OSSL notecard functions and threat level

[Min Tigerpaw]

Did some checking of the os Notecard functions in the Metaverse: osGetNotecard (threat level very high) and osMakeNotecard (threat level high)
Both are used in my MT Collegeboard to save and call written and drawn content:
If I rezz the board in sandboxes of osGrid or Germangrid the board is fully functional -
I can save complex board content to a notcard inside the board with osMakeNotecard and I can call any notecard from inside the board will osGetNotecard. Same as in Kitely in my regions.

Seems the threat-level is not causing the described restrictions in some major grids. I know that it does so in Metro though. Will continue testing and vote for removing such restrictions as much as possible. As Handy wrote - it is a real nuisance not to be able to store data related to an object in-world!
We should not copy-paste the bad concepts from SL in OpenSim.
Freedom to Kitely!

[Ilan Tochner]

Hi Min,

This configuration can be per region, it doesn't have to be grid wide. I suspect different regions on bring-your-own-server grids such as OSgrid and Metropolis will have different settings for this.

[Kayaker Magic]

My experience is that lots of grid owners see the THREAT LEVEL of many osFunction calls and set the threat level very low for fear of some unknown threat. OSGrid for example is made of 1000's of different computers (many running older versions of OpenSim) and you have no way of knowing what threat level is set on each. As a result, my assumption is that I CAN NEVER USE ANY osFunction calls if I expect to be able to export the resulting script. Even the functions with a treat level of "may startle kittens and small children" will be disabled somewhere in OSgrid.

I also lack confidence that the threat levels of many osFunction calls were set the way they are for ANY GOOD REASON. For example, I can see that osGetNoteCardLine() has the potential that some idiot scripter will call it in an infinite loop and tie up one of the script threads. (But llSleep() does the same thing, without the loop. I've seen flashing holiday lights based on llSleep delays bring a SIM to its knees. llSleep should have a THREAT LEVEL of "SUPER PROMPT CRITICAL MELTDOWN"!) But osMakeNotecard() and osGetNumberOfNotecardLines() are never called in a loop, they should not have the same treat level as osGetNoteCardLine(). It looks to me like someone just lumped them all together by association, not for any good reason.

Other grids add just a MakeNoteCard() function (InWorldz for example) and you have to use the normal LL dataserver process to read the notecards. Things work great over there, never heard of anyone griefing with making notecards.

[Jack Spade]

Handy, I expect the concern is that, because they bypass the dataserver event and execute synchronously, large number of notecard reads could be used to lag the simulator. It jumps out to me as an obvious griefing vector.

[Handy Low]

And yet there's nothing to stop someone calling, say, llGetInventoryNumber() in a tight loop.

I've asked in the OpenSim user group - perhaps someone there will clarify the reasoning.