Beta Server Office Hours/Minutes/2013-09-05

From Second Life Wiki
Jump to: navigation, search

[2013/09/05 15:01]  Yuzuru Jewell: Hello, Maestro.
[2013/09/05 15:01]  Maestro Linden: Hi guys
[2013/09/05 15:01]  arton Rotaru: Rex, do you keep those logs, from other user groups as well?
[2013/09/05 15:01]  Inara Breen: Hi, Maestro
[2013/09/05 15:02]  Rex Cronon: well. it possible but i can't afford:(
[2013/09/05 15:02]  Yuzuru Jewell: Hello, Ardy.
[2013/09/05 15:02]  arton Rotaru: ah ok
[2013/09/05 15:02]  Ardy Lay: Hi
[2013/09/05 15:02]  Yuzuru Jewell: Hello, Baker.
[2013/09/05 15:02]  Rex Cronon: hi maestro
[2013/09/05 15:03]  Jenna Felton: hello all and all Lindens :)
[2013/09/05 15:03]  Yuzuru Jewell: Hello, Nalates.
[2013/09/05 15:03]  Baker Linden: Hello everyone!
[2013/09/05 15:03]  Maestro Linden: Okay, I think we have quorum
[2013/09/05 15:03]  Maestro Linden: Here's the agenda:
[2013/09/05 15:03]  Nalates Urriah: Hi Yuz
[2013/09/05 15:03]  Rex Cronon: hi baker
[2013/09/05 15:04]  Baker Linden: /afk
[2013/09/05 15:05]  Maestro Linden: So this week's releases were pretty straightforward, except for the timing (rolls were delayed a day because Monday was a holiday for us)
[2013/09/05 15:05]  Maestro Linden: The main channel got the project that was in all three RC's last week
[2013/09/05 15:06]  Maestro Linden: which.. we went over last week, mostly
[2013/09/05 15:06]  Rex Cronon: imo its a little illogic. labor day should be celebrated by woring hard:)
[2013/09/05 15:06]  arton Rotaru: lol
[2013/09/05 15:06]  Rex Cronon: working*
[2013/09/05 15:06]  Kennylex Luckless: Oi!
[2013/09/05 15:06]  Maestro Linden: I'd take that over going into labor..
[2013/09/05 15:07]  Lucia Nightfire: I read whoring for some reason
[2013/09/05 15:07]  Rex Cronon: bad lucia:)
[2013/09/05 15:07]  Jenna Felton: hehe
[2013/09/05 15:08]  Maestro Linden: all three RCs got monty's new project, which adds support for smarter asset fetching (which future viewers will use)
[2013/09/05 15:09]  Lucia Nightfire: where does that shine?
[2013/09/05 15:09]  Ardy Lay: Good.  We need smarter assets.
[2013/09/05 15:09]  Maestro Linden: Monty had some details here:
[2013/09/05 15:09]  Kennylex Luckless: No, not smarter, just faster.
[2013/09/05 15:09]  Kennylex Luckless: Smart is over rated.
[2013/09/05 15:10]  Lucia Nightfire: what is getmesh2? just another pipeline?
[2013/09/05 15:10]  Lucia Nightfire: or something else
[2013/09/05 15:10]  Maestro Linden: GetMesh2 is like GetMesh, but pipelined
[2013/09/05 15:10]  Maestro Linden: and it will use fewer connections
[2013/09/05 15:10]  Ardy Lay: Many consumer broadband routers will thank you.
[2013/09/05 15:10]  Rex Cronon: with getmesh2 u click on any mesh and download a local copy;)
[2013/09/05 15:11]  Maestro Linden: the throttling actions with 503 errors will mostly benefit high bandwidth, high-latency connections
[2013/09/05 15:11]  Lucia Nightfire: idk I've seen less agent/image/simulator timing in the RC's yet
[2013/09/05 15:11]  Lucia Nightfire: *idt
[2013/09/05 15:11]  Maestro Linden: Monty can explain things better
[2013/09/05 15:11]  Monty Linden: okay, dropped in in the middle
[2013/09/05 15:12]  Yuzuru Jewell: I want hear about it more.
[2013/09/05 15:12]  Maestro Linden: I just did a high level overview, from the release notes
[2013/09/05 15:12]  Monty Linden: new RCs have a GetMesh2 cap that's mostly equivalent to GetMesh
[2013/09/05 15:12]  Monty Linden: This is used to retrieve meshes.
[2013/09/05 15:12]  Monty Linden: For reasons lost in time, viewers have been using far, far
[2013/09/05 15:12]  Monty Linden: too many connections to do mesh downloads.
[2013/09/05 15:12]  Yuzuru Jewell: How about texture transfar.
[2013/09/05 15:13]  Monty Linden: 32 by default, extended over 500 by certain users
[2013/09/05 15:13]  Yuzuru Jewell: ?
[2013/09/05 15:13]  Monty Linden: completely unconstrained by certain debug/developer functions
[2013/09/05 15:13]  Lucia Nightfire: so what is getmesh and getmesh2 doing differently from each other? if anything
[2013/09/05 15:13]  Monty Linden: these have, umm, negatively impacted the experience for everyone.
[2013/09/05 15:13]  Rex Cronon: downloading 500 meshes at a time? who has a gigabyte interned connection?
[2013/09/05 15:13]  Monty Linden: GetMesh2 is part of fixing this.
[2013/09/05 15:13]  Ardy Lay: I do.
[2013/09/05 15:14]  Rex Cronon: wow
[2013/09/05 15:14]  Ardy Lay: But I don't think I am doing the abusing thing.
[2013/09/05 15:14]  Monty Linden: It includes:  default 8 connections per viewer plus keepalives plus better retry logic.
[2013/09/05 15:14]  Monty Linden: On the server side, GetMesh2 now is serviced with
[2013/09/05 15:14]  Monty Linden: keepalive connections and the viewer I'm currently using
[2013/09/05 15:14]  Monty Linden: makes use of it.
[2013/09/05 15:15]  Yuzuru Jewell: Can I set the count of connection? Only by server?
[2013/09/05 15:15]  arton Rotaru: so setting it to 8 connections in the current viewer will imporove things as well?
[2013/09/05 15:15]  Monty Linden: Not directly, not yet.
[2013/09/05 15:15]  Lucia Nightfire: was about to ask if it required a update in teh viewer to capitalize on, heh
[2013/09/05 15:15]  Monty Linden: New texture console is going to have a line of status
[2013/09/05 15:15]  Monty Linden: about mesh activities so you can see what it is doing.
[2013/09/05 15:15]  Inara Breen: Monty, is "get2Mesh" separate to the "Mesh2MaxCocurrentRequests" you referenced at the last TPV dev meeting, or one in the same?
[2013/09/05 15:16]  Maestro Linden: Yuzuru, for GetMesh it's configurable viewer-side via 'MeshMaxConcurrentRequests' debug setting
[2013/09/05 15:16]  Monty Linden: And there will be debug settings (Mesh2MaxConcurrentRequests or something like that)
[2013/09/05 15:16]  Yuzuru Jewell: THank you.
[2013/09/05 15:16]  Inara Breen: Ah... answered :)
[2013/09/05 15:16]  Monty Linden: GetMesh2 is the cap/URL
[2013/09/05 15:16]  Inara Breen: Yup, got it :)
[2013/09/05 15:16]  Monty Linden: And there are other tools to monitor what viewers have been
[2013/09/05 15:16]  Monty Linden: doing.
[2013/09/05 15:17]  Monty Linden: You can kick up a 'netstat' command and filter for port 12046 activity to see what is happening
[2013/09/05 15:17]  Monty Linden: on a viewer machine.  It can be quite shocking when you
[2013/09/05 15:17]  Monty Linden: TP to a mesh rich region.
[2013/09/05 15:17]  Monty Linden: Or if you use certain features in the develop menu.
[2013/09/05 15:18]  Monty Linden: One other interesting change is that some 503 responses
[2013/09/05 15:18]  Monty Linden: from our web services may include a 'Retry-After' header with a delta value (simple seconds).
[2013/09/05 15:18]  Rex Cronon: is the mesh downloading on different thread in the viewer? to me it looks like the viewer has one big loop and each time viewer downloads something the rendering is affected:(
[2013/09/05 15:18]  Monty Linden: Indicating a time at which it will be likely, but not guaranteed, for
[2013/09/05 15:18]  Monty Linden: a new request to succeed.
[2013/09/05 15:19]  Monty Linden: This applies to any request via cap subject to throttling including
[2013/09/05 15:19]  Monty Linden: http-in.
[2013/09/05 15:19]  Monty Linden: Though that's typically not the throttle of interest there.
[2013/09/05 15:19]  Jenna Felton: is there a way to tell a viewer, that opened too much connection, to close some of open conections?
[2013/09/05 15:20]  Monty Linden: Rex:  old and new viewers both did the downloads in a
[2013/09/05 15:20]  Monty Linden: separate thread but there is a lot of mutex locking in the code.
[2013/09/05 15:20]  Monty Linden: New viewer will have some interesting comments for those
[2013/09/05 15:20]  Monty Linden: who want to read.
[2013/09/05 15:20]  Ardy Lay: More cursing?
[2013/09/05 15:20]  Monty Linden: I believe there are less stalls with the new code which cleaned up much but we'll see.
[2013/09/05 15:21]  Monty Linden: Jenna:  in a way.
[2013/09/05 15:21]  Monty Linden: The extreme cases can't even get in to services, they'll simply reset when they try to connect.
[2013/09/05 15:21]  Jenna Felton: ok
[2013/09/05 15:21]  Monty Linden: Other connections will result in 503s and a connection closed by the server.
[2013/09/05 15:22]  Jenna Felton: so setting up the max number in debug setings will not harm the server then
[2013/09/05 15:22]  Monty Linden: Problem is that the original viewer coded a retry-as-fast-as-possible
[2013/09/05 15:22]  Monty Linden: policy which made things worse.
[2013/09/05 15:22]  Monty Linden: There is a pathological mesh case:
[2013/09/05 15:22]  Jenna Felton: ok
[2013/09/05 15:22]  Monty Linden: retries happen so fast that *all* are 503'd and you never
[2013/09/05 15:22]  Monty Linden: see any mesh progress.
[2013/09/05 15:22]  Monty Linden: I'm curious if anyone here has experienced this.
[2013/09/05 15:22]  Monty Linden: I can reproduce it fairly easily.
[2013/09/05 15:22]  Ardy Lay: Logging goes nuts too.
[2013/09/05 15:23]  Lucia Nightfire: I think many have implicitly seen it on many occasion
[2013/09/05 15:23]  Monty Linden: Yes, logging goes absolutely nuts.
[2013/09/05 15:23]  Jenna Felton: i not touched any debug settings for mesh... and i download mesh fast and fine :)
[2013/09/05 15:23]  arton Rotaru: I don't
[2013/09/05 15:23]  Ardy Lay: I have been changing logging level when that happens.
[2013/09/05 15:23]  Monty Linden: You'll have 10,000 'retrying'
[2013/09/05 15:23]  Nalates Urriah: I have seen sculpties and mesh that never rezzed. Come back later and all is fine.
[2013/09/05 15:23]  Monty Linden: messages from a single login/TP.
[2013/09/05 15:23]  Yuzuru Jewell: Is it expanded translating textures?
[2013/09/05 15:23]  Rex Cronon: i think i have seen it. each time i go to a sim the ping sim time in statistics window inside the viewer can go to over 9000
[2013/09/05 15:24]  Monty Linden: default settings are permissive enough to cause the failure
[2013/09/05 15:24]  Maestro Linden: Yuzuru: yes, the 503 & Retry-After behavior applies to HTTP texture fetching as well
[2013/09/05 15:24]  Monty Linden: I'm curious if large ping times make it worse or better....
[2013/09/05 15:25]  Lucia Nightfire: I see it mainly in heavy attach mesh regions that also have high object/texture/script count, but the latter may or may not be a factor.
[2013/09/05 15:25]  Rex Cronon: the is stuck in a mutex lock and that is why it show so high ping times
[2013/09/05 15:25]  Rex Cronon: the viewer is*
[2013/09/05 15:25]  Lucia Nightfire: ie high traffic sandboxes/malls/stores
[2013/09/05 15:25]  Monty Linden: the viewer is looking good and I'm trying to get it out
[2013/09/05 15:25]  Maestro Linden: We should be clear that the viewer which respects Retry-After or uses the new GetMesh2 capability is not yet available.  I believe the plan is to have a project viewer at some point?
[2013/09/05 15:26]  Monty Linden: as quickly as possible.
[2013/09/05 15:26]  Monty Linden: Also trying to get a better technical description out describing
[2013/09/05 15:26]  Monty Linden: what I've done and what I'd like to do.
[2013/09/05 15:26]  Rex Cronon: i heard UML diagrams can be quite useful:)
[2013/09/05 15:26]  Monty Linden: All of this is preparation for HTTP pipelining which should be a very nice
[2013/09/05 15:26]  Monty Linden: change, especially for high-ping-time locations (yuzuru).
[2013/09/05 15:27]  Monty Linden: what maestro said, project viewer or maybe straight to RC
[2013/09/05 15:27]  Monty Linden: not certain yet.
[2013/09/05 15:27]  Monty Linden: new comments in llmeshrepository actually include a simplified sequence diagram, so there, Rex
[2013/09/05 15:28]  Rex Cronon: cool:)
[2013/09/05 15:28]  Monty Linden: that's the main thing
[2013/09/05 15:28]  Ardy Lay: I don't suppose this is related:
[2013/09/05 15:28]  Ardy Lay: 2013-09-05T22:25:19Z WARNING: LLMotionController::purgeExcessMotions: ONCE (4900th time seen): > 64 Loaded Motions
[2013/09/05 15:28]  Monty Linden: one other small change is to ensure that final requests from services from capabilities
[2013/09/05 15:28]  Ardy Lay: Whatever that was has topped now.
[2013/09/05 15:29]  Monty Linden: are transmitted with 'chunked' encoding
[2013/09/05 15:29]  Monty Linden: the value of this is that it provides redundant length information
[2013/09/05 15:29]  Monty Linden: for reseponse data which allows libcurl and the viewer code
[2013/09/05 15:29]  Maestro Linden: Ardy: Hm, 'motions' sounds like an avatar animation issue
[2013/09/05 15:29]  Monty Linden: to validate that it really received a full response.
[2013/09/05 15:29]  Monty Linden: If someone has a bad ISP or a bad network driver, they
[2013/09/05 15:29]  Monty Linden: may see error messages indicating short data.
[2013/09/05 15:30]  Monty Linden: (partial file transfer, or range header invalid)
[2013/09/05 15:30]  Monty Linden: we'd like to hear about these but they likely indicate a
[2013/09/05 15:30]  Monty Linden: weakness in the networking path.
[2013/09/05 15:30]  Ardy Lay: "bad ISP"?  WHat might they be doing wrong?
[2013/09/05 15:31]  Lucia Nightfire: throttling, lol
[2013/09/05 15:31]  Rex Cronon: how can u get a partial file transfer when u r using http?
[2013/09/05 15:31]  Monty Linden: Ardy, I don't recognize the motions problem
[2013/09/05 15:31]  Maestro Linden: yeah, LLMotionController is avatar animation playback.. purely viewer-side
[2013/09/05 15:32]  Monty Linden: Rex, if actual data received on closed connection is shorter than that declared in content-length, you have a short response.
[2013/09/05 15:32]  Monty Linden: Ditto if a connection closes in the middle of a chunk without completing the chunk.
[2013/09/05 15:32]  Monty Linden: In fact, we have a jira where this happens.
[2013/09/05 15:33]  Monty Linden: that's basically my story
[2013/09/05 15:33]  Rex Cronon: i understand how it work, but in theory by using http it should prevent that from happening
[2013/09/05 15:33]  Monty Linden: Anything inserting a RST into the stream will do it.
[2013/09/05 15:34]  Monty Linden: Have seen bad HTTP caches that react badly to Range headers.
[2013/09/05 15:34]  Monty Linden: It's a network.  Things go wrong.
[2013/09/05 15:34]  Rex Cronon: u could be using udp and u could get same performance:)
[2013/09/05 15:35]  Lucia Nightfire: whats on the horizon, whats coming down teh pipe
[2013/09/05 15:36]  Ardy Lay: I have seen some service providers use buffered generic traffic shaping to delay TCP traffic instead of discarding it to reduce link utilization.  This does not work well at all.
[2013/09/05 15:36]  arton Rotaru: ribbon particles :p
[2013/09/05 15:36]  Lucia Nightfire: when is that exp thing gonna be presented
[2013/09/05 15:36]  Lucia Nightfire: and explained
[2013/09/05 15:38]  Ardy Lay: GTS is commonly the cause of really high round trip delays.
[2013/09/05 15:38]  Monty Linden: /me doesn't have the answers to those - not that he can say
[2013/09/05 15:39]  Maestro Linden: Experience tools?  That will probably be presented when it's close to release... I'm not sure where that is in the pipeline right now
[2013/09/05 15:39]  Ardy Lay: Properly engineered traffic management doesn't significantly delay traffic.
[2013/09/05 15:39]  arton Rotaru: anyway, thank you Monty for taking the time to explain it to us. I think it's all good work you are doing
[2013/09/05 15:39]  Maestro Linden: Yes, thanks for explaining the changes, Monty!
[2013/09/05 15:39]  Yuzuru Jewell: Thank you, Monty.
[2013/09/05 15:40]  Monty Linden: yw
[2013/09/05 15:40]  Maestro Linden: I can talk about the upcoming server maintenance project..
[2013/09/05 15:40]  Monty Linden: I should have a picture sometime but we lost the poster wall here in Morris
[2013/09/05 15:40]  Ardy Lay: It discards that which does not fit, usually making an effort to be fair about which flows lose a packet.
[2013/09/05 15:42]  Rex Cronon: monty. u need to write a press release right before group meeting and during meeting have people ask u questions about it:)
[2013/09/05 15:42]  Maestro Linden: So the upcoming maintenance project, which may go to RC next week, has several bugfix goodies:
[2013/09/05 15:42]  Lucia Nightfire: any bugs that have been around for years?
[2013/09/05 15:42]  Maestro Linden: fix for "llXorBase64 produce different result of llXorBase64StringsCorrect"  (llXorBase64 had an issue with password-string wrapping)
[2013/09/05 15:43]  Rex Cronon: :)
[2013/09/05 15:43]  Lucia Nightfire: thats just code change, not a function ya?
[2013/09/05 15:43]  Maestro Linden: it's a fix for that LSL function
[2013/09/05 15:44]  Maestro Linden: hm, there are some crash fixes, but I'm not sure how old the crash modes are
[2013/09/05 15:44]  Jenna Felton: with llXorBase4 we run slowly out of names
[2013/09/05 15:44]  Jenna Felton: llXorBase64
[2013/09/05 15:44]  Lucia Nightfire: great, so anyone sending code from RC and MC regions will have to have their target servers expect two different returns temporarily
[2013/09/05 15:44]  Maestro Linden: yeah, we won't make yet another XOR function :)
[2013/09/05 15:44]  Lucia Nightfire: idt I'll have to worry though
[2013/09/05 15:44]  Ardy Lay: Hey, let's fix the changed event again.
[2013/09/05 15:45]  Jenna Felton: whats wrong with the event?
[2013/09/05 15:45]  Lucia Nightfire: that the only bug fix?
[2013/09/05 15:45]  Maestro Linden: Kelly also fixed a few JSON function issues, for somewhat obscure edge cases: "JSON implementation treats blank string and JSON_NULL interchangably", "Issue with llJsonSetValue with strings that end in ']'", "RFC 4627 Non-Compliance by llJsonSetValue() with out-of-range Indices"
[2013/09/05 15:46]  Ardy Lay: It wasn't detecting face texture changes.  When that was fixed a "popular vendor" went nuts.
[2013/09/05 15:46]  Jenna Felton: oh
[2013/09/05 15:46]  Maestro Linden: Right, there was a changed() event bug that if you did llSetTexture() on a prim, CHANGED_TEXTURE wouldn't trigger
[2013/09/05 15:47]  Lucia Nightfire: so its working correctly now, lol
[2013/09/05 15:47]  Maestro Linden: we fixed the bug 2 or 3 years ago, but had to undo the fix since it broke a good portion of inworld vendors
[2013/09/05 15:47]  Ardy Lay: To && or to ||, that is the question.
[2013/09/05 15:47]  Maestro Linden: the vendor code had   ...  if(change | CHANGED_TEXTURE){llResetScript();}
[2013/09/05 15:47]  Maestro Linden: in the changed() event
[2013/09/05 15:47]  Lucia Nightfire: ...
[2013/09/05 15:48]  Maestro Linden: and the vendor would also change the texture each time you cycled to a new product..
[2013/09/05 15:48]  Jenna Felton: talking about changed event, is there a plan for filter event cases, so when your script not wants hear on teleport change for example, the teleport change can be filtered and the changed event handler is not bugged by unused changes
[2013/09/05 15:48]  Lucia Nightfire: idt I want to even comment on that
[2013/09/05 15:48]  Rex Cronon: i gues making a small lsl script change in the vendor was harder than rolling back the fix:)
[2013/09/05 15:48]  Maestro Linden: I'm not aware of any plans like that, Jenna, though I could see how that could be useful
[2013/09/05 15:49]  Maestro Linden: kind of like llCollisionFilter()
[2013/09/05 15:49]  Jenna Felton: yes :)
[2013/09/05 15:49]  Lucia Nightfire: llPassChanges()
[2013/09/05 15:49]  Ardy Lay: Mæstro, I have to wonder how many of those vendors are still out there.  Probably way too many.
[2013/09/05 15:49]  Jenna Felton: than that fix could be combined with the filter
[2013/09/05 15:50]  Maestro Linden: Yes, probably quite a few, Ardy..
[2013/09/05 15:50]  Lucia Nightfire: I would like certain changes to not be detected by the change event too, lol
[2013/09/05 15:50]  Maestro Linden: Well, I would recommend the scripter to change their bitwise OR to an AND, if they're editing the script :)
[2013/09/05 15:50]  Lucia Nightfire: lol
[2013/09/05 15:51]  Maestro Linden: another fix is "Issue with agents icons showing at 0,0 on the minimap when they are actually sitting at high altitude." - there's a weird case where another avatar would appear near the origin both in the minimap and in-world
[2013/09/05 15:51]  Jenna Felton: you can replace a script by a new one, if you detect it by uuid? like you replaced animations by replacements :)
[2013/09/05 15:51]  Lucia Nightfire: sounds familiar, heh
[2013/09/05 15:51]  Jenna Felton: aha
[2013/09/05 15:51]  Maestro Linden: heh, might be tricky to do a replacement like that on a live-running script
[2013/09/05 15:52]  Maestro Linden: and it would be really bad if we accidentally swizzled bits on the wrong scripts :)
[2013/09/05 15:53]  Jenna Felton: ok, was bad suggestion :)
[2013/09/05 15:53]  arton Rotaru: so it will never be fixed?
[2013/09/05 15:53]  Rex Cronon: talking about avatars showing at weird locations. i find it very weird that sometimes when i am killed on damage enabled sim, instead of being sent home i end in a weird place from where i can't to, nor IM others and i have to logoff:(
[2013/09/05 15:53]  Ardy Lay: Tricky as in "Tricky.  I'll have to think about it." -- Deep Thought - Hitchhiker's Guide to the Galaxy
[2013/09/05 15:54]  Rex Cronon: i can't tp*
[2013/09/05 15:54]  Jenna Felton: Rex, you was killed a coma
[2013/09/05 15:54]  Lucia Nightfire: he was killed in Rausch where anything can happen
[2013/09/05 15:54]  Jenna Felton: not dead but not alive :)
[2013/09/05 15:54]  Rex Cronon: no coma. i can type and what i type shows on screen
[2013/09/05 15:54]  Maestro Linden: arton, I wouldn't promise "never", but I don't see that changing unless priorities suddenly change
[2013/09/05 15:54]  arton Rotaru: hm odd, but hey...
[2013/09/05 15:55]  Jenna Felton: perhaps it is possible to send an owner of a script with certain UUID a message via email talking to replace the vendor or correct the bug
[2013/09/05 15:57]  Ardy Lay: The people using the script in discussion here are not the author and likely do not have access to the source.
[2013/09/05 15:57]  Maestro Linden: Finally, there's a fix to close an exploit in which certain types of objects made themselves immune from autoreturn.  It looks like the kind of method that a griefer would use for annoying spam boxes, so that should be a welcome fix..
[2013/09/05 15:57]  Ardy Lay: Yes!
[2013/09/05 15:57]  Rex Cronon: self-replicators maestro?
[2013/09/05 15:57]  Lucia Nightfire: what was the final decision on thresholds there?
[2013/09/05 15:57]  Lucia Nightfire: I'm assuming you're referrign to self-replication
[2013/09/05 15:57]  Lucia Nightfire: although there are 5-6 ways to defeat autoreturn easily
[2013/09/05 15:57]  Maestro Linden: Was that the one Andrew discussed?
[2013/09/05 15:58]  Lucia Nightfire: ya
[2013/09/05 15:58]  Maestro Linden: Okay, then yes
[2013/09/05 15:58]  Rex Cronon: he talked about it last week
[2013/09/05 15:58]  Lucia Nightfire: is it time limited?
[2013/09/05 15:58]  Lucia Nightfire: like, deny accepting offered inv w/i # minutes of creation or what?
[2013/09/05 15:58]  Maestro Linden: /me asks Andrew
[2013/09/05 15:59]  Lucia Nightfire: I've been seeing him in and out for awhile
[2013/09/05 15:59]  arton Rotaru: we need to know how the fix work, to work around it :p
[2013/09/05 15:59]  Rex Cronon: lol arton
[2013/09/05 16:01]  Rex Cronon: did andrew talk about replicators this week too? i couldn't make it and last i checked the transcript wasn't available
[2013/09/05 16:01]  Maestro Linden: Okay, "the decision was to nerf the autoreturn bypass technology completely -- since it violates the spirit of the parcel settings"
[2013/09/05 16:01]  Rex Cronon: the important thing in that quote is the word "was" :)
[2013/09/05 16:01]  Lucia Nightfire: so what specifically is being blocked? objects accepting scripted inventory offers?
[2013/09/05 16:02]  Baker Linden: I want to try to finish this last bit I wanted to get done today, so I'm out. Have a fantastic weekend everyone!
[2013/09/05 16:02]  Maestro Linden: If ObjectA rezzes ObjectB, ObjectB will inherit the temp-on-rez and parcel time of ObjectA
[2013/09/05 16:02]  Rex Cronon: tc baker
[2013/09/05 16:02]  Lucia Nightfire: or blocking scripted rezzing of what was recieved
[2013/09/05 16:02]  Jenna Felton: ah
[2013/09/05 16:02]  arton Rotaru: by baker
[2013/09/05 16:02]  Maestro Linden: So both ObjectA and ObjectB would be returned to the owner simultaneously, if autoreturn was in effect
[2013/09/05 16:02]  Ardy Lay: Good.
[2013/09/05 16:02]  Jenna Felton: i know someone who would be not very happy by this fis
[2013/09/05 16:02]  Lucia Nightfire: ok, thats great
[2013/09/05 16:02]  Jenna Felton: fix
[2013/09/05 16:03]  Lucia Nightfire: now you need to address the other 5 ways to defeat atuoreturn, lol
[2013/09/05 16:03]  Jenna Felton: but it sounds reasonable
[2013/09/05 16:03]  Rex Cronon: nice. that kills my replicating working-platform:(
[2013/09/05 16:03]  Lucia Nightfire: it makes sense to nerf it anyway Rex
[2013/09/05 16:03]  Jenna Felton: now i know one more
[2013/09/05 16:03]  Ardy Lay: Rex, the ground is not subject to return.
[2013/09/05 16:04]  Maestro Linden: could be funny at times if it were :)
[2013/09/05 16:04]  Lucia Nightfire: if you're going that way you can nerf at least 3 other means that rely on "rezzed time"
[2013/09/05 16:04]  Rex Cronon: unless u can make me a floating island at 2km it useless to me
[2013/09/05 16:04]  Ardy Lay: WHy 2k up?
[2013/09/05 16:04]  Lucia Nightfire: or rather rezzed time resetting
[2013/09/05 16:05]  Rex Cronon: usuallly neither people nor bullets fly that high:)
[2013/09/05 16:05]  Maestro Linden: Okay, we're out of time
[2013/09/05 16:05]  Maestro Linden: slightly past, even
[2013/09/05 16:05]  arton Rotaru: yup, thanks for the meeting
[2013/09/05 16:06]  Ardy Lay: There are some little-used sandboxes you can move to if you are in one with undesirable activities in it.
[2013/09/05 16:06]  Rex Cronon: stopping loggin
[2013/09/05 16:06]  Maestro Linden: time to get back to testing - thanks for coming, everybody!
[2013/09/05 16:06]  Ardy Lay: Thanks!
[2013/09/05 16:06]  Rex Cronon: tc maestro
[2013/09/05 16:06]  Lucia Nightfire: laters
[2013/09/05 16:06]  Yuzuru Jewell: Thank you, Maestro.
[2013/09/05 16:06]  Inara Breen: Thanks, Maestro, Monty
[2013/09/05 16:06]  Nalates Urriah: Thx Lindens
[2013/09/05 16:06]  Maestro Linden: bye guys!
[2013/09/05 16:06]  Ardy Lay: Tink!
[2013/09/05 16:06]  Monty Linden: g'day
[2013/09/05 16:06]  Jenna Felton: have a great weekend all of you