MetaHarpers

Over the weekend I experimented with yet another method for avoiding script lag from rezzed, scripted objects.

Does this effect you?

You may not realize your region suffers from rezzed object script lag. This is a large topic but for a quick test, do this: Wait till your region has only a few avatars on it. Open up the performance monitor in your viewer with control-shift-1. Scroll down about 2/3rd of the window, and find the line that says ‘Scripts Run”.  If this number is lower than 80%, Your region likely suffers unduly from rezzed object script lag.

Ok, so what’s the news?

I’ve been experimenting injecting a script into rezzed objects. The idea is that this script that can stop and stop the other scripts inside that object, on demand.

This would allow you to do things like stop all non-essential scripts in the sim during the hours it was hosting a performance.

I tested pausing and resuming scripts within objects managed this way, as well as setting up exclusion zones, so you could say “I want to pause all scripted objects more than 100m away from my theater”. This would allow all your performance objects to continue to run.

The starting and stopping mechanism is more like a “pause” and is not destructive like a script reset is. When the managed objects are resumed, they pick up processing where they left off.

One drawback to this method is that it can only be used for objects where the owner has MOD rights, and is only efficient for managing objects with 2 or more scripts inside them. Even so, early research is that it could make a significant impact on script lag from rezzed objects.

What might be next?

At this point I’m like to hear from you if you’re a region owner who might be interested in using this method to control script lag. If there’s interest there is some advanced potential, for example building a system that pauses some kinds of low-priority scripted objects when their owners are away and unable to notice.

Until Soon, -Arrehn

I’d like for the MST 6.0 release to focus on ease of use, and reliability in high-lag environments. I’m open to rewriting pieces, if it makes them easier or simpler to explain. For example getting rid of the concept of boxing/unboxing of the performance engine.

I’ve started to capture the various ideas that may inspire the MST 6.0 release below. I’d love to hear your ideas and your feedback!

Mover Improvements

• Install the ‘never drop an invite’ system, brought over from the Choreo HUD project.
  
• Teach movers to register avatars who sit on them to the MST Choreo HUD, not just performance engine.
• Allow movers to be more remote-controlled (remotely set save status, and remotely set home status. Both of these may be handy for HUD control)
  
• Stop giving out the sync tool by default. I’m thinking this can be an option, which would allow people to choose whether to use synctools.
  

MST Performance Engine Improvements

• Adopt a new, improved-efficiency event scheduler from MST Choreo HUD project. Should save script time.
  
• Allow a method of disabling internal animation control, for  if you want to use the choreo hud for animations, instead of the engine.
  
• Prettier texture and size for the engine
• deprecate the box/unbox functionality
• warn when multiple valid centerpoints are present
• warn when no valid centerpoints are preset (report if centerpoints are present but we’re not allowed)
  
• Fix the precache command so that it can be called by chat command.
  
• Make the menu auto-reopen after making a choice by default
  

Centerpoint Improvements

• Auto stop / start the communication scripts that can be used to project voice on stage, that no everyone uses, to save script time when they aren’t in use.
• Automatically set the centerpoint name to the venue name, and vice versa
  

General Suite Improvements

• Include the new, mesh MST syncTool in the suite package
• Add a subfolder for venue-owner gear
• Use a new mesh box for the tools

Documentation Improvements

• Document all the possible eventlist commands, similar to the MST choreo HUD.
• See if anyone wants to make quick vid tutorials
  

Greetings Harpers! Here’s a quick catch up of what we’ve been up to and what’s coming soon.

Quick news in 2018…

The last many months have been quiet for us in Second Life as some of us have been experimenting with some personal projects. I’ve been experimenting with new seating animation systems that go beyond the capabilities of the usual things you see out there, along with tricking out house rafts and developing lag detectors.  Vicki’s been exploring odd artistic behaviors when you move your avatars into negative altitudes as well as building some unusual marine vehicles, tributes to Roger Zelazny, and Seamonsters. Candace has been building scavenger hunts and quirky clothes.  Fifi and Harter have some of the newest mini-exhibits in our group land in the Miramare region. I apologize if I’ve left anyone out, let me know so I can make updates!

On the performance side, our creative partners and fellow harpers at Idle Rogue continue to experiment with different theater formats and show styles and most recently launched a high-production sci-fi build with dance shows and space particle shows for the Burn2 festival.  Our friends at Beach Shack Productions have a built a few more shows with their signature incorporation of voice translation and comedy. Caledonia Skytower and friends produced a very polished, well acted machinema short where some of our tools made key contributions. Check it out and be inspired! Finally if you are yearning for more high fantasy steampunk inspiration, our friends at Flygearz launched a new regular social shindig of music and dance a few Saturdays each month on a luxurious steampunk airship in the skies of the famous Calais Galadon park regions.

So what’s coming next?

I’ve had a few bursts of activity lately and am excited to share what’s coming and also give the opportunity for anyone interested to reach out and join in if they so desire.

Interactive Theater Experience 2.0!

As many of you know we’ve been hosting the Interactive Stage Gallery in the Idle Rogue region for years. This is, to my knowledge, the only gallery that takes theater shows, and then allows them to be replayed with your avatar acting in the starring roles.

In 2019 we’re planning to redesign this concept to be even bigger and better. It will provide a wrap-around, immersive experience for audiences and no longer be constrained to a traditional stage and seats. It will also use new technology designed to make even the audience experience an interactive one. Finally we would envision this theater to be used for multiple purposes, a mix of both live performances and on-demand. If this is the kind of thing that interests you and you want to help in some way, speak up. We’re intending this to be a community project.

MetaHarper Choreography HUD!

I’m really excited about this one. This project is fairly advanced and will get its own post in the very near future. It aims to to be the best group freestyle dance HUD in SL and the best tool out there for designing animation and command sequences in SL.  Have you wanted a slick graphical interface to the MetaHarper Show Tools? This will be that. You’ll start to see documentation pop up for it on the website.

New 6.0 MetaHarper Show Tools Release

The MST Suite is due for some technology upgrades that have been proven out in our other projects. The “never lose an invite” animation system will be added to the movers, and a high-performance event list processor from the MST Choreo HUD will make it into the performance engines. I’m also looking at using some new script capabilities to make access setup easier, and possibly adding the option to change the channel for direct chat commands. If you are an MST suite user and have suggestions for the next release, this is your chance! I want to hear from you.

Lag Management Tools!

As the Second Life platform has evolved and live performances become more sophisticated, squeezing every available drop of performance has become increasingly important.  My goal for this project is to make it easy to understand the performance of your region, and more importantly make it possible to plan ahead of time to ensure a good live show experience.

I did some work on this over the past summer and hit a few interesting challenges. It turns out that the public tools that SL provides you for detecting script and region performance become unreliable when that region suffers performance degradation. So the worse off a region is, the less it is able to explain to you why it is under-performing. As bad as this situation sounds,  I’ve prototyped a couple possible solutions and plan to try them out in the months ahead. Stay turned for more info.

That’s all for now, happy exploring

-Arrehn

If you are trying to put on a performance in Second Life that depends on well-timed scripting, you know it can sometimes be a challenge. Dance Shows, Theater, Sports Events, Combat Events, Amusement Parks, Particle Shows, many kinds of events fall into this category. This article will talk about some of the unique issues and also outline some of the tools I have created to handle them.

Enemy #1 : Script Lag

Unfortunately for those of us who depend on well-timed scripted logic, Regions give scripts the lowest possible priority. Regions take care of everything they need to do first: simulate physics, track avatars, send content to viewers, etc.. and with whatever time is left they process scripts that ask for attention. If the region doesn’t have enough time  left to handle all the scripts that want attention, all scripts get delayed and we call this script lag.

In the past, script lag could actually cause avatars to crash. Luckily this is no longer the case, but it can still ruin your event, if that event depends on scripts running in a timely fashion!

One saving grace is that regions start with a large amount of unused capacity. Assuming you have the authority to manage objects on a region, with proper planning you can insure that you have enough resources to run a successful script-intensive event, even after many guests with their own scripted attachments arrive.

Strategy #1, Groom your Region

The first and best strategy should be to simply avoid using up the pool of sim script time in the first place by removing costly scripted objects that are not needed. This can be easier said than done, because traditionally it is difficult to determine which objects are the most costly, where they are in the region, and exactly what the costs mean.

To help solve this issue I’ve created a tool “MetaHarpers Region Script Performance Report” which inventories all scripted objects on a sim, measures their impact, and gives you a handy dashboard.

The dashboard tells you two very important pieces of information: Firstly “How much of my script time am I using right now?” and secondly “Based on this usage, how many typical visitors can the region host without script lag?”. These two pieces of information will give you the information you need to know if your next big event is likely to experience script lag or not.

If the data is telling you that you are likely to experience script lag, you have a handy sorted list showing you the most expensive objects, where they are, who owns them, and how many more visitors you could host if you removed them. For an example, you might see that a scripted object “My Texture Organizer” uses the equivalent of 2 typical visitor’s worth of processing time.

Strategy #2 Encourage your guests to be Low Impact

Since visitors often (always?) bring their own scripted HUDs and attachments with them, which consume your region’s resources, it makes sense that if you can reduce visitor’s usage, you’ll have more left over on the region for your own events.

The traditional difficulty with this strategy is that oftentimes it is not known or clearly explained what is “good” and what is “bad” script usage. Typical tools such as “get script info” can be cumbersome to use when you’re wearing a dozen scripted attachments. I frequently see arbitrary limits enforced at event venues such as “You must have less than 100 scripts” or “You must use less than 1MB of script memory” but these targets usually do not tie directly to script time actually consumed. They might have an impact, or they might not. This indirect relation to performance can lead to frustration, for example when a knowledgable user knows that their attachments are low-impact but a script counter insists they are over a limit, or when a visitor knows they are using too many resources but doesn’t know which attachment or HUD is best to remove.

To help provide a more accurate idea of “good” and “bad” I’ve created a few different tools that can determine whether a visitor is using more than “their share” of a region’s script resources. The visitor’s share is determined by a model based on their actual consumed script time, as well as how many other avatars in the region the visitor has to share resources with.

Not only can these tools provide a quick “good/bad” reading but they can also give you a sorted listing of an avatar’s attachments, showing how expensive each one is, along with HUDs as a whole.

One version of these tools can show a “top 5 most costly avatars” listing, along with the share of resources consumed by each one. For example, you might see that “Eve” is using “300%” of her share of script time. Another version of these tools can be quickly worn by avatars and gives simple “good/bad” feedback based on the region they are in along with the ability to optionally see more details, so they can determine for themselves which of their attachments are most expensive and efficient to remove.

Strategy #3 Extreme Measures

One strategy sometimes used to improve the pool of script time is  disabling scripting features on a given parcel or given region. This will reduce the amount of resources consumed by visitors, but it can also be intrusive and lead to visitor frustration and confusion when their avatar does not behave as expected. Your guests may be so preoccupied trying to figure out what is wrong with their avatars that they distract attention from your event. For this reason it is not a recommended  approach to take. However if you do decide to go this route there are a few things to know:

If you disable scripting on a parcel, the scripting restriction only comes into effect within 50m of the ground terrain. So if your event is in the sky, disabling scripts will have no effect. Also be aware that some attachments and HUDs that have keyboard-operated controls can continue running in no-script parcels.

If you disable scripting for everyone, make sure that scripting is enabled for your “land group”. Additionally, you should ensure that your land group cannot be joined without an invite or the no-script restriction can be easily bypassed.

Conclusion

There are new tools that make it easier than ever before to avoid script lag. You can use them to groom your region with efficient effort, and communicate with your visitors to encourage them to be responsible in an accurate, helpful manner. Use the strategies above to keep your large events running smoothly.