Wanted: Questions for Tomorrow's Town Hall with Cory Linden

In an attempt to better organize the Linden Town Hall format, we're going to try taking a group of questions before hand. As such, I'd like for you to post your technical questions below for Cory to answer. The Town Hall will open with these questions and an open Q/A will follow the answers.

Three Guidelines for Posting Questions

1. Technical-related questions only
2. No Tech support-type questions (ie: "My computer is broken"
3. If you'd like to ask a question that is already posted - 'second' it instead of posting it again

I'll be heavily moderating this thread, so please stay on topic and post only the questions you are interesting in knowing more about!

Thanks!
(more info on Town Hall here )

So... is Havock 2 still on track for (last I heard) 1.7?

When can we realistically expect an increase in prim limits? We have remained at the same level for more than a year, which as you know on the low end especially is extremely inadequate for an enjoyable experience.

Not really a question but more of a foot note about the prim limit not changing in a year is the fact that they have been promising an increase for a year every time that question is asked.

Now for my question "Is there some reason why the group limit is so low. Between business and pleasure 15 is not even near enough especially because to run some things effectively you need to be in a group thanks to various ingame dynamics."

I tried to ask Philip this question last week, but he didn't get the difference between usability testing and preview/user testing. So trying again with the uber-geek...

I'm curious if LL uses live usability testing in evaluating feature design. Do you? If not, why not, and would you consider setting up a usability testing program? As you can see in this forum poll (/120/0c/43081/1.html), many users would be happy to play test for nothing more than pie!

For anyone reading who is not familar with usability testing, it involves bringing in live users and observing them as they use the product. This differs from preview or user or beta testing, which lack the crucial component of observing the user. This is important as it lets you see when there are problems that the users doesn't understand is even a problem. I've used usability testing in many projects, and it always makes a huge difference in the quality of our user experience.

The new render engine , Is it ready ?

Is it going to utilize new hardware and software standards.?

Is it being built to utilize multi Cpu or multithreading?

Will it utilize more VPU hardware?

Any thought to 3d interfaces ie screens or 3d glasses?

currently I see no method to subtract two priitives ie cut out a hole in a will for a window.
Will such a thing be possible in the future?

will spot lights and be supported in the future?

I am a photographer. I currently use a primitive with light type as the attribute to simulate a controllable light source. It is somwhat cumbersom as the size of the primitive apparently impacts the amount of light it emits. Will light attributes such as intensity and falloff of the omindiretinal ligtht be modifieable independent of pirimitve size?

I am a photographer and i use your engine to practice. Is it or will it be possible to freeze an anim or pose and use IK to position limbs for a finer level of control over the pose?

Currently this is possible in an engine such as Half-Life2's source engine

Could it be possible for a user to set a waypoint for another av? Hi me again-- photographer.

I often work with other gamers to do photoshoots in the engine. the most diificult part is getting the other player's av into position. Can the engine be modified to allow a user to set a waypoint and for the other player to opt to navigate to it?

Is there a limit to the xyx dimension of say a cube? I have noticed that my cube seems constrained in dimensiosns. I can't say what that is as I am unable to log in currently.

what is the size limit and is there a method to enable/disable-- in case i disabled something?

Is it possible through scripting to gain control of the FOV and or rendertargts for screenshots to enable simulating different camera lenses and/or cropping aspect ratios? Also will the source ever be realeased to the mod community?

What would be the minimum required adapter to use in conjunction with a 2GHZ main processor on XP to enable the full graphics capabilites to perform reasonably will in say Hawaii-- where scipting is fairly minamal-- in order to gain the most photograpically rich screen shots?

What would be the suggested VIDEO Ram?

Larry, it seems you are asking a lot of newbie type questions. It's great you are eager to learn, but there are plenty of other ways to find out the answers to most of those questions, like the Building and Scripting forums. You'll probably get a better response there too. Cory is one of the lead developers at LL, and his town hall will be focused on answering technical questions that only the Lindens can answer for us.

Will it be possible to use a an effects filter to simulate depth of field and/or lens filter effects-- either in the current engine or in a future engine in the works?

would love to be there to ask questions but cant log in sighs

We're currently using XML-RPC to bring in data from devices other than keyboard and mouse, but it's sllllooooooowwwww. We get about one packet every 2-4 seconds, even when the rest of the current system lag seems minimal. Would LL consider opening up some kind of data channel through the client for devices in addition to mouse and keyboard, e.g. joysticks, steering wheels, foot pedals, and of course, my favorite, biometric devices? It seems that this would give much better response than the more general-purpose XML-RPC, should have minimal risk to the world in general, as it would be data coming directly from the client's machine, and would allow for a much wider range of interactivity in-world. Something generic that we could use to write and distribute drivers for our favorite devices, with capture functions in LSL and Mouselook llTakeControls, would be ideal.

Thanks,

neko

When might we expect to see Flash click events?

Thanks,

neko

I think the question he is asking is are there plans to allow for larger prims? The current constraints of 10 meters are tied to the physics engine, but a lot of prims are wasted just on making larger surfaces because of this limit. It is not a newbie question, but a rather valid one - removing the 10 m limit would open up some creative possibilities that are difficult and prim heavy now.

Would it be possible to add support for avatars sitting on prims attached to another avatar? Right now, we have to use some clunky vehicle scripts to achieve certain types of inter-avatar interactivty. If one avatar could sit on the attachments on another avatar, it would allow for new types of interactions to be created more easily, in combination with custom animations.

Is LL more likely to want to support non-human avatars and/or scripted avatars (NPCs), or allow import of custom avatar meshes with scripting? Or neither? It seems a fair number of people would like to have alien, animal, or otherwise non-human avs, and similarly quite a few people would like to be able to have easier-to-create animals, monsters, humanoid robots, etc. (though I would expect a high prim cost for such objects). What is the LL position on this, especially from a technical standpoint? Which would be less problematic to implement?

Thanks,

neko

Ok, this should be my last question....

How's the permissions overhaul coming? A lot of interesting ideas were posted earlier about Creative Commons, wrappers, copyright protections, etc. Other, more pressing issues seem to have taken center stage since then. When should we expect this issue to be revisited? Is there a tentative timeline for the permissions overhaul? Some knotty issues have been raised since then regarding creator's rights vs. owner's rights, that I think will need to be taken into account, whatever happens.

Thanks,

neko

Regarding the vote system that was implemented,

Will items that appear to have a simple fix, such as "Default Streaming Audio to Stop", be prioritized over the inevitable "Havok 2" requests, and how much impact would prioritizing smaller things that could be put out with a point release have on future updates and heavy code changes?

Also:

Will support for Portable Network Graphics (PNG) and Ogg Vorbis audio uploading be realized at any point before 2.0?
Also, will longer sound clips be allowed (again?) at any time in the future? Right now, it's a bit of a kludge to get a longer sound clip to play: Split wave file into 10 second bites, upload each bite, place in object and use script that plays these clips sequentially.

Cory, could you talk a bit about how the development process works within LL?

1. What tools do you use for version control, bug tracking, and build automation?
   
2. Do you do nightly builds?
   
3. Do you use automated unit testing (JUnit, PyUnit, and the equivalent)?
   
4. What role, if any, do agile languages (Python, Perl, Ruby, etc) and agile methodologies (XP, SCRUM) play in your development?