Chest vs. spine

The only consensus for attachment points I've managed to suss out is that tails go on the stomach.

For other possible points - I suggest including two copies of your product in every sale. One that fits the chest and is already positioned for it, and one for the spine.

- If you must put something on stomach, because you also need something on pelvis, put the less important to the look item on stomach. Because every Neko, Usagi, Furry, Kistune, Faun, Satyr, and etc in SL won't be wearing the part that's on the stomach - and that's a big crowd.
- I keep a list of designers I consider 'Neko hostile' because I've had to deal with moving an item off of stomach and onto pelvis with them. If in so doing I have to lose whatever was on pelvis, and doing so destroys the look of the item, its a 'blacklisted' designer - I won't be back and I'll warn others away.

Its very easy to sort them out by just adding (chest) or (spine) onto the end of the object name of each.

I found it easier to make sure I always have to versions of my tails, stomach and pelvis, so I didn't have to be concerned with clothing, belts etc.

Just for the record: I had another test with that attachment script in a different sim and by rezzing the object first, and now the script works perfectly fine

So I'm definitely using this from now on. Problem solved
Thanks again, Innula.

Yes, in terms of sim lag, it's not worth worrying about. And it's probably not worth thinking about otherwise, either, but just to cover all bases:

Scripts in attachments are special because they move with you to other sims, so they have to get removed from one sim and loaded into another. Both of those operations seem to be more problematic for Mono- than LSL-compiled scripts, at least right now (and to some degree always will be, given the Mono architecture).

Really, transfering any rezzed asset from one sim to another involves some overhead, so it's a win to minimize the number of prims worn, and the number of things inside those prims. So there'd be some small benefit to combining tiny scripts as long as the resulting script doesn't grow to use a larger increment of memory.

In the middle future, each agent and parcel will be alloted a finite amount of script memory. At that point, there will be some advantage to Mono-compiling tiny scripts because they'll be able to use less than the fixed 16KB of memory of every LSL-compiled script. But for attachments, that advantage is weighed against the above disadvantage for TPing or moving from one sim to another.

Finally, every script uses a very tiny amount of sim time even if it has no active event handler. This is an oddity of the way scripts are scheduled, and is negligible for this application. It really only matters for extreme cases, such as thousands of sleeping scripts per sim--a condition that can arise, maybe, with enough fake "water" prims still holding a dormant texture animation script, or the like.

Concensus has been reached, nice! And not a single negative or off-topic post in the thread, how cool is that, hehe. Yep, that script is only active when attached/detached and only stores the data for the points it has actually be attached to - so only try/set it on 'sensible' attachment points if you want to minimise memory-use.

One other point that hasn't been made - with mod+copy items I will sometimes link copies of my favourite combinations of, say, collar and necklace, into a single object so that they can both be worn on the same attachment point at the same time.