The following notes are primarily taken from the TPV Developer (TPVD) meeting held on Friday, May 15th, and from the Server Beta meeting held on Thursday, May 14th. A video of the TPVD meeting is included below, with any time stamps in the following text referring to it. My thanks as always to North for the recording and providing it for embedding,
Server Deployments, Week 20 – Recap
As always, please refer to the server deployment thread for the latest updates / news.
There was no Main (SLS) channel deployment on Tuesday, May 11th. On Wednesday, May 12th, the three RC channels were all updated with a new server maintenance package, comprising internal server logging changes, back-end system bug fixes, reply-to email changed in postcard sends (see below for more).
The Server Beta User Group meeting on Thursday, May 14th, saw a test of a group chat update it is hoped will fix the issue of some people not seeing all or some chat when the group chat window is open (see BUG-9130). The test appeared to yield positive results, with Simon reporting no unusual event logging. The problem here is that the instances of the problem seem to be so rare, it’s hard to guarantee a small sampling of testers will catch any problems which might still exist.
[00:15] It has been a quiet week on the viewer front. As noted in part 1 of this week’s report, the attachment viewer (Project big Bird) reached RC status, other than that there has been not additional movement with either the current selection of RC and project viewers.
A further update to the mesh importer project viewer (currently at version 184.108.40.2060878) is with LL’s QA team and should be released relatively soon. Updates are also being made to the Viewer-Managed Marketplace project viewer (which is likely to go to RC status once through LL’s QA process) and the Oculus Rift project viewer.
Snapshots to E-mail
[02:22] Commenting at the TPV Developer meeting, Oz Linden gave a little more information on the “reply-to email changed in postcard sends” update deployed to the RCs, indicating this was indeed a fix aimed at preventing snapshot to e-mail being tagged as spam by ISPs and a/v software due to the way they handle the “from” field (see my TPV Developer meeting report for week #17), which had caused the Lab to consider removing the snapshot to e-mail capability.
“Instead,” Oz told the meeting, “we found we could get around that by sending the e-mail differently … So the way it’s changed now is that instead of sending the ‘from’ address as the sender’s address, we send the ‘reply-to’ address; and the ‘from’ address is ‘email@example.com’ … so that ducks the problem of us looking like spammers who are forging invalid addresses.”
This should hopefully negate any need to remove the snapshot to e-mail capability, and retain compatibility with sending snapshots to the likes of Snapzilla and the SLU forums.
Unified Snapshot Floater
[04:40] NiranV Dean, who submitted the unified snapshot floater to LL (and which has most recently been integrated into Firestorm among the TPVs) asked if there had been any feedback on it. Both Oz and Grumpity Linden indicated that overall, feedback has been positive, although some have complained at the amount of screen real estate it takes up with the preview panel open. As this allows the snapshot preview to match the aspect ratio of the user’s screen, there’s not a lot that can be done about it – and the preview window can always be closed / the panel minimised when initially setting-up shots.
Niran is proposing a further set of updates (one of which, a fix for auto snapshot, is in the works at the Lab), including possibly making the preview screen detachable from the main floater. Cinder Roxley also indicated she is working on fully integrating the Facebook, Twitter and Flickr options into the main snapshot floater (they currently retain their own floaters due to the authentication workflow required for each. This work will be contributed to the Lab for consideration / integration when complete.
Viewer-Managed Marketplace (VMM)
[08:09] Over the last two weeks, as a part of the on-going beta of VMM, around 15-20 volunteers have had their stores migrated by the Lab from Direct Delivery to VMM. Brooke linden reports that the exercise has uncovered “pretty much minor issues” which the Lab can address. A further batch of volunteer migrations is planned to help further test the robustness of the process in the next week or so.
As noted above, the VMM viewer is now heading for an RC release once it has cleared LL’s QA testing. However, the time frame on when this might happen is a little vague; it might be in the next week or so, or it might be longer.
[17:34] The next Snowstorm contributions viewer is progressing internally at the Lab. This is the viewer which includes the new Avatar Complexity (aka “Jelly Babies” or “rainbow avatars”) functionality which allows users to define a level of complexity (a weighting number) which will render any avatar exceeding that value as a solid colour, rather than a full avatar. The aim of this is to help reduce the rendering load placed on people’s computers, particularly in very busy locations. The value is adjustable, as so can of course be varied to suit your current needs.
A slight hiccup has occurred in that in making some changes to the code, Oz accidentally broke the code such that instead of rendering as a solid colour, avatars exceeding the limit are currently rendering as transparent, and this is yet to be fixed. Code has been added to the viewer to report how many people around you are rendering your avatar as a solid colour (should your avatar be complex enough to be rendered thus), but this has yet to be made visible through the viewer UI, and simulator support for this is now in place on the RC channels and will be rolled to the Main channel in the coming week.
[19:36] Cinder Roxley has a set of contributions for using the viewer with Mac Retina displays ready to go to the Lab, and it seems likely these will flow into a further Snowstorm contributions viewer in development alongside the one containing the Avatar Complexity updates.
A question was also asked whether there were plans to update the Mac viewer to use a newer OpenGL core profile. The Lab is not working on this, as their rendering team believe there is little or no benefit to be gained from it. However, they would accept any contributions offered for consideration (subject to a “long, terrible QA process”). However, a good part of this would require working through some eight years of OpenGL code.