The price of a pretty sky

Businesswise, as someone who has run a moderately sized engineering department before, it's not a matter of either 'fix it' or 'add features'.

It is: Fix it AND add features!

That's the nature of business. It has to be functional AND relevant on the market.

Sort of like: cars need the engine to work AND have a radio.

Fun, huh? Nobody much cares if you have finite resources. Everybody's got finite resources.

* * * * *

As controversial as this opinion might be, it could be a lot, lot, lot worse.

And if you look at the competition (moment of silence as everyone looks around, tumbleweeds blowing in the dust) the competition really *is* much worse.

So much worse, that nobody's got anything even close to Second Life yet.

I think a common misconception here, as well as in nearly every online game/environment I've been part of, is the idea that a company can only really work on one thing at a time, or at least, should do that as a more effective model.

Generally, companies that work on game programs or the like are not bottlenecked into one activity funnel for a reason. It's still impressive to me that SL has 70% of its technical staff working on bug resolution; that's a pretty high ratio. The fact that other people are not working on said bugs is the result of differing departments and responsibilities. Not everyone a company like this hires is best deployed in code ripping; some are there for artistic enhancement, interaction, 'new ideas' and the like.

I highly doubt the acquisition of pretty skies has slowed down their other efforts.

I would guess that LL have to tread a tight-rope between fixing performance issues with SL and racing to stay well ahead of any opposition in terms of capabilities and realism.

Also it's quite possible that if LL ironed out all the bug we'd still have lag, crashes etc because of internet or home PC problems.

Imagine; if game manufacturers had stuck with bip-bop diagrammatic black and white computer tennis, it would work 100% perfectly by now!!!

Actually that sounds about right to me, coming from electronics hardware / firmware / software design.

For instance, in the automotive industry one can redesign some of the simpler parts in a week, have 'em rapid-proto'd in another week, and installed on a pilot fleet of vehicles a week after that.

Then it takes 2-3 months of testing to see if what you designed has held up as everything ages, interacts with other aged parts, and comes across odd situations. This is just as true for a bracket as it is for a firmware change to the engine computer.

The second life grid also has a testing problem. The aggregate behaviour of 40,000 people online for weeks, on thousands of servers using multiple clients on thousands of unique client machines can't be simulated very well.

So what that means is that pretty much no matter what, if development progresses there is only one true test to see if something works (unless you have 40,000 people, 40,000 computers, oodles of bandwidth and 10,000 servers lying around separately for testing).

They *have* to test new stuff on all of us, and probably, this is just how it's gonna be. Am I happy about that? No. It's just how it is.

You would at least hope that if we are all the Lindens Guinea Pigs, that they would pay some attention to us when we squeal........

I get the sense that they're paying aLOT of attention to us.

I'm kind of amazed by this thread. Look at some of the other games out there.. the skies in my Unreal Tournament 2004 game are vastly superior then what we see in sl.. heck, the skies on the original Unreal game from about 7 years ago are vastly superior to what we see in the SL. it's FAR bnehiond the times, graphically speaking, and any shot in the arm is a good thing, IMHO.

The game engine that drives the Farcry game is being upgraded, and there are tech demos popping up on it, and it's jaw dropingly real, and has a full blown editor to create your own maps.

I'd say graphically, SL is in about 1997 right now, and REALLY needs as much improveent as possible.

Good example: If I own land, why can I drop down my own land textures, theat blend at the borders? On a mountain, you'd have patches of rock, and even vegitation areas.. on an island, you'd have oasises, and rocky areas sticking up. In sl, the land is what you get, unless you build prim ground, whioch looks like crap. (Hopefully sculpies will cure some of this.), but we should be able to drop textures on ground as we see fit, especially for land owners.

I agree with you on the sky quality. But if everyone could apply their own textures on the ground, then sl would look even more cut up than it already is. It is the common ground textures that make sl still look like one world, despite all the crap builds everywhere.

I kind of got the impression that it was cached or streaming video to the great big skyprim.

Just because a lot of peopel have no common sense, does not mean everyone else should be limited because of that.

Which is better: someone who applies some ground textures to make an oasis and builds his house and garden on it, or someone who drops a BIg SQUARE MEGAPRIM with ground textures and builds his home on top of that, because the land cannot be textured?

The last land I owned had exactly that happed.. EXCEPT the guy bought land around a mountains, and dropped about 30 megaprims to make it a huge plateau. All the land around it had massive cube grass walls around them, and the Lindens did nothign to remedy the situation at all.

And of course building land that way also uses more prims, and puts a huge amount of lag on the sever.. some megaprims take up about 100 "prims" load on a server, and some are MUCH worst!

I have a freind now who's mountain top mainland, and she has about 10 megaprims on her land now, converting it from a rock mountain top to a large square grass area, with fountains and gardens. Inside, it looks great.. from outside, it's inorganic and looks like hell, but it was the only choice she had with that land.

It's quite possible that this technology will prove to be less
of a burden to clients than the existing sky generation and
look better at the same time.

Better programming can do that.

Yep, plus many advanced video cards might be handling the burden, intead of the main processors.

I know with UT2004, I have gorgious skies, and run 1600x1200 at 80+ fps.

With the current crap sl sky system, I run 1024x768 I think, and I'm having a great day if I get 12 fps.. LOL!

To me it seems that lag is only caused by having a bad internet connection to the particular sim server you are on. If my internet is seeming speedy in general (in terms of web browsing, etc..) then I usually find that SL is speedy. My graphics card isn't the best, just some $100 nvidia card... but it has no problem keeping up with SL graphics and effects (bring them on, utilize my graphics card to the fullest). The bottleneck is just the internet connection.

That's because the graphic tech in sl is WAY far behind the times.. any decent graphics card (note decent, NOT great!) can handle anything sl tosses at it without issues.

the problem is packet loss, and bandwidth issues in most cases.

A question. How many of you have actually thought 'SL skies are not photorealisitc and I am not enjoying SL as much as I could'. I'm sure 99% of us haven't given that thought.

Will we see more signups / retentikon when we will have prettier skies? I doubt these figures will be affected.

I agree that fixes and new developments can happen at the same time. But LL have limited resources and need to prioritise work (by their own admission).

A refer-a-friend program with the proper tools and reporting can help SL sign up premium accounts, which are the most valuable to them. I would love to invest time and money marketing SL to profit from the scheme but I am not in a position to do so as I have no metrics to work with. Instead I read up about a new feature that will have very little impact on gameplay or the business.

I'd like to hear the opion of anyone who signed the open letter (http://www.projectopenletter.com/) as well.

Actually, I signed the open letter. And since I live almost exclusively in skyboxes I often find myself saying "Holy crap that sky looks horrid"

There's always Linux )

I'm looking forward to it...maybe SL will start looking like a game from 2001 instead of from 1997!

Since it's all client-side and pretty much pre-developed by a company that specialized in this feature, I'm down!

You know, with all this complaining about lag and bugs and whatnot, I'd like to ask this question:

If LL told you they would fix 50% of the major issues IF they took the main grid down for 6 months...how many of you would actually be okay with that?

...and how many of you are lying?

But then it will be no good with most other programs!

Open Source

Do they charge us tier during that 6 months or not?

I agree, I have a 486/66, a Celeron 400 and Pentium166 that won't run SL and a 850 that has problems, it only runs on my 2.6 and my 2.2 dualcore so think we should be removing features to make SL run on older machines...........................
<sarcasm mode off>

Yes, because they still need to pay their employees for the work they're doing during that time.

Many times, I have looked up at the sky, and wondered where the hell the sky WAS. Why? There isn't one. Now there will be!

Again, this is going to be more than just prettier skies. MUCH more.

Yes, they have limited resources and need to prioritize. This integration looks to be close to completion, and it happened as a result of open sourcing the viewer software - the guys over at WMI have already done most of the work. Great way to prioritize.. fast track the most phenominal visual improvement Second Life has seen to date, and not hardly have to lift a linden developer's finger to do it!

And with all due respect, you're way off base when you say it will have little impact - it will have a HUGE impact on the user experience. It's a substantial step toward bringing Second Life out of the visual stone age.

<-- Didn't sign the open letter. I do not support obstructing progress, as doing so will not make the fixes come any faster.

Onward, Linden Lab. Onward Second Life. Onward.

I think the sky will be a good thing, unless it means more and more people can't run SL.

As for "obstructing progress," I would have to disagree. Imagine if my town decided they needed sprucing up, and painted all the buildings, and put new things in the park - all of that wonderful.

EXCEPT - you couldn't drive to get there, because the streets were full of potholes, and the city electricity wasn't working, etc.

In that case, I imagine the citizens might start saying, "Hey, enough of the painting and the park benches! Fix the damn roads and get the electricity back on!"

Don't you think? Or - do you think those citizens would be "against progress?"

coco