User:Babbage Linden/Office Hours/2009 07 15

From Second Life Wiki
Jump to navigation Jump to search

Transcript of Babbage Linden's office hours:

[3:08] Nock Forager: Here he come.
[3:08] Nock Forager: Hi Babbage
[3:08] Phantom Ninetails: There he is, hi Babbage
[3:09] Xugu Madison: Hey Babbage *ambush* :)
[3:09] Morgaine Dinova: Woohoo, it's Babbage!
[3:09] Babbage Linden: hi everyone
[3:09] Imaze Rhiano: 9 mins
[3:09] Babbage Linden: sorry i'm 9 minutes late
[3:09] Phantom Ninetails: Nice seat, babbage
[3:09] Sebastean Steamweaver: Hey Babbage :)
[3:09] Babbage Linden: *squelch*
[3:09] Morgaine Dinova: 'Morning Babbage :-)
[3:09] Qie Niangao: Hi Babbage. :)
[3:09] Phantom Ninetails: lol
[3:09] Xugu Madison: That's messing with the unmessed with parts of my brain. I was saving them for stuff
[3:09] Sebastean Steamweaver: Strange question. was Babbage named after you, Babbage?
[3:09] Babbage Linden: how's it going everyone?
[3:10] Sebastean Steamweaver: Seeing as how it's a steampunk sim, and you're dressed in a rather steampunk fashion, and it has your name.
[3:10] Xugu Madison: I've been doing error handling in LSL, and my brain hurts. Other than that, good
[3:10] Imaze Rhiano: pretty well
[3:10] Babbage Linden: i think new babbage was named after the same babbage that i was
[3:10] Babbage Linden: inventor guy
[3:10] Babbage Linden: played with cogs a lot
[3:10] Babbage Linden: burned workshops down
[3:10] Sebastean Steamweaver: Yay!
[3:10] Morgaine Dinova: Babbage: going fine thanks. At least it *was* fine, until I listened to Tits of Death from your Myspace page ^_^
[3:10] Xugu Madison: I have a flatmate like that
[3:11] Imaze Rhiano: awww.... I hope that you don't burn SL down
[3:11] Babbage Linden: heh
[3:11] Phantom Ninetails: People use Myspace?
[3:11] Babbage Linden: the tits of death were awesome
[3:11] Babbage Linden: bands use myspace still
[3:11] Babbage Linden: that's about it
[3:11] Morgaine Dinova: Yeah, bands use it a lot
[3:11] Phantom Ninetails: That saddens me
[3:11] Xugu Madison: Anyone got an agenda we should stick to, or can I wade on in with script memory quotas?
[3:12] Sebastean Steamweaver:
[3:12] Babbage Linden: want me to do an update
[3:12] Morgaine Dinova: Please!
[3:12] Babbage Linden: then we can talk about script memory quotas
[3:12] Phantom Ninetails: Ooh that's a good one to mention here, Sebastean.
[3:12] Xugu Madison: Update good!
[3:12] Morgaine Dinova: Super!
[3:12] Babbage Linden: anything else people want to talk about today?
[3:12] Sebastean Steamweaver: Just that one :)
[3:12] Sebastean Steamweaver: Lemme post it again, ,I wasn't wearing my JIRA helper
[3:12] Sebastean Steamweaver: WEll, I'll post it after update.
[3:13] Qie Niangao: List of Linden-Confirmed Easy "Featurettes," With Large Returns
[3:13] Xugu Madison: (Need a llGetLinkPrimitiveParams function) would be a good one to bring up later
[3:13] JIRA-helper: [#SVC-224] Need a llGetLinkPrimitiveParams function
[3:14] Babbage Linden: ok, sounds good
[3:14] Babbage Linden: update:
[3:14] Babbage Linden: i've spent some time working on 1.27 issues this week
[3:14] Babbage Linden: the 1.27 simulator is using the new region presence service
[3:14] Imaze Rhiano: what is Jira entry for that "Featurettes"?
[3:15] Babbage Linden: instead of the space server
[3:15] Sebastean Steamweaver:
[3:15] JIRA-helper: [#MISC-3077] List of Linden-Confirmed Easy "Featurettes," With Large Returns
[3:15] Babbage Linden: which is great news
[3:15] Phantom Ninetails: Region presence service?
[3:15] Babbage Linden: as we're getting rid of the last single process service in SL
[3:15] Morgaine Dinova: Aye, Peri mentioned that.
[3:16] Babbage Linden: the region presence service allows you to ask which simulator is running which region
[3:16] Babbage Linden: simulators use it to find their neighbours
[3:16] Babbage Linden: to coordinate l$ transfer notifications with remote avatars
[3:16] Phantom Ninetails: Neat
[3:16] Babbage Linden: to send estate messages
[3:17] Qie Niangao: oh, please please please let there be a scripting interface to that, to find the global x,y of an arbitrary sim
[3:17] Babbage Linden: and to get the number of avatars in a region
[3:17] Gunter Vandyke: Hi all
[3:17] Babbage Linden: that last one was causing problems last week
[3:17] Phantom Ninetails: Interesting
[3:17] Babbage Linden: as the viewer requests information for each region one at a time
[3:17] Sebastean Steamweaver: Would this enable a grid-wide llGiveInventoryList?
[3:17] Babbage Linden: when you open the map
[3:18] Babbage Linden: previously the sim just forwarded the udp message to the space server
[3:18] Babbage Linden: which had an in memory database of all the grid activity
[3:18] Babbage Linden: with the region presence service, the sim does a region presence service look up
[3:18] Babbage Linden: to find the process running the region
[3:19] Babbage Linden: and then asks the region how many people are in the region
[3:19] Babbage Linden: so, 2 XML over HTTP lookups instead of one UDP packet forward
[3:19] Babbage Linden: which we found to be roughly 50x more expensive
[3:19] Babbage Linden: and caused the sim to slow down
[3:20] Phantom Ninetails: Crazy
[3:20] Imaze Rhiano: UDP was 50x more expensive?
[3:20] Babbage Linden: so, for the interim, we've rebranded the space server as the map item request aggregation service ;-)
[3:20] Phantom Ninetails: Was that what caused the sim lag when you zoomed into the map then?
[3:20] Sebastean Steamweaver: So that's what Vektor was having us test at QA hours with the map thing.
[3:20] Babbage Linden: map item requests still go to the space server (map item aggregration server)
[3:21] Babbage Linden: everything else goes to the region presence service
[3:21] Babbage Linden: which means that if the space server is overwhelmed
[3:21] Babbage Linden: only map requests are affected, not region connectivity
[3:21] Babbage Linden: so, SL should work better
[3:21] Phantom Ninetails: Neat
[3:21] Norton Burns: nice
[3:21] Xugu Madison: Good to see critical and non-critical stuff being isolated :)
[3:22] Babbage Linden: i think what happened is that a "bug fix" allowed the viewer to show green dots at wider zooms
[3:22] Babbage Linden: which caused more load on the space server
[3:22] Babbage Linden: which caused region connectivity problems recently
[3:22] Babbage Linden: that shouldn't happen any more
[3:22] Phantom Ninetails: Ah
[3:22] Babbage Linden: longer term, we need to look at the way map item requests work
[3:22] Qie Niangao: ("anymore" meaning after 1.27 server, I assume)
[3:23] Babbage Linden: to make them more scalable
[3:23] Phantom Ninetails: What's a map item request
[3:23] Babbage Linden: yes
[3:23] Babbage Linden: (having a feature which allows you to ask a question to all regions in a theoretically infinite grid, is not exactly scalable)
[3:23] Babbage Linden: we may have to cache the data longer
[3:24] Babbage Linden: or have it display an average region activity
[3:24] Babbage Linden: and treat it like another map layer
[3:24] Babbage Linden: or make it so that you can only see regions around your current region
[3:24] Babbage Linden: to make it scale geographically
[3:24] Babbage Linden: anyway, 1.27 should improve the situation in the mean time while we work that out
[3:25] Gunter Vandyke: Babbage, you know there is still abug in1.27?
[3:25] Gunter Vandyke: i just restart a 1.27 region, after its back you cant enter it from eighbour region without a crash
[3:25] Babbage Linden: there were a couple of show stoppers in 1.27 last I heard
[3:25] Babbage Linden: which were being worked on
[3:26] Babbage Linden: kelly was looking at http-in headers
[3:26] Babbage Linden: wow, that sounds like a big one gunter that i hadn't heard off
[3:26] Babbage Linden: is there a JIRA?
[3:26] Xugu Madison: Which is a point, we ought to have easily findable main-grid version servers on preview, to test hopping across with. Also, Babbage, if I take a script that uses http-in and rez it on 1.26 sim, what happens?
[3:27] Babbage Linden: if it's mono you will likely get a crash with a .NET stack trace
[3:27] Gunter Vandyke: the sim comes back, suddenly loose the surface, if you cross border you may crash, if you tp in you stand on water, if you relog your avatar all is fine
[3:27] Babbage Linden: with LSL i'm not sure
[3:27] Babbage Linden: i'll ask kelly
[3:28] Imaze Rhiano: Any news about script memory limits?
[3:28] Gunter Vandyke: was on first 1.27 and the new one too
[3:28] Babbage Linden: gunter the viewer crashes, or the sim?
[3:28] Gunter Vandyke: the viwer
[3:28] Babbage Linden: gunter, if there isn't a JIRA, please file one
[3:28] Norton Burns: I was about to ask about memory limits too
[3:28] Sebastean Steamweaver: Eesh,, that was wierd.
[3:29] Gunter Vandyke: if you not crash you see no surface and have to relog
[3:29] Sebastean Steamweaver: Uh oh, I think I just TPed stick into the water.
[3:29] Babbage Linden: so, script limits
[3:29] Xugu Madison: Oh, yes, script limits. Please, please, PLEASE include bytecode sharing in the usage stats. I'm about to put out two new scripts at 40kb and 45kb each, in bytecode, and a couple of kb of data. They'll get pummelled if bytecode sharing isn't taken into account
[3:29] Stickman Ingmann: YOU DID. :(
[3:29] Sebastean Steamweaver: Sorry Stick
[3:29] Babbage Linden: we're on the 2nd and last week of our script limits information sprint
[3:29] Stickman Ingmann: I'll never forgive you.
[3:29] Norton Burns: copy/paste - sorry...
[3:29] Norton Burns: I have been working towards monolithic scripting in my weapons, every weapon containing the same script, a self-read system that grabs parameters for each individual weapon in a short list...
After the new limits are imposed, how will this affect large numbers of identical scripts running simultaneously in a sim? Strictly, the memory footprint could be quite large - but all are running the same script, though outwardly different behaviours.
[3:30] Gunter Vandyke: no surface is there, you stand on it, but not visible
[3:30] Techwolf Lupindo: I just found out about this one.
[3:30] Babbage Linden: we're adding capabilties to the server that will allow viewers to request information about scripts
[3:30] Babbage Linden: everyone will be able to see used and available script resources
[3:30] Morgaine Dinova: That's cool!
[3:31] Babbage Linden: parcel owners will be able to see a list of scripted objects in their parcel with usage
[3:31] Qie Niangao: We need that urgently--for at least six months before any limits are in effect.
[3:31] Liandra Ceawlin: Excellent. :o
[3:31] Babbage Linden: and be able to select them to show with beacons and return them
[3:31] Babbage Linden: estate owners will be able to see the same for all scripts in the region
[3:31] Norton Burns: for the slightly less-techy amongst us, will 2 mono scripts have only one footprint, or 2, as far as totals are concerned?
[3:31] Norton Burns: (identical, I mean)
[3:32] Babbage Linden: residents will also be able to their attachment script usage and availability
[3:32] Babbage Linden: with a break down by attachment
[3:32] Phantom Ninetails: I want to see region script memory usage in the statistics floater still
[3:32] Kephra Nurmi: able to see usage = requires client update?
[3:32] Techwolf Lupindo: Good. Then end user/customers can show the shops that there scripts are not so "low-lag" as they advertized. :-)
[3:33] Babbage Linden: we're getting this work done now so it can be in server 1.30
[3:33] Babbage Linden: (which is the next big release after 1.27)
[3:33] Liandra Ceawlin: I spologize if this had already been gone over, but regarding the script memory limits: When that goes into effect, will we be able to like say "This script needs 256k", and then exceed the current 64k limit and not have to deal with script splitting foolishness?
[3:33] Liandra Ceawlin: *apologize, too. :O
[3:33] Babbage Linden: so it can be used be the redesigned viewer that's being worked on now
[3:33] Babbage Linden: liandra, yes that's the plan
[3:33] Liandra Ceawlin: Wewt! ^_^
[3:34] Babbage Linden: mono scripts will be able to request different memory reservations
[3:34] Norton Burns: low-lag does not necessarily mean low memory, though, does it? That's why it's important to know if 2 avs wearing the same script will show twice in the memory
[3:34] Babbage Linden: as with URLs for HTTP-In
[3:34] Morgaine Dinova: Babbage: I'm not so sure it's a good idea to provide that public info through caps though. That's forcing people to use login+TP bots to gather the info, which is a heavy load. If info is public and not sensitive, it's generally best (for everyone) to provide it over a plain web service.
[3:34] Babbage Linden: LSL scripts will stuck with 16K, whatever they request
[3:34] Babbage Linden: morgaine, i agree
[3:35] Phantom Ninetails: I want to see region script memory usage in the statistics floater still
[3:35] Stickman Ingmann: I heard the subject touched on earlier, and I had a question about how script memory is and will be. Is there any current condition where two scripts with the same code only load their bytecode into memory once, and will such a condition exist when "script limits" are put into place?
[3:35] Babbage Linden: but exposing services without caps is a bigger, architectural change
[3:35] Babbage Linden: and it gets in to SL API territory
[3:35] Norton Burns: that's what I need to know Stickman - my whole future development is based on that answer
[3:35] Babbage Linden: which is another big issue
[3:36] Xugu Madison: Stickman, (and Norton from earlier). Scripts with the same asset UUID running on the same sim will share bytecode. Not sure how that will work with quotas, though...
[3:36] Babbage Linden: xugu, correct
[3:36] Babbage Linden: it helps to have multiple scripts sharing UUIDs for mono scripts
[3:36] Phantom Ninetails: I must be the only person here who uses the statistics floater.
[3:37] Techwolf Lupindo: Phantom, I do.
[3:37] Gunter Vandyke: me too phantom
[3:37] Norton Burns: nope, Phantom, I live in it;-))
[3:37] Babbage Linden: but whether we can expose that in the limits without making it too complex is a different matter
[3:37] Morgaine Dinova: Stats Bar? I have it up permanently
[3:37] Phantom Ninetails: Ah good to hear. Am I the only one who wants to see region script memory usage in it?
[3:37] Xugu Madison: and we really, REALLY need bytecode sharing to affect quotas, most of my scripts are designed to be used repeatedly across a sim but are quite big as a result
[3:37] Stickman Ingmann: Is bytecode sharing current behavior, or future behavior?
[3:37] Babbage Linden: i'll bring it up with jack when i next meet with him
[3:37] Norton Burns: mine too Xugu
[3:37] Techwolf Lupindo: If I could de-tach it and put in a second monitor, that floater would be expanded all the way out.
[3:37] Xugu Madison: Stickman, current
[3:38] Stickman Ingmann: Fantastic, thank you Xugu.
[3:38] Phantom Ninetails: Am I the only one who wants to see region script memory usage in statistics?
[3:38] Xugu Madison: Babbage, I'm aware it's going to make it complicated, but without what will happen is we'll end up writing a dozen tiny scripts with minor variations, which won't bytecode share...
[3:38] Norton Burns: but like you Xugu, i don't want to build massive memory multi-purpose scripts to share bytecode, if I'm hit by the memory limit instead
[3:38] Techwolf Lupindo: Question: How can one make sure the same bytecode is being used acrosss two script?
[3:38] Stickman Ingmann: Phantom, I'm up for as much information exposed as possible.
[3:39] Imaze Rhiano: about mono memory usage: I have done some preliminary testing and bare functions without even parameters or return values are consuming like 500 bytes - is there any chance to lower it future? (Or it is just llGetFreeMemory that is laughing for me)
[3:39] Xugu Madison: Techwolf, if it's the same asset (so you save the script and use that copy without recompiling), on the same sim, they should share
[3:39] Babbage Linden: imaze, there seems to a minimum of 4K for an assembly produced by ILASM
[3:39] Babbage Linden: and by Reflection.Emit
[3:40] Babbage Linden: i'll look to see if we can squeeze that down
[3:40] Norton Burns: does that work if I hand the script to someone else & they put it in their own product? no re-compile involved.
[3:40] Babbage Linden: norton, yes
[3:40] Techwolf Lupindo: Ok. I have one tool that uses many scripts to get around some silly limits.
[3:40] Xugu Madison: Compile time inlining of functions only called once (used to make the code more readable) would be nice :)
[3:40] Norton Burns: nice, ty
[3:40] Morgaine Dinova: Babbage: how much extra is added by the instrumentation you inject into assemblies on average?
[3:40] Qie Niangao: so, the downside of mem limits taking into account bytecode sharing is the ill-defined behavior if an owner sheds sq.m. of land in a sim such that it can't support the shared segment anymore. Then every instance would go poof at the same time (I guess).
[3:40] Imaze Rhiano: also constanst would be nice :P
[3:40] Babbage Linden: morgaine, quite a lot :-D
[3:40] Morgaine Dinova: Yield points and stuff.
[3:41] Xugu Madison: Qie, yes, which is really bad, but not quite as bad as the alternative IMHO
[3:41] Babbage Linden: i'd like to be able to use Generics for it
[3:41] Babbage Linden: now that Mono supports generics sharing
[3:41] Babbage Linden: but that will require a Mono upgrade
[3:41] Xugu Madison: Babbage, could we have hard and soft quotas, so going over due to losing land doesn't blow stuff up for a few days if it's not over a higher "hard" quota?
[3:41] Norton Burns: ? - for the uninitiated... generics?
[3:42] Babbage Linden: which will probably happen when they have their verifier done and we do C#
[3:42] Liandra Ceawlin: Are there still any plans seriously in the works to provide us with languages other than LSL, with the mono VM? Maybe C#? Or perhaps even the ability for us to upload our own assemblies someday? I mean, LSL is cool and all, but no offense, it is a bit crusty in some places. >_>
[3:42] Liandra Ceawlin: Ah, you answered my question I guess. :P
[3:42] Xugu Madison: Liandra, blocked by work on Moonlight 2 from the Mono project, AFAIK.
[3:42] Liandra Ceawlin: Okies.
[3:42] Babbage Linden: yes, i spoke to kumpera the other day
[3:42] Babbage Linden: he's planning to have it all done by September
[3:42] Liandra Ceawlin: :O
[3:42] Liandra Ceawlin: Dayum. Hehe.
[3:43] Babbage Linden: it's an open source project, so you can all help of course...
[3:43] Sebastean Steamweaver: Babbage, before we run out of time, I'd like to bring up MISC-3077 to you. we actualyl discussed it at Andrew's office hours, and he liked the fact we were gathering it in one place. Andrew also mentioned the idea of putting a dev on it for three weeks just to clear it out. Could this be done?
[3:43] Xugu Madison: and there's awesome SL-geek points for getting involved with the underlying tech :)
[3:43] Sebastean Steamweaver:
[3:43] JIRA-helper: [#MISC-3077] List of Linden-Confirmed Easy "Featurettes," With Large Returns
[3:44] Sebastean Steamweaver: actually*
[3:44] ThunderBolt Kanto: hi Techie, Seb
[3:44] Stickman Ingmann: Hey, MISC-3077 has my name on it as the creator. Maybe I should have brought it up instead of letting Sebstean. :p
[3:45] Sebastean Steamweaver: :P
[3:45] Babbage Linden: looking
[3:45] Xugu Madison: Can we also get some attention as a good way of reducing number of scripts used
[3:45] JIRA-helper: [#SVC-224] Need a llGetLinkPrimitiveParams function
[3:46] Norton Burns: llName2Key anyone?
[3:46] Babbage Linden: this is great, thanks
[3:46] Stickman Ingmann: I doubt we'll ever see llGetLinkPrimitiveParams().
[3:46] Xugu Madison: With the accompanying
[3:46] JIRA-helper: [#SVC-4270] We DESPERATELY need llGetLinkPrimitiveParams
[3:46] Sebastean Steamweaver: :P
[3:46] Babbage Linden: i will point T and the new platform product manager at it
[3:46] Sebastean Steamweaver: Lets just let Babbage deal with MISC-3077 for the moment.
[3:46] Liandra Ceawlin: llSetLinkText would be cool too. ;)
[3:46] Babbage Linden: and talk to Andrew too
[3:46] Xugu Madison: There's security concerns, but they're not actually real
[3:47] Norton Burns: in getting keys? You can get them anyway, just not easily
[3:47] Qie Niangao: Yeah, I'd really like SVC-224 / -4270 first, perhaps as a preliminary step toward some of the MISC-3077 work.
[3:47] Stickman Ingmann: Oh, Babbage! Do we know who the new product manager is? I heard we were getting one.
[3:47] Babbage Linden: I'm going to bring up the general approach of reducing numbers of script by operating on multiple objects too
[3:48] Sebastean Steamweaver: SVC-2885 :D
[3:48] JIRA-helper: http: //
[#SVC-2885] llSetObjectScale and llGetObjectScale
[3:48] Stickman Ingmann: Qie, I would LOVE SVC-224. If you can convince someone to sneak it in without getting yours and their head chopped off, that would be awesome.
[3:48] Babbage Linden: the PM hasn't started yet, but I'll invite them along to an office hour soon
[3:48] Babbage Linden: so I can introduce them
[3:49] Qie Niangao: oh, that would be a huge win, Babbage, to be able to specify which prim we want to act on... not just for scale, but for text and especially particles and... other stuff not in llSetLinkPrimParams
[3:49] Xugu Madison: Babbage, something to think about for when C# scripting hits. this.prim.links[15].setText() - just saying :-D
[3:49] Stickman Ingmann: Operating on multiple objects? How feasible is that? I wasn't aware SL's internal makeup allowed for something like that.
[3:49] Sebastean Steamweaver: I will be happy to be present :)
[3:49] Techwolf Lupindo: -224 -4270 can be restricted to object/script creaters matching only to allivate security consernces.
[3:49] Babbage Linden: stickman, initial multiple prims
[3:49] Babbage Linden: but multiple objects would also be helpful
[3:49] Babbage Linden: if more work
[3:50] Stickman Ingmann: Sounds wonderful to me, either way.
[3:50] Stickman Ingmann: Also, I'd love to meet the PM when they exist.
[3:50] Sebastean Steamweaver: I would love to meet him too.
[3:50] Sebastean Steamweaver: I'll take him out for a cup of coffee or five.
[3:51] Stickman Ingmann: You can bribe better than coffee, Sebastean.
[3:51] Babbage Linden: ok, i think we got through everything on the agenda
[3:51] Sebastean Steamweaver: I suppose I can.
[3:51] Sebastean Steamweaver: Oh!
[3:51] Babbage Linden: anything else we should talk about for 8 minutes?
[3:51] Sebastean Steamweaver: Babbage, I had one last thing to ask.
[3:52] Norton Burns: SetRot on non-phys, with Tau??
[3:52] Imaze Rhiano: what was thingy that needed to do in Mono to get C# here? I might take a look at it...
[3:52] Sebastean Steamweaver: Would anything like a llGetAgentGroupKey(key) or llMatchGroup() (for checking a group without having to use the group of the object the script is in) be possible?
[3:52] Stickman Ingmann: Norton, Sebastean has a few Jiras in relation with that.
[3:52] Babbage Linden: finish the metadata verifier
[3:52] Norton Burns: kk, tks Stickman
[3:52] Babbage Linden: ask kumpera about it on #mono
[3:53] Babbage Linden: let him know i sent you ;-)
[3:53] Techwolf Lupindo: I though mono and C# was teh same thing? One was done by M$ and the other FOSS done.
[3:53] Gunter Vandyke: yes thats would be great Sebastian
[3:53] Norton Burns: though, yes, group info improvements would be nice
[3:53] Imaze Rhiano: okay thanks
[3:53] Kephra Nurmi still has one point for last minutes: <- whats plan roadmap for llTextBox?
[3:53] JIRA-helper: [#VWR-8759] llTextBox in client Non-Functional - Functionality Patch Attached
[3:53] Gunter Vandyke: and more groups possibe (25 not enough btw)
[3:53] Sebastean Steamweaver: Who are you sending Babbage? You're talking to a large group of peole, hehe.
[3:53] Xugu Madison: Techwolf, C# is a language, Mono is the engine that runs programs compiled from C#. So there's plenty of other languages that you can run under Mono (F#, for example)
[3:53] Sebastean Steamweaver: people*
[3:53] Xugu Madison: Yes! Good grief, yes, llTextBox!
[3:53] Stickman Ingmann: Group MANAGEMENT would be nice. Changing people's roles, role permissions, inviting people, ejecting people, the works.
[3:54] Sebastean Steamweaver: That's on my list Stickman :)
[3:54] Stickman Ingmann: Kephra, Qarl did llTextBox. He said it may be getting some love in the next viewer.
[3:54] Sebastean Steamweaver: Anyone who would like a copy of my notecard with permformance and maintance JIRAs, please just ask.
[3:54] Techwolf Lupindo: I want top scripts. :-)
[3:54] Stickman Ingmann: Magic Viewer 2009, of which no one will say anything in relation to it. :<
[3:54] Babbage Linden: stickman, yes, but that could also be done using web services
[3:54] Liandra Ceawlin: What do y'all think the attachment script memory limit is gonna be? I'd like to start making sure my stuff will be easy to get working with the memory limits as soon as I can. >_>
[3:54] Morgaine Dinova: Babbage: I read that Miguel has added microthreads into Mono (derived from Tomi's Monoco). I'd like to try instrumenting that with RAIL ... wondering how big a project that would be though :-)
[3:54] Gunter Vandyke: yes Sebastion, please
[3:54] Babbage Linden: and interesting question is, what should be scripting APIs
[3:54] Babbage Linden: and what should be services you can use via web calls
[3:54] Stickman Ingmann: Babbage, any way to do it -- web or LSL, would be fantastic. Anything automatable.
[3:55] Kephra Nurmi: the problem with llTextBox ... its a really usefull feature ... implemented on servers side ... lacking client viewer side support for a year
[3:55] Babbage Linden: and if you have web services, do you just let scripts using them http out
[3:55] Babbage Linden: or do you wrap them in library calls
[3:55] Phantom Ninetails: Whoever brought up getting the avatar's group:
[3:55] JIRA-helper: [#SVC-1612] llGetObjectDetails works on avatars but does not return with OBJECT_GROUP
[3:55] Babbage Linden: (or do you let people build libraries that wrap them <--- my favourite)
[3:55] Babbage Linden: ;-)
[3:55] Stickman Ingmann: Liandra, we were already told the "limits" in the memory will be far higher than cause sim degregation as present. It was mentioned on sldev by ... Kelly Linden?
[3:56] Xugu Madison: I think do http-out and focus on making http-out as easy as possible. Although trying to get everything into the memory limits is my biggest challenge ATM. How's library support coming, BTW? :)
[3:56] Babbage Linden: libary support will be post C#
[3:56] Babbage Linden: and only for C# scripts
[3:56] Babbage Linden: LSL has no idea about includes
[3:57] Gunter Vandyke: thx Sebastian
[3:57] Xugu Madison: Oh, Babbage, locale support, for that matter. We need locale assets that can be sent to the client as you would with textures/anims/sounds, and the script simply says llOwnerSayLocale(integer messageID, ... parameters);
[3:57] Babbage Linden: or functions outside the current script
[3:57] Stickman Ingmann: This is all very exciting stuff.
[3:58] Babbage Linden: xugu, yes i agree
[3:58] Babbage Linden: (the new PM is going to have a long wish list to work through)
[3:58] Morgaine Dinova: Hehe
[3:58] Imaze Rhiano: SVC-3848
[3:58] JIRA-helper: http: //
[#SVC-3848] Script-to-viewer communication: llViewerSay, llViewerOwnerSay, llViewerInstantMessage
[3:58] Babbage Linden: (and I'm going to need some more engineers)
[3:58] Morgaine Dinova: PMs live to suffer, you know that ^_^
[3:59] Sebastean Steamweaver: Well Babbage, I've said it before. My "wish list" is more to help performance and scripters overall. If I were going for "shinies", I'd be pushing particle system improvements (whichI'd love, but I don't see as important at this point).
[3:59] Xugu Madison: Babbage, I'm free this time next year, if you've got any positions going in Minnesota :)
[3:59] Babbage Linden: imaze, i'd like to have a more general script to client script communication
[3:59] Stickman Ingmann: Job requirements: Able to be drawn and quartered by multiple wish-lists.
[3:59] Morgaine Dinova: Babbage++
[3:59] Techwolf Lupindo: <-- un-employed for over a year, need good leads.
[4:00] Xugu Madison: Stickman, yeah, I have meetings like this at work already, except they're actually across a table from me
[4:00] Babbage Linden: but client scripts are a long way off, so this may be useful in the meantime
[4:00] Stickman Ingmann: There was an open source deveoper... that one you recommend to me, Sebastean. He was working on client scripts.
[4:00] Sebastean Steamweaver: Nexii Mathus
[4:00] Sebastean Steamweaver: He has some fantastic ideas for client-script integration.
[4:01] Morgaine Dinova: Babbage: I had a proposal written up on MMOX for symmetrical client-side and server-side scripting and communication between them, hehe. Nobody else was interested ^_^
[4:01] Sebastean Steamweaver: Like being able to "subscribe" a script to know where you click in a sim, whenver you click, and being able to act accordingly.
[4:01] Babbage Linden: ok, we're out of time
[4:01] Babbage Linden: i should get back to work
[4:01] Liandra Ceawlin: Thanks, Babbage!
[4:01] Babbage Linden: lots to talk about next time though, it seems
[4:01] Sebastean Steamweaver: Thank you Babbage :)
[4:01] Norton Burns: tks for your time, Babbage
[4:01] Phantom Ninetails: Fare well Babbage
[4:01] Babbage Linden: thanks for coming everyone
[4:01] Xugu Madison: Thanks Babbage, it's been great!
[4:01] Gunter Vandyke: thank Babbage
[4:01] Sebastean Steamweaver: Thank oufr thee OH
[4:01] Imaze Rhiano: thanks Babbage
[4:01] Morgaine Dinova: Thanks Babbage, have fun!
[4:02] Stickman Ingmann: Thank you Bbbage!
[4:02] Sebastean Steamweaver: Let me know when the new PM is here so I can bribe him ;)
[4:02] Stickman Ingmann has a spelling failure.
[4:02] Nock Forager: Seeya in next week.
[4:02] Sebastean Steamweaver: Did anyone else want the notecard of LSL performance improvement JIRAs?
[4:02] Norton Burns: please Seb
[4:02] Phantom Ninetails: At one point there were 21 people here.
[4:03] Norton Burns: ty
[4:03] Phantom Ninetails: Quite alot of people for this meeting
[4:03] DigitalJack Blackhawk: Yes please Sebastean
[4:03] Sebastean Steamweaver: Welcome :)