Wednesday, May 13, 2009

A Conundrum Over Graphics for Linden Labs

Location: Second Life Education Roundtable

Photo Courtesy of Olivia Hotshot

I was so proud, as a faculty member, to have two of my students, DMZ and Socrates, join our weekly meeting as part of a student panel on SL. They really did a great job.

I'll run the entire transcript soon, but one point that Socrates raised is worth an entire post. He liked using SL in a class, and felt it helped his learning. Yet at the same time, he said this:
I do not see myself using Second Life after I graduate... but I do see myself using similar technology in the future. I am attracted to the concept of a virtual world with endless opportunities but the graphics are not strong enough to hold my interest for long.
For some time, I've argued that educators need stability more than fancy graphics. I'm not a gamer so SL's graphics are fine by me. Every time they get ramped up, it means ever-increasing requirements for graphics-cards and RAM on our systems. This orphans some students in the midst of a class and makes maintaining SL in a lab setting extremely difficult.

Socrates and other potential SLers, who are used to game-style graphics, want better graphics without lag to hold their interest.

How will Linden Lab satisfy both constituencies and grow its user base? On this question hinges a lot more than the fate of one virtual world.


Anonymous said...

The young like a bit of eye-candy, it is true, but the old aren't immune to the lure of the glitz either.

Look at the experimental shadow code. On one hand, we all holler for stability, but the graphic improvements are like candy that we can't resist.

Iggy O said...

True...I'm also mesmerized by Stalin vs. Martians, a new PC game out that promises a lot of eye-candy for this WWII buff and fan of "silly games."

Speaking of PCs, I understand that the shadow code is PC-only. I bumble around SL with my time I'll get to see the shadows.

Anonymous said...

Actually, I find the shadows to be a bit overrated. But that doesn't stop me enabling them on each new version viewer to see how they perform. (A: It's a bit variable but there's a 10-95% drop in frame-rates, trending towards the larger end)

This, despite the fact that my current creative projects would require me to disable them in almost every circumstance anyway.

It's a cheap kind of excitement :)

Tenchi Morigi said...

Well to be honest all that eye candy is nice but I would at the spot trade it in for stability and less lag at once. The graphicsengine of 1.19 (the one before windlight) worked really fine and the clients back then didn´t cause me any trouble at all ... in direct opposition to what I have now.
All that eye candy won´t help when SL becomes unuseable.

Additionally I think that virtual worlds still constantly fail to draw the sharp line that seperates them from games. When they continue to stick to gaming terminology and do not offer a clear definition of what they are different. Especially Metaplace is a pretty good example there.

I think the ultimate solution would be to maintain two clients. One could hold all the fancy stuff until your eyes go sore, while the "classical" client has the old graphicsengine of 2007 which worked well for a majority of people.

Iggy O said...

Tenchi, I should press Prospero Linden for an answer about that. I think he is the server-release manager at LL, and he often attends our weekly educators' meeting (we tangled over Macs publicly at one meeting).

His anti-Mac bias aside, however, he's a decent fellow and a former faculty member. PC users do make up a majority of the user-base, and they (like you) suffer from the escalating requirements. It may be support/development costs of building two client-types for the platforms LL does support. But you'd think they could pull it off.

Tenchi Morigi said...
This comment has been removed by the author.
Tenchi Morigi said...

Maintaining a second client on the graphical standard
I agree that it would require double effort for the developers and the testers to a certain extend, but one should keep in mind that you cannot make people follow you graphic specs all the time. This will only leave a group of die hard users that would do everything to get onto the grid and I think this group is far smaller then LL thinks at the moment.

Maintaining a second client on the graphical standard of the last pre windlight version with the non graphic enhencements since then and the security updates that go along will secure the broad userbase in a long term. If they continue to further turn on that graphicsscrew they will continually narrow down they userbase.
And one shouldn´t forget one thing. We have an economics crisis out there. What are people going to do? Return to their old games they liked and still can play on their current hardware or again invest in hardwareparts to be able to remain in SL for an uncertain time longer? I think the answer is at hand.

I would also suggest to hold a gridwide poll (which would be an easy thing for LL) to see what people would prefer in first place stability and performance or new fancy graphics stuff. I would be surprised if stability and performance wouldn´t win.