Sculptie Dev

when i was first pitching the sculptie idea inside LL, there was substantial concern that without an in-viwer editor, and with only a maya exporter - the sculpties were too "raw" to be released. i argued that we ought to put them out anyway, and let our advanced residents use them as they could.

what i didn't expect was the literal avalanche of resident collaboration to bring sculpties to the masses. from tutorials to walk-throughs to plugins to external applications - you guys have really come through. it's made me very very happy. bravo. and thank you.

now i'm inspired to take this curious situation one step further. as an experiment.

below i've listed a variety of sculptie bugs/tasks/projects that still need work in the second life viewer. they range in complexity from day-long projects to multi-month projects. if you happen to be a decent software engineer, have a copy of our open-source client, and are bored - i suggest you take a look. perhaps there's something in here that "moves" you.

and because i think it's unfair to take without giving back - here's my promise to you: as a developer on the "inside" of LL - i'll do my part to enable you guys to extend/advance SL in the ways you want. first up is a small change that, had it existed six months ago, would have let the OSS community implement their OWN sculpties. hopefully, with changes like these, you won't need to wait for LL to implement the features you want.

and like i said, this is an experiment. if we fail - then no harm, no foul.

best,

K.

--Qarl Linden 11:40, 22 June 2007 (PDT)

Feedback

Qarl, I'd like to congratulate you on how you handled the release of sculpties from the Linden side. I can only hope that other Linden developers (and your managers) take it as a model.

To expand on this cooperation, we're still going to need an active dialog with you. I might have what I think is a great idea, and be willing to code it up and submit it as a patch, if I had good reason to believe that it would be incorporated into the standard client in a timely fashion. But that's a big "if". There are lots of valid reasons why it might not be picked up -- it would require too much user support, it has an adverse impact on some other part of the system, it conflicts with other features being developed internally, Linden testers have higher priorities, ... . Are you willing to play the role of Linden insider to give us your take on whether a feature would be picked up? And if you think not, what is your reasoning?

As a concrete example, there is a concensus (among at least two of us) that when lowering LOD, points should be sampled instead of interpolated, so that the sculpty creator can manage LOD without sacrificing vertices. If we created a mechanism for first building a larger concensus on the merits of the proposal, and then submitted a patch, would you be there to give us the insider's view, and then to help usher it through the Linden process, should it make it that far? --Omei Turnbull 16:06, 23 June 2007 (PDT)

Omei - sorry it took me so long to get back to you - i hadn't really published this page and i didn't expect for it to reach an audience yet. (you tricky guys.)

regarding dialog: YES. i agree 100% (which is why my delay here is so embarrassing.) you absolutely positively need that sort of feedback - and that's part of my goal in creating this page and starting this experiment. (and documenting the hell out of it, per your comment above about my co-developers/managers.)

where do you think this dialog should occur? i'm not a big fan of wiki discussion pages - perhaps on the sldev mailing list? or a new mailing list? we should decide quickly.

regarding the sampling issue - yes, let's discuss it. if we have consensus - i'll make sure it gets into the client. --Qarl Linden 11:15, 26 June 2007 (PDT)

Tasks

• VWR-866 adjust jpeg compression to improve sculptie accuracy

• VWR-1335 improve sculptie degenerate surface detection

• VWR-1336 use streaming media image-feed to drive sculpties

• VWR-698 (and related) create in-world based sculptie editors

for Qarl:

• SVC-350 create new "extra prim parameter" type for open-source (outside LL) use.