Or, I am going to be a wet blanket on a generally damn fine idea.
Conversations in the SL Bloggers' group about yesterday's Grid snarlups led my esteemed colleague Snickers Snook to support a proposal by Caitlin Poultry-Seale for a Premium service that'd actually make Premium membership look like a good deal; SL property insurance - backups of your inventory, restorable as required.
It'd be a good idea. I have not, personally, had much in the way of inventory losses (barring missing out on that Chocolate Lucky Fish, but that was due to circumstances beyond either my or LL's control). But I know those who have; my friend Nyree Rain occasionally goes to the Aditi testing grid to view her no-copy items, wiped off the main grid by some random foulup. Why shouldn't that data be restored? It's all there on backup media somewhere, right?
So, I thought about it, and I asked myself: why not? And, unfortunately, I came up with an answer.
It comes down to those no-copy items, and how SL's object permissions things have been set up, and how people have worked with them over the years. So, you have lost an item from your inventory, have you? If you have Premium property insurance, you just fire up your "restore object" dialogue, and bish-bang-wallop, there it is again. Simple, right? ... Except SL is just chock-a-block with items that you are meant to lose, and there are even entire business models depending on that. I do that 7Seas fishing game, remember? How long would they last if, every time I filled my rod up with bait, instead of buying more, I just restored my last bait bucket from inventory backup?
Or, what about no-copy items? If you give one away, can you then restore it from backup? If so, expect the bottom to drop out of the breedable animals market in about ten seconds flat.
Now, none of this constitutes a technically insoluble problem. What it does mean, though, is that any inventory backup system would have to be hedged about with precautions and bookkeeping procedures to make sure that you're actually entitled to the object you're trying to restore. And this elevates the whole process from "fairly simple to do" to "a huge, huge amount of work". (Not to mention that it might involve checking up on individual contracts between Residents, which raises some privacy concerns, and which also falls under the heading of "intervening in disputes between Residents", which the Lab - not unwisely - has always fought shy of.)
This is, if I may wax philosophical for a bit, kind of an endemic problem in SL: the format presents limitations, the Residents work with and around those limitations in manifold and ingenious ways... and this means we get sort of stuck with those limitations because of the sheer inertia of all the existing content that depends on them. Sure, we could take the current permissions set (and we know, right, that "copy/mod/transfer" is only a crude approximation of what actually goes on?) and tack "restore" on as a fourth permission... who decides what existing content becomes restore or no-restore?
There is a similar thing with the whole idea of "prims", which the Lab seems to be trying to wean us off - a move which I approve of; the "prim" is an abstraction of, and in many ways a distraction from, the real processing cost of rendering objects in a 3D virtual space. But so much stuff has been made using that abstraction that the Lab would face a king-sized headache revamping it now! It would be really nice to swap that "prim count" for a polygon-vertex budget, but could LL actually implement this without pricing thousands upon thousands of current builds out of existence?... There may be more on this topic. Consider yourselves warned.