[opensource-dev] CAN WE PLEASE STOP VIEWER DEVELOPMENT FOR 5 MINUTES

[Daniel Smith]

I agree with Oz on this as well.  There are different folks working on the
client and server sides.  Throwing the client folks over to fix server side
problems runs the risk of actually slowing things down.

The huge, looming issue on the client side is simply this: evolution.

Where could the 2.x viewer be, 2 years from now?

Where could a 3.x viewer, with a plugin architecture, be in 2 years?

Where could a Unity-based viewer be in 2 years?

Without a good plugin architecture, this viewer is going to keep growing.
It will become even more of a monolithic piece of code that wont run well on
lower end machines.

With a good plugin architecture, there is the opportunity for a solid core +
customization.  Chrome and Firefox come to mind as examples on the web side.

Having said all that though, think about this:

It is a Sunday.  Go look at what Unity 3.x (you can download it for free) is
capable of.   Can any 2.x codebase get to that level of quality in the next
2 years?  I dont think so.

Consider what it would take to have a Unity foundation, layered with the
SL-specific experience on top.  Then think about it the other way (trying to
get the existing standard of 2.x to that level of quality, and on all of
those platforms).

My thought would be: get mesh out there on 2.x viewer, and then put on the
brakes and consider direction.

Daniel  -
daniel.org/blog
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.secondlife.com/pipermail/opensource-dev/attachments/20101003/001c71b1/attachment.htm