Travelling between grids is something that really generates a lot of excitement - we've only had it a couple of months and already there's a lot of interest surrounding it.
Very good to see that you're cautious about it too RightAsRain - it is a very new facility and there are various ways in which it can be exploited. However, the contribution of community members such as yourself will hopefully get traction around clarifying exactly what the issues are and what can be done to address them, as well as encourage code contributions to fix the various bugs which are technical rather than architectural in nature. It may be that we find out that certain issues are intractable in the Hypergrid model, but we probably won't fully appreciate them without trying the experiment (personally, I tend to favour a model where inventory data is supplied from the client rather than the grid side since it better addresses the problem of exposing inventory data, but this is much more difficult to implement in the current environment).
I'm not sure precisely what's happened to OGP since I'm not that close to the process (only so many hours in a day, etc.). However, it seems to have slowed down markedly and I'm prepared to guess that it won't be revived much in 2009, since the benefits are fairly long term and the immedate risks are too great, at least compared with other initiatives that Linden Lab can devote it's manpower to).
Very cool to see experimentation and your very sensible caution
Sun, 01/11/2009 - 19:50 — Justin Clark-Casey (not verified)Travelling between grids is something that really generates a lot of excitement - we've only had it a couple of months and already there's a lot of interest surrounding it.
Very good to see that you're cautious about it too RightAsRain - it is a very new facility and there are various ways in which it can be exploited. However, the contribution of community members such as yourself will hopefully get traction around clarifying exactly what the issues are and what can be done to address them, as well as encourage code contributions to fix the various bugs which are technical rather than architectural in nature. It may be that we find out that certain issues are intractable in the Hypergrid model, but we probably won't fully appreciate them without trying the experiment (personally, I tend to favour a model where inventory data is supplied from the client rather than the grid side since it better addresses the problem of exposing inventory data, but this is much more difficult to implement in the current environment).
I'm not sure precisely what's happened to OGP since I'm not that close to the process (only so many hours in a day, etc.). However, it seems to have slowed down markedly and I'm prepared to guess that it won't be revived much in 2009, since the benefits are fairly long term and the immedate risks are too great, at least compared with other initiatives that Linden Lab can devote it's manpower to).