qee Nishi Rocks the House at Shengri La

Greetings and Salutations Avies!

It's been a busy week, chock-full of fun and good times for all! so lets get the ball a-rollin' as we take a walk forward through time from this time last week...

Zaius Plaza Build
Zaius Plaza, currently under construction (of course, isnt everything?) is rapidly becoming freebie heaven for OSGrid. Contributions are coming in en-masse from Shenlei Winkler and Nebadon Izumi, both importing content via SecondInventory they have created over the years on the Linden Labs grid. Both have been content creators on the SL platform for three years or so. Shenlei's mall will contain her offerings, and Nebadon has set up the bottom three floors of OKC Tower as a freebie-zone and Art Gallery. Both creators are donating the body of their imported work to OSGrid users, if you seem 'em in world be sure to say thanks!

Additionally, OKC Tower is also a community freebie-zone, so if you have content you'd like to share, come on by Zaius Plaza and drop it off at OKC Tower.

Bade Plaza Build
Bade Plaza is a build I'm conducting pretty much solo, so it goes in fits and starts as I move around between evangelizing and supporting OpenSim and OSGrid, creating content, and otherwise taking care of business.

Having gotten a little tired of working on all these techie-looking projects in the midst of my more or less pastoral digs at Hoffman Sim, I decided to make use of a work plat I had established some time ago at 500m altitude; going to and from rapidly became a pain.

Bade Plaza Under Construction

So, on a whim, I decided to make a decent teleportation system (I know, this tale is supposed to be about Bade Plaza - bear with me a sec).

This actually came together quite nicely; I'd recently tested bulk upload with a ton of rich textures from Filter Forge, one of my favorite sources of free textures. Great software too from the look of it - but I dont do windows.

The scripting evolved quite naturally from a combination of a waterfall script, an alpha-adjusted tie-die texture, a posing stand script, and a couple of calls too llRezObject and llSetPos. Look for the system in a package (real soon now) over at OKC Tower in Zaius Plaza, or pop over to Bade Plaza and take a copy of one of the operating units.

Bade Plaza Teleportation Device

And with that feeble seguey I'll get back to Bade Plaza. Bade is an unusual build; it's a small city (tiny really) consisting of four approximately identical apartment buildings in each of the four corners; the region is layed out approximately symmetrically, and each of the buildings are joined to the others by walkways couched in gardens and aqueducts. The region is surrounded and bisected on north and south axis by waterways, with a central pool.

The apartment buildings provide ready dwelling spaces for region squatters/homesteaders. They also provide the-ground level support for a massive 200m+ skyscraper called Blade Tower that dominates the entire rest of the city. Tower space is available only by special arrangement (contact yours truly with your ideas if interested).

Oh, and you can see the teleporters in action there too.

AdamZaius comes to OSGrid and Brings His Regions With Him
Boy, Howdy!

Apparently, AdamZ doesn't believe in doing things small. So far, he's dropped a few tens of servers on the grid, practically doubling the number of regions attached. What's more, he isn't finished yet. By my reckoning, from the hints he's dropped on #osgrid at FreeNode IRC, he will own fully 4/5ths of OSGrid's regions by the time he's finished. What a freakin' load test for the UGMAI! Go AdamZ!

The InterOp Love Fest Continues
After week-before-last's Linden Lab powows, Which Linden's Office Hours were a little anticlimactic and highly technical. Only the AWG hardcores were in attendance; I think John Q. and Mrs. Public are growing weary of detail, as I note their attendance is beginning to slowly fall off. Probably such things go in bursts anyway.

Which's Office Hours were more a brainstorming session than not, though I think (s)he may have preferred it to be more of a presentation. I don't think (s)he got finished, and was perhaps a bit worn out from the difficulties of trying to communicate clearly and effectively one's ideas in cleartext. Of course, anyone who cares about the issue has their own ideas about how all this should work; not all of them are cluefull. 'Nuff Said. Here's a link to a transcript of Which's Office Hours.

And Finally....Party at Shenlei's Shengri La!
Ok, lets get a couple of things out of the way right off the bat, and then we'll get to the pix:
1. This was hands-down the best party I have ever attended in Second Life. No question about it. The Host (Rez Tone) and Hostess (Shenlei Winkler), were phenomenal, the people in attendance were notable and photogenic, as was the venue; and then there was ...
2. QEE NISHI. Qee spins the techno. Qee spins the rave. Qee spins the trance. Qee spins it all in a contiuous flowing stream of cosmic notification that she is rockin' the house. I just can't say enough about Qee. Her av is the picture of her sound. A total package - absolutely incredible.

And the last bit of party news before we get into the photojournalism:
Shenlei and Rez are planning to collaborate with Qee on OSGrid's first major event. Heavy iron is being donated to carry the load of 40+ highly decorated and animated avatars and scripts. It's gonna be a party to remember! Sometime in mid to late August. Keep your eyes glued to this blog for more details on this event as it unfolds.

So, here's the pics from Shenlei's Party this past weekend at Shengri La:

This is Shengri La on the night of the party:

Here's our beloved host and hostess in proper form. Rez veiled in particles on the left, and Shenlei on the right in the articulated wings.

Nebadon Izumi (one of my OSGrid Ops colleagues) and his good friend Coke Supply Get Down on the dancefloor:

The inimitable Qee Nishi!

Qee show's us what it is that's sexy about technology:

Qee against a backdrop of Rez's molecular construct. By any chance would that be a caffiene molecule Rez?

That's it for now, a fun and busy week!

Celebrate content! We sure did!

aka Laz/Hiro/daTwitch

Tuesday, July 1, 2008

Dateline 2008-Jul-01: A Day in the history of VR

Greetings from the interwebs and farflung VR spaces!

Tonight I'll be relating a set of VR history-making events (at least *I* think they are history-making) that occurred over the course of a very busy day in cyberspace.

First, was the Architecture Working Group meeting on Linden Lab's grid, led by Zha Ewry of IBM. This was one of the best meetings I have yet attended on the Linden grid - the participants were passionate, reasonably polite, very focused, and very well spoken. It was almost completely devoid of backchatter and heckling, and no one showed up to fling poo prims. +1!

Next was OpenSim Office Hours, held each Tuesday at 12:00 noon PST. Most all of the OpenSim Principals were in attendance, in addition to several linden labs folk who made their first appearance on OSGrid. And, we broke our previous record for concurrent logins to a single region server! Read on to find out just how many folk attended.

Finally, winding up the afternoon, was Zero Linden's Office Hours, where he made a brave attempt to present his outline for a permissions system that would provide the necesary mechanisms for implementing DRM obligations in an open grid services environment. More on this later though. Let's get into the thick of it!

AWG Meeting re: Domain Membership and Trust, with Zha Ewry of IBM

This was an awesome meeting! The best LL-sponsored meeting I've yet attended. I dont know how much design work was actually done, but a lot of understanding was passed around. Link to the transcript here. The thing that made this meeting great was there was no real heckling; everyone there was focused on the issue before us; no one was really ill informed, or at least, no one was unwilling to expand their understanding. It was a fantastic meeting with an equally fantastic group. Great Job Zha, Saijani and AWG!

Linden Labs vs. OpenSim Office Hours on OSGrid, with a broken record!

That introductory blurb doesn't do proper justice to what happened - it makes it sound like a musical experiment gone way south! Here's what *actually* happened:

We set up for office hours, as per usual - except perhaps there was more of us in attendance than usual. I dunno, maybe everyone was just free at that time for once. In any case, we piled on in and had just begun to discuss our agenda, when the first Linden Labs representative showed up. Unless I am mistaken, this was one Hamilton Linden. Next came Infinity Linden, followed at some interval by Whump Linden (he came in a bit late, and unfortunately we did not get to chat with him at much length). Zha Ewry also made it by, to round out the party.

With so many special guests at Office Hours, we were behooved to some flexibility with respect to our agenda, finishing up quickly with our more routine discussion, and turning the floor over to our guests. Hamilton did most of the talking, desiring to confirm that we had some marginal level of interest in assisting with their interop testing; I wonder if he wasn't a little overtaken by the strength of our response. All expressed interest in assisting with testing in whatever way possible, and several of us made one or more of our regions available for R&D. Here's a phat link to a transcript, and an even tastier link to Nebadon Izumi's video of the occasion. High point of the meeting: when Hamilton Linden looks around at Wright Plaza and asks "Did you build this build in OpenSim?" Serious Chucklehead, that Hamilton!

What? I forgot something? a broken record? Oh, that was just the soundtrack for Zero's presentation on Permissions...just kidding!

The record that was broken, for the number of consecutive logins on an OSGrid region - was at Wright Plaza - we hit 25 main agents (the prior record was 22), with no significant lag or other issues, and the region never did crash, though JustinCC did mention his inventory was flailsome - but I think he was probably the only one diggin' in inventory, so hard to say if it was anomolous or not.

Zero Linden's Office Hours Permissions Presentation

Big, heaving, tired sigh....Poor Zero - I feel for him. Seems the asshats *always* show up and disrupt his meetings. I often wonder why he even bothers. He's clearly got the patience of Job.

Today Zero attempted to deliver a presentation on the plans for a permissions system, one that would ideally complement the things Zha and AWG are doing with Region Domains and Trust Protocols. Unfortunately, most of the folk in attendance at this meeting had neither bothered to attend Zha's meeting nor read the transcript beforehand. It was a chaosfest, and that was before the first griefers showed up.

Hardly anyone there had sufficient understanding to ask intelligent questions, and when Zero was able to address their talking points directly, they often didnt understand one or more of answer and implications. Many slipped into "this is fundamentally broken" trains of thought, or "just open it up like the web" mode. Most seemed to think we were discussing user interactions with an interface when in fact, in most cases, we were discussing transparent communications exchanges between client and server components.

Then there were the griefers. I'm not sure what it was they were slinging around; but eventually, the prims were cleaned up, and the offending parties removed.

Then came the second wave.

And the third. I'm not certain whether or no it was a PN attack, but it was certainly crafted to look like one. Particle rezzer spewing 'srs nigras' e.g., a black man with a large afro in a grey dress suit - basically, the PN logo. Personally, I think it was likely a 'false flag' attack, the PN are -much- more repugnant than this.

Zero handled all this with not inconsiderable grace, and extreme patience to boot - as numerous god-mode Lindens were zipping about cleaning up and banishing identified offenders to the seventh level of hell.

Hang in there Zero, some of us actually do benefit from your presentations!

Oh, and here's a link to a transcript from Zero's meeting.

All in all, it was a great and momentous day in VR - new ground was broken, bridges were built and plans for the future were laid. Thats my kind of day - where we can all get in the pit together and show a little love.

Cheers from Hoffman on OSGrid!
Hiro Protagonist aka Lazarus Longstaff aka daTwitch aka James G. Stallings II

Monday, June 30, 2008

News You can USE


Recently, Nebadon Izumi (another of the OSGrid admins) and I conducted a very interesting and productive test on OSGrid. We employed a light-weight patch against mono trunk that enables the use of large heaps on linux, and compared it against a similarly loaded windows region running .net.

This post is a general description of the test - a more detailed analysis will follow shortly.

The test was basically comprised of DMiles' (#opensim-dev, FreeNode IRC) patch (which can be found at, and related instructions.

Our initial test involved enabling the Boehm garbage collector as per the instructions and testing against an empty sim which we then populated with multiple copies of the same scripted prim.

This reference prim contained a script which loads a radar map from on 15 minute intervals, then paints it on the prim face with the proper osFunction().

Using the patched mono I was able to load no less than 1024 of these prims into the region.

The region ran for some 12 hours without significant issue.

Memory consumption for the region was 1.3 GB

While this rendered the region unusable from a client running on the same box as the test region (my normal approach to things), it was still quite useable from a different box on the same lan, as were the other four regions running on the same hardware.

Speaking of hardware, my configuration is as follows. I'll include the software while I'm at it:

Pentium D dual core running at 3.0GHz
2 GB ram
SATA HD subsys
10/100 switched network
Tier II DSL (256 DN/96 UP)

Debian Linux unstable branch (sidux)
Mono svn (a recent revision - sorry, the best I got right now)
OpenSim r5189 (at the time of this test)
MySQL 5.2

Other sifgnificant facts about my installation:
It's my desktop machine, so typically runs an mp3 player spinning from the local disk;; and the client.

My OpenSim configuration is OSGrid attached with a MySQL backstore for the regions' assets and grid inventory.

As for more general workaday useage, it goes something like this:

My home region on OSGrid, Hoffman, has around 1000 prims, and around 100 scripts (animated fires mostly). With a single avatar on the region, it typically consumed around 120-160 MB. Since the application of this patch to mono, the footprint has decreased by some 40 MB on average. This is consistent with the reduction in footprint when the sim is unoccupied, when I have seen memory usage drop to a low point of 35MB - typical useage under these conditions pre-patch was around 70MB.

Nebadon's findings on windows/.net reflected a similar memory usage profile. His installation is a Core2duo at 2.2GHz with 2GB ram running Windows 2003 Server and .net; everything else pretty similar except he has a much faster internet connection (the winserver in question is hosted at He ran 500 or so of the same test scripts; his memory consumption bounced around a lot between 500MB and 1GB depending on whether or not he left his avatar idling. Mostly closer to 500MB.

All in all, at least anecdotally, every indication is that this is a Good Thing to do for your regions.

*** note: at the time of this writing, the patch is known to be broken due to a hand edit gone bad. A few plus signs are missing somewhere. It was trivial, however, to apply the code change with a manual insertion of the code to the mono source. Will attempt to get DMiles to polish this a bit Real Soon Now.

Hope this gets you more bang for your OpenSim buck!