Here's the forum, where's the preview?

Was this forum prematurely renamed?

As noted : To organize, brainstorm, plan and execute testing of Second Life 1.9 in Preview. Let's get together and make some incredible things happen!

This is the organize, brainstorm part...so not early at all...

So do we all have to wear red tails?

Okies then heres a couple of things I'd like:

llPointAt() (my obligatory mention of the function that rotated and av to 'point' at an arbitary postion -- long since taken away and never replace)... or at least ONE function that will let a script rotate an AV.

derez() event - just as we have a limited amount of cycles to clean up on a detatch - why not the same for a derez.. I know a few things I'd like a cleaner exit on if deleted..

The HOVER HEIGHT limit on vehicles and physical prims - (100m global - 100m over terrain) - there are sims out there now with 90m of water -- you can go swimming in a balloon if you follow terrain - or slam into the hillside when you leave..

.... is this height over terrain - the difference in terrain heights when crossing sims part of the 'falling into the ground' vehicle sim crosses experience? I dunno - maybe it does.

Updating the DIALOG box - I really appreciated it when it was expanded from 4(?) to twelve buttons - I'd like to see this evolve into something like the old Amiga Requestor Library myself - type text into a dialog box and have that delivered to the dialog channel.. building custom requestors from paramater lists - fun fun fun for all the family

Okies thats off the top of my head while at work.

Oooh, I know!

derez()
{
  llRezObject(llGetObjectName(),llGetPos(),llGetVel(),llGetRot(),1);
}

OK, OK, I know that wouldn't work, I know why, but it's funny.

Here's what I'd like to see in 1.9:

Redtail-Linden-style flexible primitives.

Especially if they do something intelligent when applied to a linkset.

llTreeSystem(list attributes);

Analogous to llParticleSystem, to let people create their own linden-style trees.

llSetAnimationTarget(string animation_name_or_key,vector offset,rotation rotation);

When llStartAnimation(same_animation_name); is run, the avatar is displayed with the specified offset and rotation. This would allow animation overiders to apply the same kinds of corrections for avatar size as furniture and poseballs can.

llSetAnimationScope(string animation_name_or_key, integer bonemask);

Similar in application, to limit the animation to only moving the named bones. This would allow you to use a full-body animation (like, a seated piano player) but only run part of it (like, the upper body only) for a specific animation (say, typing).

llRequestor(key id,string title,list objects,string format,integer channel);

Your smart dialog box. I fully endorse this proposal.

Smarter loading textures.

The default texture for phantom objects (phantom prims, particles, attachments) should be the old smokey one.

if you go and read how to do cleanup on detach you'll see that it states you have a small number of cycles to do a limited cleanup..

... now go and read rez object - and it's delays...

Of course not stating the obvious - that people already to this using more inefficient methods - such as a set timer, rez object, and die..

If they're gonna fux with it anyways - at least let them do it efficiently.. besides I think such a state -however implimented would do more good than harm.

*obligatory mention of case - structs - arrays*

I'm still waiting for object to object communication. Preferably grid-wide. E-mail is way too heavy for simple messages.

Another cool thing would have to be:

llSQLConnect(string host, string username, string password)
could return a key to ID the connection.
and
llSQLQuery(key connection, string query)
could return the response from the SQL server, or use a on_SQLResponse (string response) event.

LL could even sell SQL server space for it, and just add the cost to the tier. I wouldn't mind using a LL server for it, as it would probably respond faster than an external server.

I was just coming back to suggest that - yes I think that is incredibly important.. Email for the amount usually sent by a link message seems incredibly wasteful..

what was the other thing I was thinking of - oh yes:

rezzed (key id)

We get on_rez (key id) when we rez something - why not a callback on the key of the object that rezzed me?

Again - one of those things we cludge around daily - especially with objects sending info back n forth - doing it simply would be great.

I already said that this isn't a real problem, and I think on_derez() would be a useful tool, but I don't think delays work the way you're implying.Delays don't actually use up any cycles. They delay the script, so the statement after the rez would never get executed, but the rez would still happen. And, yes, it would be useful to be able to create a "successor" object like this.

A bigger problem is that on_derez isn't going to be reliable. What would happen if the object went off-world? That's not the same as being derezzed.

Yes, I was very dissapointed with the lack of obj-obj comm in the last two updates like we were promised. It was even finished they said (the single sim version). They stated its release in 1.7 as a fact not a hope. Bah. It would make things so much better in so many ways.

and as far as cases and structs: not gonna happen. No major changes to LSL until Mono. This is slated for 2.0, but we'll see.

I've actually discussed an SL webhosting service with a couple of ppl. Think it would be a great idea. Have a built-in method of communicating to/from SL for the resident business owners who aren't as saavy in such areas as XML-RPC or e-mail parsing. Whether it's hosting by LL or third party would bring a lot to SL.

I'd like to see an enhancement to snapshots.

In particular the ability to change the savename of "save to disk" snapshots after you set the name of the initial snapshot.

Eg I dont really want all my snapshots being friends_001 friends_002 etc where friends_002 may be an unrelated picture.

True, but it works great for 2005-01-14_001, 2005-01-14_002, and so on. Changeable on demand would be nice, but I do enjoy my semi-automated date keeping.

I have a HUGE request that's been bugging me since I started just over a year ago.

MOUSE OPTIONS!!!
(not talking about mouselook btw, just the ordinary use of the mouse for interacting with the viewer)

1. The mouse speed is way too slow for me. I feel it would be SO easy to implement a slider for this, as in any graphical OS out there. It's essential for SW ergonomics.

2. Swapping of mouse buttons should be possible. (Mac users need not appy! ) I am left-handed and SOLELY for SL I had to move my mouse over to the right-hand side. Surely this must be dead-easy to implement, no?

To me, this is a serious issue - I've developed an inflammation in my right elbow (like "tennis elbow" and since all I ever do with my right hand is use it for the mouse in SL (in work I use my left hand) I feel my pain is directly linked to playing SL. When I play, I click on the wrong objects and chatting becomes painful very quickly. Very soon, I will need to take a break from playing for something that could be resolved so easily!

PLEEEEASE????

Thor

Can't you just swap the mouse buttons in Settings? Or am I not understanding something?

Well unless I'm blind - no, you can't. There is an option "Invert mouse" but that applies to Mouselook and means that when you move the mouse down, your view goes up.

Mouse buttons and mouse speed are such BASIC features, I can't understand why they are not there -- particularly since overall, the viewer interface is really pretty neat.

Thor

What kinda mouse are you using, Thor? Many mice have drivers with application-specific settings to customize. For example, my Microsoft Intellimouse Optical does. I can't set tracking speed differently within just SL, but I certainly can swap buttons.

I meant in Windows settings - in mouse - you can swap the buttons.

I've been in SL a short time, short enough to not be entirely accustomed to and no-longer-bugged-by some of the silly things in SL.

SL has some awsome points, but compared to other "games" SL is severly lacking in a few basic areas.

Top 5 things that kill the experience in SL:

5) Water? Flying under water, perhaps an AV could swim? No splash sound when you enter the water. The horizon looks the same. SL water is some strange mater that is basically air.

4) AVs aren't able to go up stairs smoothly. Currently you have to bump into each stair and keep pressing against it to make it up the stair.
if (nextstep.pos.z < current.pos.z + max_stair)
{
nextstep.pos.z = current.pos.z + max_stair;
}
?

3) Teleporting takes you to somewhere other than where you landmarked with a red beacon to guide you. Some might say computers are my forte, but I still wasn't able to figure this one out for quite a while..

Recommended solution, after being teleported and not landing where you had expected to go, land owners can optionally set a timer/yes-no dialog box/scripted process to automatically jump you to your final destination once you've taken in the warning about mature content.. or what ever needs were met by the intermediate location.

While on the topic, it would probably be good to also have an option to allow land oweners to permit people who have already been through the process to bypass the process/intermediate location and jump directly to where they want to go.

2) Search results are limited to 100 and full of spam for shops, casinos, clubs, etc.. (the high level categories are generally not granular enough to weed out irrelevant stuff)

Rather than/in addition to providing a search interface for places[/events] an organized directory/hierarchy would help new users find what they're looking for.
eg.

SL/
>Games/
>>RPG/
>>Casinos/
>> ActivitySims/
>>> Abbotts..
>>> River Walk..
> Clubs/
Mature/
>> Themed/
>>> Furries/
>>> Vampires/
...

Perhaps the LL could even give users "jobs" helping them classify, and maintain, various places/events.

1) Why is SL touted as a ready-made game development platform, when the only real successful games in SL are 2D non-mporg games? (SLingo anyone?)

Aside from Sims like Abbotts, the best actual game I've see in SL is Darklife. No offense to it's creators who have done an awsome job with-in the constraints of Second Life, its a far cry from what you might expect, considering the "ready-made-game-development-platform-touting".

Whoa! This is some kinda sugartrain. What an apt way to describe yourself... and a warm welcome to the SL Forums too, Heather!

First, the usual state that my sightspeed is lateral. Trains that can travel sideways are normal to me and I'm always excited to see ideas bustling 'bout.



An adjunct is the introduction of Ripple Water in 1.8. Before 1.8 (and my current computer), I used to really cringe at the horrible sun reflection that ate through the landscape. And then in 1.7 it was limited to the water... and then in 1.8 it's improved oodles. I'm eager to see what's next!

But for now, in the here+present... Resis have come up with some kewl additives to this. At the basic level, you have animation overrides that can detect when you're in the water, and then put your av into a swim animation. Splashers? Jillian Callahan has made splashable water and BuhBuhCuh Fairchild made a splasher attachment. Siggy Romulus has a swimmer that simulates buoyancy so water behaves differently too. Cid Jacobs and I once did a bunch of tests to "float" watermelons. Any other inventions and I'd LURVE to hear about them so I know more about what's out there. For water, this can't be a vertical market.



Yeah, the old bumpity-bump-bump eh? A couple of methods of dealing with this are 1) put a script inside the staircase that mutes the sounds walking on each makes or 2) overlay an invisible prim that's like a smooth ramp upwards. It can't be seen but you'll glide.

And where are all the ladders too, right? Would lurve to see more climbable ladders and fireperson's poles. A couple of many possibilities.



It's great to know where you're coming from. I just threw my mind into that gear and I had such a vivid mental image.

I can understand why this would be confusing! Here's what's happening: some parcels of land (not all) have Landing Points, which send you to where the landowner wants. For example, a store may have a landing point set up at the New Products display. If a parcel doesn't have a landing point, you'll be TPed directly to a specific coordinate of your choice sans red beacon.

To check, click on menu bar on name of land ---> OPTIONS tab then look next to LANDING POINT and TELEPORT ROUTING.

And you may already be familiar with history, but just a few months ago, we didn't have such direct teleportation.




Oy, better search to get people connected with where they wanna go. ("Granular" is a fave word of mine, BTW.)

Residents are creating their own solutions, like http://www.slwebsearch.com/ and http://www.roamsearch.com/ and if there are more send them my way so I can search too.

And re: the builtin system search tools, that is important and constantly being thought+felt about. One key already has been an improved Map in recent times.

What do you think of "tag-based taxonomy" (as much as a mouthful as that is for me) classification?



Some peeps have brought this up to me too on recurring occasions, and it's one of those observations that brings me back to how one person can come along and do so much. And do things with other people, and do more. (That's a lot of doing.)

I enjoy playing a lot of games within SL, and it is kind of a funny analog: two avatars sit down at a chess table and play. And chess, of course, is planar.

Sometimes in SL, I believe we're starting from scratch in implementing so many new things within this online environment. We're not reinventing the wheel, strictly, but there are things that look like wheels--and physics engine willing, hopefully will behave like wheels. And then it's onto more growth, blossoming, the next generation, and soforth. Once mastery is attained of the simpler experiences, then as dimensions expand, so can we advance together.

I think a key operator to add is "yet". I've seen some bodacious experiments that were hampered by technical limitations (i.e. first-person shooters getting hampered by lag or not more of a flexible combat system, adventure games not being able to hide treasures 'cuz of the prevalence of camscanning into walls), but these baby steps are just gonna keep--stepping!--and I think it is sooo important to bring up schtuff like this to encourage positive creation in this world.

Beautiful!

Great ideas, really great. Given the history of the last year, isn't it a little misguided to assume that there is going to be any work in 1.9 that isn't sheer fluff or targeted to keep SL from failing even worse under increasing load?

Sure, I'd like 1.9 to wash my car too, but here is a heavily excerpted list of some of the "major" changes that almost the whole last year has brought to SL:And who says there is no such thing as progress?

Thanks for the responses.

Again, I'd like to emphasize that I think SL has amazing potential, if it could solve a few basic problems that seem within reach if they're A) identified and B) given proper attention.

A couple points that I may not have emphasized enough in my previous post:

1) The things I listed in the previous post kill the SL experience for a newbie. If you want more people to stick around/pay money, I wouldn't develop under the expectation that new users are going to stick around for a month or two to learn where to get the right third party stuff to make the experience something that isn't a little disenchanting.

LL could A) endorse third parties in some obvious place that new users can get to without much searching, or B) even buy their IP from them so that they could make it available to new users on their first login.

2) Related to #1, as a software architect/developer, its good practice to provide "reasonable defaults" that fit with the majority of use cases, while providing the flexibility for users to modify these defaults for the uncommon cases. Does it make more sense for every one who wants stairs to put a script in their objects to walk people up stairs, or does it make sense to make the ability to properly walk up stairs in the module that controls the AV? Same for water. Does every object have to be scripted to behave reasonably around water or is that a "reasonable default", that people can disable/override if they need some other behavior? It would seem this might help alleviate some of the pain from the lag caused by too many scripts running...

3) SL is not going to reach its potential to be a ready-made mmporg development environment until it:
A) provides reasonable defaults so that developers don't have to do the heavy lifting to get something decent/reasonable behavior working
B) provide a full fledge programming language, my recommendation would be to plug-in (something like) java bean shell which already has a lot of community/IDE support. I would think this would take a significant burden off LL, so they can focus on other aspects. To this point I haven't been able to figure out a way to make a library of common functions that I can reuse between all my objects. Instead I have to cut-and-paste (a big no-no in my book of software development practices) code into each script. Not to mention the value of OO, especially for game development.

--

I'd be interested to know what LL has planed for future releases (more than the next minor release). Improving the physics engine is good and all, but my general feeling is that there are a few basic/more important things that can be done to significantly improve the user experience even before this happens AND that things like the physics engine aren't going to help all that much unless these basic things are resolved.

Likewise, scalability isn't going to matter, until you've removed the barriers that are keeping SL from needing to scale. Plan for it, yes, but focusing development effort on it seems a bit like wishful thinking.

Thor go buy a left handed mouse or like has been suggested change the mouse settings

I would like to see something like llSetPrimStreamingURL that works in exactly the same way the llPlaySound does with the volume etc but instead of playing a tiny sound clip from inventory it plays whatever is being played on x URL.

after she picks herself up off the floor laughing says they cannot even get SL grid working right let alone anything else more complex