This is a difficult one, over at SLUniverse Bronxelf has posted that Linden Lab are going to fix an issue that allows people to find out someone else’s online status, this is detailed in Jira SVC-4823. I agree that someone’s online status should be what the person themself set it to, so if you set your setting to friends and groups only, that’s what should be respected. People would like to see this go further and have an invisible to everyone option, currently you seem to need to do this individually for your friends, but we’re looking at the privacy settings in preferences such as this:
The simple fact of the matter is that your privacy settings can easily be made useless by some simple scripting and people, rightly don’t like this. I agree with Bronxelf on this issue, the settings should be resepected, if you want to appear invisible, you should. Linden Lab agree too, but the solution is now creating a great deal of gnashing of teeth, because the solution will break content if Linden Lab go ahead with it, they might not yet go ahead with it.
The proposed solution is to only allow people to find the online status of themself, therefore if you have a board that has an online status indicator inworld, you need to be the owner or creator of the board for the call to the Linden Lab servers to reveal your true online status. This rightly has content creators howling because there are several decent use cases for finding someone’s true online status, which aren’t designed to subvert privacy settings but to deal with issues of llGiveInventory() not working when messages are capped and therefore people missing out on deliveries.
The decent use cases for being able to find someone’s true online status have caused Linden Lab to pause, even if it’s only temporarily, Oz Linden stating in the Jira:
“The feedback here is very useful. Thank you to those who are providing descriptions of use cases that don’t unduly violate the privacy of the subject of this LSL method.
We will discuss this further and provide updates, and will try to do so with enough time for existing usages to be adapted if needed.”
Is this a sign of Oz following through with his claims of better collaboration? Maybe, I’ve actually never found Oz to be bad to be honest, although I have seen complaints about his attitude at times, he always seems willing to engage, sometimes he’s maybe a bit too forthright as a customer facing Linden but I don’t think that’s always a bad thing.
The solution to this issue requires Linden Lab to think systemically. The reason there are decent use cases for finding someone’s online status is because the system doesn’t cope with capped messages, they need to look at the whole issue here, they are right to want to respect people’s privacy wishes and they should ensure that TPV’s remove true online status from their tools as that isn’t really about scripted calls but they should most definitely pay heed to the legitimate uses of this tool and address why those legitimate uses exist.
We do have to accept that over time content may break in Second Life, that scripted functions will be changed, or even fixed, I can recall quite a while ago Linden Lab fixing something that made people howl, but Linden Lab pointed out that it hadn’t been working as intended. However how and why content is broken needs to be taken seriously, in this case the initial idea has good intentions but has unearthed some wider issues and hopefully those wider issues will be addressed.
Personally I think this calls for a two pronged approach, one to deal with the capped messages issue, which is a much larger and longer term issue and one to deal with the call only revealing someone’s status as they set it, if someone misses out on deliveries because of how they set their status, that’s their issue, we used to be told to ensure we were logged in to order items from Xstreet SL, so it’s not as if people are incapable of understanding the issue, game makers, online staff board makers etc. could send out updates telling people that their online status could impact performance of their tools, that would be a far better solution than simply breaking lots of existing content.
Key thing here is for LL to make their intentions well-known in advance so that people affected can make the necessary steps to minimise impact.
From listening to the Viewer Dev meeting, I get the impression the functionality of llRequestAgentStatus isn’t going to happen immediately – certainly not within the next 2+ weeks. There was also the implication that LL presume that issues relating to deliveries failed due to capped IMs would be bypassed as the new Received Items / Direct Delivery mechanism(s) will “make it go away”.
Of course, whether LL are time-tabling the changes to llRequestAgentStatus and the intro of the Received Items section of inventory such that the former rolls out ahead of the latter and has time to propagate through all TPVs is open to debate.
Really, though, the issue does come down to communication to their wider audience as a whole. Not everyone watches every JIRA or gets to attend all the relevant meetings where these things are discussed.
Communications, again. Testers have confirmed that items do still get capped during their testing of received items beta, so I’m not sure the issue will be fixed by that.
TPV’s don’t need to display true online status, I’m with Oz on that point but more and more people are posting use cases on the Jira, from mailing lists, to card games being impacted by this proposed change.