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

I would like to see XML-RPC working in both directions. I have heard this is being worked on. What is the timeframe looking like?

Also, one of the boons of XML-RPC is that you can set arbitrarily-named variables. Currently we only get to set a string and I think an integer. Would be nice to have some hooks into whatever XML-RPC libs you are using, to set and get various values.

Examples:
setter - llSetXMLValue(hndRPC, "/info/auth/username", "John Doe";
getter - password = llGetXMLValue(hndRPC, "/info/auth/password";

This would save us a bit of time having to parse individual tokens out of a single string, and also allow for quicker and easier interoperability between systems made by different vencors, since we'd all be using the same protocol rather than everyone having to write parsers to talk to everyone else.

The SL client uses the fmod library (for sound output I guess) - can we get script interface to this library? So that we can play (and create and control) mods, midis and other sound formats. Streaming audio/video is woefully inadequate for many purposes. I'd like to be able to create a good audio tracker in SL

Garnet Psaltery and I have a project for building a railway station. The problem is the inability of building near the railway tracks because the land does not allow the creation of objects. Is it possible to somehow get the permission to build next to portion of railway track, close enough for the station to have access to the train? Or at least no to return object that extend over the railway land? This project is for community improvement.

An old bugbear but ...

Why does the system allow us to rez an item that will overlow our parcel prim allocation instead of not rezzing it and warning us if an overflow is likely ?

This old one is still wide open to griefers and accidents and can cause great upset to a landowner.

I would like to know the reasoning behind the design decision that allows new prims to be created and old ones to be stolen instead of not creating the new ones.

Seconded.

Architecture - Dynamic sim management: Why are sims and servers linked hard to each other? Most of the sims are empty or so crowded that lag occurs. IMHO in sims without users nobody cares about what happens there and no capacity is needed. The server capacities should be concentrated around user online activities to make big meetings ands events performant and hence enyoable.

Addition: I understand that certain elements to implement this are
"arriving" in 1.6.x and later version. Do you have any idea when
dynamic workload management will be, if ever, implemented in Second
Life?

Architecture - Prerendering: It would help for performance (especially for
online games zones) to prerender environments. Users must not create prims all the time and therefore the environment is more or less static. If a player wants to enter such a game zone - the cached environment is downloaded - parallel to the first orientation time or tutorial. Additional users can select their interest in advance by booking the event and can download the environment in advance. Large shops could profit from this when the environment would be preloaded whereas the sellable items are "life-rendered".

this would be, for example, extremly feasible for situations like the here adressed town hall meeting(s) - which often are plaged by lag...

Architecture - Non-real time creation: It is absolutely perfect for collaboration to have a real time prim creation or modification. But for some (common) use cases it is not necessary to make everybody see the changes in real-time. Even sometimes (work in progress on public places) it would be better to make an invisible change until everything is perfect and than "commit the change" as a whole. This would enhance performance and make local preparation (offline client) possible. Additional: Are there any plans to implement such a "offfline" sandbox client with synchronisation to the online world? For example, I could download my own peace of land, prepare it offline and "upload" it when going online the next time.

Are there plans to support more demanding, modern
rendering techniques? More extensive use of shaders and some features
of top-of-the-notch graphic hardware could bring the environment on
par with applications like everquest II or even Half-Life...

Is this possible while still providing a "less demanding" version for low entry barriers?

Seconded by myself;
Was also a request of half of the residents of the QLD - an offline build tool, or at least some way to import things built in other tools. Whichever is easier and faster to implement.

Are there any functionalities planned to create and manage user quests?

We would like to have the possibilities to give people more long-term motivation by having a quest logbook. Quest could be good for adventure games but even for jobs.

A quest should have a description, a final condition (object returned/object created, avatar at place XY) and a defined reward (Lindendollar).

Is is planned to add a jobs category to the find functionality or even a more sophisticated skill search possibility?

One problem of a lot of players it to raise money and finding a source to do so.

Another (for exampe for a group like us) it is hard to get in contact to the people who could do some jobs in bigger (and cool) projects. Maybe a tagging approach like in social software as delicious or flicker...

SL seems to be pretty empty in european daytime, due to the fact of timezones.

The small number of people online hardly ever meet each other.

Wouldnt it be good to establish timezone dependend meeting places (like telehubs but with some space)?

A worldwide understandable time definition would be good either, because the schedule for events is misleading for european (or asian) people. Swatch made a trial with beats once.

Wouldnt that be cool to use a nettime?

Is it planned to realize private rooms without all this scripting for doors etc.?

This so nice for storage my goodies or having private meetings. A kind of doorbell that contacts the inhabitant (with a sound, a IM / mail etc.) would be cool to. By this you can contact people easier and no longer are feeling like an ugly trespasser everytime you enter a house!

Thirded. :>

ad 1. We are doing our business meetings in SL. It is working out quite well, but the office integration lacks a lot.

How can we upload graphical information much quicker? (maybe with a preprocessing or something)?

ad 2. What reasons speak against a Voice over IP Integration?
If this is technically a problem, could the skype account be part of the profile and "Skype!" be part of the right click menu?
It would be good, if you could switch from chat to real voice sometimes for better understanding and deeper communication. Direct talk makes sense then!

see next one - sorry wanted to quote it - but was mislead...

Aibu

Second IT! (we would like to have non-human avatars to use for
- welcome reception
- guides to follow
- clerks in shops
- crowds some place for more lifeliness

Aibu

Would it be possible for 'select texture' mode to work like it used to in 1.5? Right now we have to click on 'select texture' -and- 'select individual parts' to select a single texture. 'Select texture' by itself now changes textures on the whole item (select object already did this).

In 1.5, you used to be able to 'select individual' and unlink a number of prims from a linked set in a row (without unlinking the whole thing). Now, each time you unlink a prim, the 'select individual' option turns itself off. Any way this can be an option?

- I've heard that the underlying LSL VM will move to Mono - how will this affect LSL? Is there a timeframe for this migration?

- Is it possible to offer differentiated sims? For instance, an old machine/void sim for sale, instead of a full-blown island sim? Myself, I have no need for an entire sim, but wouldn't mind a discounted price on a void sim. (Somewhere to hide from people

- Why is there still no unlit material prim option?

- What proportion of development efforts is focused on new features and how much for bugfixing/stability?

there seems to be a general perception that SL performs less well than people think it should. even more so with 1.6, i have personally noticed a significant drop in client FPS.

if someone had a computer with the minimum requirements


and installed SL using the auto detected default settings what kind of performance should be expected?

also if SL is completely uninstalled and reinstalled and allowed to reset all its prefs to defaults, what should performance be like?


thanks,

jester



.

1. There's been some talk of a full-featured method to display client-side rendered text and/or graphics on a prim face, produced by XML, Flash, or something similar. Obviously, this would be a huge benefit to everything from games to stores. What is the likelihood that we could see something like that?

2. Assuming that you don't have the time to actually get into that, what would be involved in adding llDetectedTouchFace() and llDetectedTouchCoords(), most recently asked for here? They obviously wouldn't be as good without text-on-a-prim, but they might be a reasonable stop-gap measure until more time could be allocated.

I'm guessing the position of a click isn't sent, as can be seen by other avatars' particle trails when editing objects. They all go into the centre of the object, while one's own particle trail goes wherever the user has clicked. Obviously that would need to be changed, but what else?

3. What would it take to implement a way for users to add simple features to the UI? I know much of the UI has already been switched to use XML, but there are a few gaps. If there were a way to first create a toolbar or window, (or even just edit the existing toolbar) then to make the client llSay something on a given channel, it'd be all that was necessary to make a game that used custom UI elements. Sure, there'd be lag, and no feedback from the client to change icons, and certainly the user would have to download a file and install it in their SL directory, but it'd be a really good start.

I second both of these.

This is sort of a followon to Tinker's post. Right after the feature voting system went up, Philip asked us to help prioritize bugs, but after the general outcry, the LL sentiment seemed to be that this should happen separately from new feature prioritization (which I agree with). In the hotline and in the Feature Suggestions forum, I offered a suggestion about community feedback for bug prioritization. Any further thoughts on this? It really would help if we could start from your existing bug lists, rather than having to file them all over again.

Thanks,

neko