In ancient days we used the simplistic Linden plants. The alternative was prim plants - not any better. The sculpt plants, and since a while, mesh plants. At each stage the download size increased on a nonlinear curve - as did the quality, assuming a talented builder.
Also in the deep past there were thoughts about incorporating a dynamic tree generator, a la SpeedTree. From what I've read LL had purchased a license for it. From a technical standpoint I'd love such a system: back to small descriptive assets to download and the ability to dynamically create/adjust LODs.
I suspect there was a few roadblocks, some of which could be:
* Creator outrage - taking away jobs, etc. Same outrage at the addition of sculpts and mesh. Just remember that just like torturing a prim into exactly the shape you need, adjusting all the variables into the tree you want would still take time and skill. Not to mention all the texture work!
* Open source viewer development - if the tree generator is closed-source and expensive, like SpeedTree, what would be a fallback for TPVs that don't have the resources to license it?
* Developer time - I don't know what other projects are much further up the pipeline.
Questions to be resolved:
What other problems did I miss?
Are there solutions to the problems, and what compromises do those entail?
Is there a better solution, or set of solutions, to the core problem of trying to reduce server and network loads while increasing the quality of plant-forms?
Thoughts?
Ricky C, aka Cron Stardust