1.8 Already?

That's feted inner star chamber to you, plebian!

Aww, I thought that was 'add to land bun list'. You know, land here and get a sticky bun. Or a small cake, or a muffin.

Woohooooo we are moving on!

Good things:
- Clear features
- Features are desired
- Features don't impact the internal architecture too much -> less "omg the patch broke SL" time.
- The main feature has been discussed for a long, long time and the community has really been involved in it
- Early announcement of preview will allow for longer testing time -> better results
- Less features in patches mean easier implementations - and smoother

Bad things:
- Less features in patches means more frequent "gah, the grid is down for 2 days for patching".

Both:
- Less features in patches means less hoopla about patches. That can be a good or a bad thing. Perhaps it's good because there's less "OMG this patch sucked" because there's less expectation. Perhaps it's bad because less hoopla and more patches means more testing required, and people may tire of going to Preview over and over

Time Dilation of .04 , .1, .2 in Taber

Hiro, what that means is that major releases should not have many features? Just one or two to make us a bit happier?

Hmm.

I wonder what 2.0 will promise us then... a faster map renderer?

I'm starting to agree with Moon Adamant, who predicts in-world HTML for 2039 (I was hoping for 2019). Clearly something has to be done about the rate of introduction of new features. One day I'll set up an old LambdaMOO server again and add a few pictures, and hey, that'll be Second Life 3.0

Jokes beside, and I'm obiously eagerly awaiting the upcoming Town Hall with the project management team, but it seems that we're seriously needing a quicker rate of deployment of new features in SL. It's not as if they only have 3-4 developers now — they have 15!

Perhaps it's time for a new grid. Call it Third Life, and make it crammed full with untested features. People would be able to join either old, clumsy Second Life or bright, shiny and prone to crashing Third Life. So many software products go this route these days. May I just remind you of a few: Apache, PHP, MySQL, FreeBSD... all of them have "stable releases" and "way cool new and improved features" releases. Let people pick the one they're more comfortable with, and eventually allow some sort of communication between both grids. Looking at past things — like previews of the 2.0 rendering engine last year, or in-world HTML 6 months ago, or Babbage's almost finished Mono engine — I'd say that LL has tons of new features to deploy, but can't do so, because they fear for the grid stability, and are only fixing bugs, bugs, and more bugs.

Has anyone tried IMVU? It crashed every 5 minutes when I tried (and takes about 10 minutes to load every time I restart it). That doesn't mean that over 40,000 people use it every day. IMVU's creators couldn't care less about a few annoying bugs — you either use it and get the full range of nifty features, or it doesn't work for you — bye bye then. In their case, they don't give you a choice.

LL is much friendlier. They tiredlessly try to fix all bugs, even for people running Windows 98 with Matrox Millenium cards, or 1995 iMacs. Despite so many complains, these days you can run SL on top of almost everything with proper tinkering. I've seen reports from users with ~1.5 FPS on completely outdated and unsupported hardware (some complain, some don't), crashing every 10 minutes or so. This means that SL works well! It's more a question of fine-tuning your machine. So, give us a SL version for the ones that aren't afraid of fine-tuning their machines and are willing to trade-off a few annoying bugs in return of a fantastic experience (even if shortlived with many crashes).

Imagine a world with one million prims per server, Poser 5.0 avatars, programming in C++ with XML-RPC calls between scripts/objects/external servers, browsing with Flash enabled on in-world browsers, and with Havok 3 as the physical engine. It looks that you could have all of those by Christmas — if you only hadn't to support thousands of badly tuned computers with unsupported cards.

*Sighs* I guess that in three years, most if not all current computers will have melted down their graphic cards (I know that my old PowerBook did!), people will have to buy new ones, and so the rate of new feature introduction should start to increase. I can wait until 2009, no problem with that. But when I see startups doing this I sometimes worry if everybody is going to be as patient as I am...

I'm trying out PTP in the Vesion 1.8 Preview. Now that you can port just 1m away, there is no longer any need to...

1. Walk
2. Fly
3. Use a train
4. Use a vehicle

It's convenient, but what is the purpose of a contiguous world now? Why not just have 90k 512ksqm islands?

Wouldn't any modern version of a LamdaMOO server have much more sophisticated chat features than SL will have by 2039?

The ripples looked fine in Taber but after turning it off and then trying again in Gibson the water vanished.

For those who missed it, here's an update on the release methodology:

/invalid_link.html

The key quote:



It makes a lot more sense to me given the size of the dev team.

-Flip

Sounds like a bug. You shouldn't be able to port somewhere unless there's a landing point there, if the release notes are correct. If I'm misunderstanding them and you CAN port somewhere without a landing point, that's a problem.

Or maybe they just set all the landing points random instead of unset when they set up the 1.8 grid. If so, they should do it differently when they do the full release.

I think they should allow you to set this effect on any prim surface, like you can shiny and bump maps. Either through the editor or through a call [llSetRipple(float intensity, float wavelength, float amplitude, integer side)]. Think o' the brae banners ye could fly over yon castle in the sky!

Not possible. There's no rendering done server side. You'd have to have a client actually go there.

I wouldn't mind an [llTakeSnapshot(list snapshot_options);] to trigger snapshots on schedule as a script drives my av over a track, though.

I ported into the house, then clicked on the map as close as I could to my present location without being right on top of it and ported into structure of the wall of the house. Did it again and ported right next to house.

they should add more options on the menu.

Like there is "Go home"

Should have at least 3 more...

"Go fav place 1" and so far, lol so you dont have to look it up or get a landmark

Sounds plausible ... in a way. On the other hand we have a nice rendering engine now running on the servers (maybe not part of the core server engine but ...) which sometimes, when the moon is right, produces new images of the world map. I guess its not completely impossible to do rendering on the server side or at least initiate it.

An even nicer idea: my av controlled by a script. More of that please! Maybe in 3.7?

Maybe Lindens could add a Favorites menu like World->Landmarks folder. make it a little like IE, NN, etc. I hate having to dig through my inventory for those. I can create a landmark from that menu, but not access them?

The water ripple is nice but for me it seems to degrade over time, getting less ripples and then eventually going to almost smooth. I turn off ripple and then turn it back on, but after a little while it starts to degrade again.

I used to look forward to updates. Since 1.7, I dread them.

I figure my time in SL is limited - not by me, but by them. Meanwhile, I just continue to enjoy it, even though it's a lot worse for me. Mainly cause I don't have to do much to enjoy myself.

1.7 taught me that things will get worse and worse. That is all.

coco

The water looks very cool. The only thing is, that it doesn't match the "art style" of the rest of the SL Landscape. The only places the water will look cool is maybe the new northern sims and the private sims using good sand textures. It makes the land textures on the main land look even more dated than they are. This seems like a feature that should wait for some updated land textures and maybe speedtree also. It just doesn't seems to make much since to update water to version 2.0 then leave everything else looking version 1.3.

jello fight!

everyone get on their bikinis & thongs! FIGHT!

This is a great idea. The only concern it brings up is how big a FPS hit it will be. Imagine someone making a whole house out of this water, while it would look really cool, would it cause a huge dent in your FPS?

I'm not exactly sure how cost-effective pixel shaders are for most cards. While turning on water ripple doesn't really hit me so bad, for others it may be a different story if their card has to work harder to produce all of it.

So if it really wouldn't hurt people performance wise, I'm all for it. I'd actualy be up for taking it up one more notch and allowing users to control pixel shaders directly themselves.

My laptop doesn't pixel shade so I'll have to wait to see the ripple. P2P is nice though, especially honouring the Z coord. llMapDestination() is interesting but I think llTeleportAgent() will be a much more useful tool for ferrying people around.

The really cool thing is that I haven't encountered a single bug yet. I like these smaller releases way better.

Have you telelported anywhere?

We noticed last night when teleporting people to you, they seems to arrive somewhere within about 50M of where you are. Last night when I was being telelported places I have to track buddies just to find where they were because I landed so far away from where they were. I also noticed with Point to Point, I was teleporting about 20-50 meters away from the selected location.

Why, that's just crazed, Cocoanut.