llWarnUser(key id, string message)

llWarnUser(key id, string message)

Would bring up a dialog box with a /!\ icon, and would display something like
"This is a warning from the land owner:" +land_owner_name+"\n\n" + message

The message string could be up to 1024 characters and the box would be larger then the standard dialog box as to better allow for the text to fit.
There would be two buttons: [Ok] [Go Home]

After this function has been called, the users key could be used by any and all security scripts on the parcel owned by the owner (not just the one issuing the warning). Warning could only be issued by land owners. Until this function is called, normal security functions would only result in an ejection.

A smooth transition could be done; for a one month period, an error could be issued to the land owner on the calling of land security functions saying they need to call this function first (but still have it execute the function as intended). After that month, they remove the error message), then have the security functions result in an ejection instead.

There would be no delay associated with calling this functions (so you could be a turd and just ban people after calling it).

I'd suggest making it have 3 buttons...

1 - OK (acknowledge the notice, do notice)
2 - Go Home

Two possibilities for the 3rd button

3a - Pushback (move me X metres directly away from the scripted object and bring me to a complete stop)
3b - Pushback (move me to the nearest edge of the parcel the scripted object is in, and bring me to a complete stop)


This way it won't disrupt random exploration so much.

VERY interesting idea. How about three stages?

1. Warnng comes up, unwarned security scripts generate script error but work.
2. Warning comes up, unwarned security calls eject agent.
3. Warning comes up, unwarned security calls generate the dialog and delay 30s before ejecting.

I would suggest that llUnSit() for avatars not sitting on the task be considered a security function.

Don't forget return objects function as well in case they were driving around like a fool ignoring the warning.

I agree with this correction/addition/expansion/something ending in ion
Except that another possibility for a script not calling llWarn correctly (e.g current products) could perhaps call a generic warning as part of the security option? These aren't used frequently so the overhead of building in a generic warning wouldn't be much.

My low-lag skybox security system is designed to do basically this already (warn the first time then after 30 seconds or so go for it), so it wouldn't need much changed.

Moved to Current Version Feedback