Code change impacts RLV functionality

Update 12th March: As can be seen from the comment from Trinity below, Brooke Linden has responded to concerns over this issue, and has confirmed that the code causing it will be rolled-back from LeTigre and BlueSteel this Wednesday (RC channel release window) and won’t be re-deployed until the problem is fixed.

Kitty Barnett reports via JIRA SVC-7748, that functionality related to the InventoryAPI maintenance project adversely impacts the widely used RLV / RLVa functionality within Second Life.

RLV provides a means by which, and under controlled conditions (the user “opts-in” to the process by clicking an acceptance button), a folder is created within the #RLV folder under MY INVENTORY. Items are then delivered into the new folder, wherein a script runs to attach the items to the recipient avatar.

While this functionality does have a direct use within the BDSM community, it can have uses elsewhere as well.  However, changes rolled-out to the BlueSteel and LeTigre RC channels this week as a part of the InventoryAPI maintenance project, have inadvertently broken the functionality – the required redirection to use #RLV doesn’t occur and the associated script fails – hence JIRA SVC-7748.

The degree of impact on RLV is debatable. As Marine Kelley states within the JIRA:

On a positive note, if LL decides not to do anything and leave things as is (i.e. in a broken state), the RLV could simply check what’s coming into the “Received Items” folder and move it automatically under #RLV if the name matches. This would be transparent to the user and would overcome this breakage. 

Nevertheless, it would be preferable for LL to ensure the functionality isn’t broken in the first place (as Marine herself goes on to state).

A potential problem here is that, despite Kitty’s own efforts to point out that Received Items itself is not the problem per se, many of the comments appearing on the JIRA are further critiques of Received Items rather than a discussion of the problem as identified by the JIRA itself.

As strong as feelings are around the subject of Received Items, what is more important here is that functionality that is key to a range of user expectations / desired experiences has been inadvertently broken within LeTigre and BlueSteel, and there is a risk that this could become more widespread if the fix is rolled-out beyond these two RC channels. As such, it is important that LL hear, read and understand the core issue itself (i.e. via use-cases where the update breaks things), in order for them to try to correct the matter.

Given it is the weekend, it will likely be a while longer before any response on this matter is heard from LL – which also gives people more time to submit specific examples on the issue that outline the problem. It’s also worthwhile pointing out that LL are prepared to reconsider proposed actions – as has been demonstrated around the concern relating to llGetAgentStatus (which Oz has indicated is on-hold as a result of the number of clear-cut use-cases received), and have shown a willingness to re-think elements of Received Items based on constructive feedback from users.

9 thoughts on “Code change impacts RLV functionality

  1. Thank you Inara.

    Following some discussion in RLV related groups in world I would just like to stress that this is an incidental breakage, no one is for a second suggesting it is intentional, and that the area of RLV that is impacted is limited to objects inserted in to the #RLV folder.

    Fixing this in the viewer can be done, However getting everyone to update to the latest RLV is no easy feat. It would be much better if the server side functionality that was broken is reverted.

    Please please please do not hijack this issue to beat the lab about the head over the new received items folder. UNHELPFUL.

    Suggesting that RLV is being killed, removed, broken on purpose is again UNHELPFUL.

    If you comment on the JIRA please keep it on civil and related to this one very specific issue.

    Thank you 🙂

    Like

    1. You’re welcome 🙂

      I was very tempted to put the “inadvertently” in the article in italics to add additional emphasis on the the fact the code-change is more accidental than deliberate.

      The important thing is – as we both say – to get LL to understand the issue, not the degree of upset, and for them to implement changes in such a way that, hopefully, the behaviour isn’t changed.

      Like

    1. Unless you have evidence to support that claim, it’s little more than hearsay, and something that isn’t helpful towards Lab / user relationships.

      While I do berate the Lab for its poor level of corporate-level communications, accusations of this nature tend to encourage a more confrontational relationship between the user community and the Lab as they give rise to speculation and rumour.

      Like

      1. No I didn’t have proof and I was wrong Inara, it wasn’t deliberate it was just plain clumsy, and they have had the grace to admit it and they are rolling it back according to Brooke.

        Like

        1. Yup. this blog post has been updated to reflect the situation, thanks to word from Trinity, and a link to Brooke’s JIRA comment is included in the update.

          Like

Comments are closed.