[Opensim-dev] [Opensim-users] OpenSim Roadmap

[Justin Clark-Casey]

On 03/07/10 12:07, Clive Gould wrote:
> Hi
>
> We're presenting the results of Bromley Colleges OpenSim Project at
> the Future Learningscapes: a 21st Century Challenge conference at
> Greenwich University (UK) next Wednesday (7th July).
>
> Just in case we are asked, can anyone explain OpenSim's roadmap? Why
> is it still in Alpha and what are the criteria for it getting into
> Beta and eventually version 1.0?

We haven't had a formal roadmap for quite a while.  I think partly this is because of unpredictability about what's 
going to happen with virtual environments, partly because the project is still very young without surplus developers to 
spend time drawing them up :) and partly because the project isn't controlled by a single organization or person that 
could impose such a thing.

However, I'll hazard a personal guess that in the medium-term future we'll see media on a prim and general compatibility 
with version 2 browsers, continued development of hypergrid, more code modularization, mesh support, continued 
improvements in base performance and reliability and scene object refactoring to enable features such as multi-level 
linking.  Some of this will be driven by viewer development.  And ultimately, everything is driven by user and developer 
requirements, desires and effort/money expenditure :)

I would say that OpenSim is still considered alpha for various reasons, including but not limited to

1)  Current instability in core OpenSim itself.
2)  Instability in supporting platforms (e.g. mono appears to have issues with high load).  This is particularly obvious 
on relatively high-use public grids such as OSGrid.
3)  Current instability in some features (e.g. setting multiple textures is hit and miss).
4)  Inconsistency in user-facing interfaces (e.g. help on the console is very unstructured).
5)  Inconsistency and immaturity of module facilities.
6)  The prospect of future changes in some fundamental structures (e.g. scene object refactoring).
7)  Reasonably expected features missing (e.g. no coalesced objects).
8)  Bad legacy communications interfaces (e.g. the 'wire' protocol for inventory service communication is low quality 
and inconsistent).

This probably sounds rather pessimistic but I think that we've come an awful long way and it's far more valuable early 
on to have rough. working code than spend much time on up-front planning.  But I also believe that we couldn't 
justifiably move to beta without addressing a reasonable number of these issues.

Good luck with the conference!  Hope you make a few OpenSim converts :)

Regards,

-- 
Justin Clark-Casey (justincc)
http://justincc.org
http://twitter.com/justincc