[sldev] Viewer Roadmap

[Steve]

*Hello everyone.*
This is Steve Linden. I am a studio director at Linden Lab, focused on 
development of the Second Life Viewer.

I don't post here very often -- not because I don't value the input and 
contributions of the open source community, but simply because we have a 
lot to do, and *good* communication takes time, lots of time. However, I 
want to at least do a better job of communicating what we are currently 
working on in the Viewer and why. Starting this quarter, we will be 
regularly updating the Viewer Roadmap 
<http://wiki.secondlife.com/wiki/Viewer_Roadmap> public wiki page, 
posting here with a summary, and answering any questions that come up.

*Viewer Roadmap*
This quarter, we are focusing our efforts on four things:
1. Stability of the client: Viewer Stability 
<https://wiki.secondlife.com/wiki/Viewer_Stability>
2. Small changes to improve the experience of new residents
3. Notifications 
<https://wiki.secondlife.com/wiki/Notifications_Redesign> and Skinning 
<https://wiki.secondlife.com/wiki/Skinning>
4. Texture Pipeline Improvements 
<https://wiki.secondlife.com/wiki/Texture_Pipeline_Improvements>

The project pages linked above describe the objectives and priority of 
the respective projects. Briefly, we are primarily focused on stability, 
while making time for improving some other key pieces of our system. As 
the Roadmap Diagram 
<http://wiki.secondlife.com/wiki/Image:Viewer_roadmap_q1_q2_pwiki.png> 
illustrates, all of these projects work towards two primary goals:
1. Make the Viewer more *stable*
2. Make the Viewer more *usable*

*FAQ*
I realize that everyone has a different perspective and many will wonder 
why we are not working on certain features that are important to some or 
many of our residents.  Let me attempt to address a couple of likely 
questions here:

Q: What about "Feature X"?
A: We have been very feature focused in our development for a while now. 
As anyone who has spent a lot of time looking through our code will 
know, much of our code base is poorly structured and difficult to 
maintain. And as every resident of SL knows, the client suffers from 
crashes and memory leaks. Improving stability requires a lot of 
resources. So, for a little while, we are keeping feature development 
down to a minimum.

Q: What about fixing bugs?
A: We will continue to address bugs as they come up and prioritize them 
as usual. Critical bugs will be fixed and others will be evaluated and 
addressed as we can get to them. Much of this effort is done during the 
Release Candidate 
<http://wiki.secondlife.com/w/index.php?title=Release_Candidate&redirect=no> 
process, and many of the bugs come from the public Jira 
<http://jira.secondlife.com> issue tracker, so any contribution to that 
process helps us considerably.

Q: What about resident provided changes?
A: These are always tricky. Writing code to solve a problem is only a 
small percentage of the work required to introduce a feature or change 
existing behavior. Broad design issues need to be considered; the code 
needs to be checked for vulnerabilities and edge cases; the changes need 
to be tested; and invariably, bugs associated with the change need to be 
fixed. Outside of straightforward bug fixes, we are unlikely in the near 
term to integrate external changes, just as we are keeping new features 
to a minimum internally this quarter.

Q: What will we work on in Q3, Q4, and beyond?
A: Right now we are proceeding cautiously when looking past the end of 
Q2. As we make progress on projects in Q2, we will start to put together 
a project list for Q3 and expect to have a Q3 roadmap available before 
the end of Q2.

*"Dazzle" and Skinning*
There has been some amount of feedback regarding the release of the 
Dazzle UI changes before we release "Skinning Phase 1 
<https://wiki.secondlife.com/wiki/Skinning#Phase_1>" (which will allow 
residents to install custom skins and switch between them, e.g. switch 
to a "classic" skin). This was an internal decision based primarily on 
the desire not to try to support more than one official skin. We plan to 
release Skinning Phase 1 before we make 1.20 mandatory (hopefully well 
before-- we are trying to keep mandatory updates to a minimum), so we 
will support switching skins before imposing this release on everyone. 
However, this is the look we are planning to support going forward.
That being said, because the majority of the objections to the new look 
appear to be contrast and color scheme issues, we are going to 
immediately prioritize a project to provide an in-client interface for 
modification of custom colors and maintenance of these customizations 
across updates, just like we do for other settings.

*More Questions?
*I will do my best to answer any direct questions, but it's a busy start 
to the quarter!  I'd like to focus my available time, so please put 
specific *feature* discussions on a different thread, thanks! In 
particular, let's continue any Dazzle related discussions on another 
thread -- other than trying to be very transparent with the roadmap I 
can't offer much more on that subject and will defer additional comments 
on that topic to others.

-Steve
Director of Studio Shiny


-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.secondlife.com/pipermail/sldev/attachments/20080423/645b57b9/attachment.htm