[11:04 AM] Andrew Hellershanks: Hello, everyone.
[11:04 AM] Ubit Umarov: sure there are serius tech issues.. but for me it is clear that, again the worse threat to OSgrid is its own administration
[11:04 AM] Gavin.Hird @grid.xmir.org:8002: Hi Andrew
[11:04 AM] Orbert.Tatham @hg.zetaworlds.com: Hello, Andrew
[11:05 AM] Andrew Hellershanks: I was rather surprised to hear the latest news about osgrid.
[11:05 AM] Ubit Umarov: not long ago we saw all the nonsense abotu Start All over.. delete everrything...
[11:05 AM] Ubit Umarov: now they went into the full termination of the grid
[11:05 AM] Vincent.Sylvester @hg.zetaworlds.com: At this point I have to wonder if they'd be able to operate a toaster
[11:05 AM] Orbert.Tatham @hg.zetaworlds.com: The chatter said that someone hacked them
[11:05 AM] Gavin.Hird @grid.xmir.org:8002: hacked?
[11:06 AM] Andrew Hellershanks: What I read is that an individual just flooded the grid with assets.
[11:06 AM] Cuga.Rajal @rajal.org:9000: Sad but not surprising. They never had a viable longevity plan, or plan for how they would scale over time
[11:06 AM] Andrew Hellershanks: I wondered why they can't just purge all that data.
[11:06 AM] Orbert.Tatham @hg.zetaworlds.com: Plan? Not sure they knew the word
[11:06 AM] Michael Timeless: seems sad that after almost 20 years griefers are still a thing
[11:06 AM] Ubit Umarov: like i said.. there are problems.. but their reaction was KILL it ALL
[11:06 AM] Gavin.Hird @grid.xmir.org:8002: I wonder too, or restore to some backup point – if they have such a thing that is
[11:07 AM] Ubit Umarov: great understanding of what theior job is..
[11:07 AM] Lyr.Lobo @cc.opensimulator.org:8002: Hello *smiles*
[11:07 AM] Ubit Umarov: again no respect for their users..
[11:07 AM] Andrew Hellershanks: Depends on how long this massive uploading of assets has been going on as to whether they would have a clean(er) backup.
[11:08 AM] Gavin.Hird @grid.xmir.org:8002: if it is that massive, one would reason someone had noticed. ...like in data on disks or tables growing abnormally fast
[11:09 AM] Ubit Umarov: they claim several TB over last year
[11:09 AM] Andrew Hellershanks: They said backups went from 10 minutes to hours.
[11:09 AM] Ubit Umarov: and they never noticed until now
[11:09 AM] Cuga.Rajal @rajal.org:9000: They should never have allowed outsiders to link to their grid and point to their DB in the first place. Their "franchise" model was never viable
[11:10 AM] Ubit Umarov: well OSgrid like opensim and many other things were created in another world
[11:10 AM] Andrew Hellershanks: It was a nice idea to allow users to connect but it sounds like they needed more control over what could be done by the end users.
[11:10 AM] Ubit Umarov: a world that did look like ruled by decent people for decent people
[11:10 AM] Andrew Hellershanks: Hello, Joe.
[11:10 AM] Ubit Umarov: sadly seems that world is gone...
[11:10 AM] Gavin.Hird @grid.xmir.org:8002: right Ubit
[11:10 AM] Lyr.Lobo @cc.opensimulator.org:8002: Hi Joe
[11:11 AM] Lyr.Lobo @cc.opensimulator.org:8002: great to see you
[11:11 AM] Ubit Umarov: trash humas get into people, one way or another
[11:11 AM] Andrew Hellershanks: It only takes one bad actor to ruin it for everyone else. Sadly that seems to be the case here.
[11:11 AM] Ubit Umarov: .. humans..
[11:11 AM] Joe Magarac: Hi all.
[11:11 AM] Lyr.Lobo @cc.opensimulator.org:8002: /me chuckles
[11:11 AM] Orbert.Tatham @hg.zetaworlds.com: Joe
[11:11 AM] Ubit Umarov: hi joe
[11:11 AM] Vincent.Sylvester @hg.zetaworlds.com: We had some commits this week
[11:11 AM] Ubit Umarov: wee like i said... worse enymy of OSgird is again its own admin
[11:11 AM] Ubit Umarov: enemy
[11:12 AM] Ubit Umarov: yeack.. my keyboard is on strike
[11:12 AM] Joe Magarac: If they had a garbage collector for the asset database, like SL does, would that help?
[11:12 AM] Ubit Umarov: we cant have that
[11:12 AM] Joe Magarac: Why not?
[11:12 AM] Ubit Umarov: specially on a grid like osgrid
[11:12 AM] Cuga.Rajal @rajal.org:9000: How do you determone what is "garbage"
[11:12 AM] Gavin.Hird @grid.xmir.org:8002: now what? are people scrambing to export their inventories?
[11:12 AM] Ubit Umarov: regions came and go and return..
[11:13 AM] Ubit Umarov: not wait to knwo what asserts they need...
[11:13 AM] Jagga.Meridith @hg.zetaworlds.com: Did that already. I got off clean
[11:13 AM] Ubit Umarov: etc
[11:14 AM] Joe Magarac: If a region doesn't go live in some period of time (a month) its assets don't get counted.. Garbage collectoin tallies every asset UUID in inventory and in world. What's not used is deleted from asset storage.
[11:14 AM] Andrew Hellershanks: The claim is one person caused the problem. Was it just one avatar account or a whole lot of different avatar accounts created by one person? That should help them determine if the bad data can be purged.
[11:14 AM] Orbert.Tatham @hg.zetaworlds.com: They have thrown up their hands now
[11:14 AM] Ubit Umarov: possible several avatars
[11:14 AM] Ubit Umarov: yeap Orbert...
[11:14 AM] Michael Timeless: Is this why Mantis is also having issues?
[11:14 AM] Andrew Hellershanks: Joe, The problem is with scripts and textures that might have gotten used by others.
[11:15 AM] Orbert.Tatham @hg.zetaworlds.com: I like Joe's idea of a garbage collector
[11:15 AM] Joe Magarac: Their admins are trying to deal with this by hand and are overworked.
[11:15 AM] Ubit Umarov: last time their solution to the assets problem was to delete all
[11:15 AM] Joe Magarac: In SL, a UUID mentioned in a script will not keep it alive.
[11:15 AM] Ubit Umarov: now it is the same.. plus delete the entire grid
[11:15 AM] Gavin.Hird @grid.xmir.org:8002: haha, my inventgory was pretty much non-existent after
[11:15 AM] Vincent.Sylvester @hg.zetaworlds.com: You cannot just delete assets, because then you end up with an asset server hammered for non-existent assets which get retried constantly further reducing performance
[11:16 AM] Gavin.Hird @grid.xmir.org:8002: inventgory - sounds like an appropriate name
[11:16 AM] Joe Magarac: We've seen seasonal trees lose textures because of that. Some textures were referenced only in scripts for 9 months of the year.
[11:16 AM] Vincent.Sylvester @hg.zetaworlds.com: This is already a problem in some cases hence my work the last few weeks on figuring out what's going on in that regard
[11:16 AM] Andrew Hellershanks: Gavin, I know what you mean. I never had much here but I did lose a lot of items.
[11:16 AM] Ubit Umarov: Joe.. lie i said.. the regions are not up all the time
[11:16 AM] Gavin.Hird @grid.xmir.org:8002: I had items back to 2008.
[11:16 AM] Ubit Umarov: impossible to know needs etc
[11:16 AM] Joe Magarac: Right, but if a regioin doesn't check in every week or every month, it should lose its common asset store. Or at least have to get it back from a backup.
[11:17 AM] Orbert.Tatham @hg.zetaworlds.com: I am curious if an upload throttling scheme might have slowed down the attackers?
[11:17 AM] Gavin.Hird @grid.xmir.org:8002: back from backp works if you have a nearline that can restore automatically
[11:17 AM] Ubit Umarov: ofc youj didn not notice but all this information abotu the shutdown will one be public sunday...
[11:17 AM] Vincent.Sylvester @hg.zetaworlds.com: Then people will oar their stuff and each time run those oars "to make sure all the assets are there" increasing the count further
[11:17 AM] Joe Magarac: There are now about 18TB of assets on OSgrid, with all the junk. that's one drive that costs under $1000. Back up the whole thing once in case something gets lost, then garbage collect.
[11:17 AM] Ubit Umarov: that was the plan
[11:18 AM] Cuga.Rajal @rajal.org:9000: Exactly Joe
[11:18 AM] Ubit Umarov: but one of the admins did went to discor tell all about it
[11:18 AM] Orbert.Tatham @hg.zetaworlds.com: And?
[11:18 AM] Joe Magarac: Writing a garbage collector might be a good job for a Fable-level AI. The task is well defined and the interfaces are defined.
[11:19 AM] Cuga.Rajal @rajal.org:9000: I have a GC script for my small grid which checks each UUID, see if it exists in user inventories or in-world regions. Obviously that only works for small grids. on a 50GB database it takes 4 hours to run.
[11:19 AM] Andrew Hellershanks: I wonder what discussions there will be about this at the OSCC.
[11:19 AM] Andrew Hellershanks: Cuga, does it also check the contents of scripts and notecards to see if assets are referenced there?
[11:19 AM] Cuga.Rajal @rajal.org:9000: Nope
[11:20 AM] Joe Magarac: Right, that's what an asset GC does. You just have to write it for scale, and maybe rent time and space on AWS to get it done fast.
[11:20 AM] Ubit Umarov: well also very sad how many ppl just react with things like "that journey is over"
[11:20 AM] Cuga.Rajal @rajal.org:9000: But if they arent foind in worl or inventories, should ne matter
[11:20 AM] Vincent.Sylvester @hg.zetaworlds.com: Can't fix stupid
[11:20 AM] Michael Timeless: That was the theme of the last meeting
[11:20 AM] Gavin.Hird @grid.xmir.org:8002: only way to actually handle it is reference counting the assets that is maintained on every link and unlink
[11:20 AM] Joe Magarac: In SL, scripts and notecards are not searched for assets.
[11:21 AM] Joe Magarac: Reference counts are hard to keep in sync. Need a full scan.
[11:21 AM] Anthropic.Claude @hg.zetaworlds.com: Joe’s not wrong -- garbage collection is a bounded problem with clear success criteria, which plays to AI strengths. The interfaces to OpenSim’s asset and inventory systems are documented. Whether the result would be production-quality is the open question, but as a starting point for a human developer to review and harden it’s plausible.
[11:21 AM] Andrew Hellershanks: They can run a DB query to count how many assets are owned by each avatar. There must be a few with an excessive number of assets in inventory to account for the many TB increase in size of DB.
[11:21 AM] Gavin.Hird @grid.xmir.org:8002: I know it is hard, but you can most likely build it right into the db as a subtable on every asset record
[11:21 AM] Cuga.Rajal @rajal.org:9000: Could add a rule "you can only have XX GB of dB space"
[11:22 AM] Ubit Umarov: i think is is possible to clean this last 'wave'
[11:22 AM] Andrew Hellershanks: Cuga, that is another thing they could have implemented. I suspect no one of the admins expected anyone to do what has been done.
[11:22 AM] Ubit Umarov: problem is the atitude.. no will to fix it.. just take the excuse to terminate all
[11:23 AM] Joe Magarac: Keeping backlinks in sync is a huge pain. Needs a redesign to make that work. GC is a bolt-on fix.
[11:23 AM] Cuga.Rajal @rajal.org:9000: I lost almost everything in OSG, but doesn;t matter, I moved everything important off years ago after their forst DB explosion in the 20-teens
[11:23 AM] Orbert.Tatham @hg.zetaworlds.com: Next question - what motivation does this attacker(s) have? That might be a clue as to how to clean it up
[11:23 AM] Joe Magarac: How broke are the OSGrid people? Can they afford, say, $5000 for this?
[11:23 AM] Ubit Umarov: ( ofc clean will take several hours of scripting code etc )
[11:23 AM] Gavin.Hird @grid.xmir.org:8002: yeah I kinda lost interest in OSG when the iventory dropped.. Only used it to test the viewer builds
[11:24 AM] Cuga.Rajal @rajal.org:9000: Its not just the OSG people, the opensim people are a different bunch now. Not "early adopter" anymore
[11:24 AM] Ubit Umarov: they claim to have good balance.. then cry say they hav eno money
[11:24 AM] Andrew Hellershanks: Orbert, they might have done it to see if they could. They might have done it because the don't like the grid, or the people who run it. They might have done it go get rid of perceived competition. Probably any number of reasons.
[11:24 AM] Joe Magarac: That's the problem. Getting out of this mess requires some big disk space and some serious compute for a few days.
[11:24 AM] Vincent.Sylvester @hg.zetaworlds.com: It requires brains
[11:24 AM] Ubit Umarov: think is a bit har to ask for donations and them make decisons to DELETE ALL the things ppl did donate for
[11:25 AM] Joe Magarac: I have no idea about the politics of OSgrid. I just log in now and then.
[11:25 AM] Ubit Umarov: and then came back to ask for more donations..
[11:25 AM] Gavin.Hird @grid.xmir.org:8002: 😎
[11:25 AM] Ubit Umarov: last week we they where going to charge $$ for connected regions...
[11:25 AM] Ubit Umarov: now this
[11:25 AM] Ubit Umarov: duhh
[11:26 AM] Michael Timeless: There was a lot of anger at last weeks meeting about the change in pricing structure. I have to wonder if this isn't related.
[11:26 AM] Ubit Umarov: hwo can they deserve any doantion?
[11:26 AM] Orbert.Tatham @hg.zetaworlds.com: The $/region thing was the end of me anyway
[11:26 AM] Joe Magarac: Well, it does cost money to operate the asset store. More if it has to be defended against attacks.
[11:26 AM] Ubit Umarov: yeap
[11:27 AM] Ubit Umarov: thats the major cost
[11:27 AM] Cuga.Rajal @rajal.org:9000: Need to end allowing external servers to connect and be part of grid.. but moot now
[11:27 AM] Andrew Hellershanks: The $/region was supposed to be when you wanted to have more than one region.
[11:27 AM] Joe Magarac: Anyway, I've said my piece. This is a solveable problem, with costs. I'm, viewer side, not sim side, and can't really help much personally.
[11:27 AM] Ubit Umarov: evne considering how ' memry is cheap'we keep many devs say is last years :)
[11:28 AM] Ubit Umarov: ...memory...
[11:28 AM] Ubit Umarov: well about opensim
[11:28 AM] Ubit Umarov: i did add a few bugs
[11:28 AM] Joe Magarac: At least make sure that someone backs up the whole asset database.
[11:28 AM] Gavin.Hird @grid.xmir.org:8002: memory is almost as expensive as core used to be
[11:28 AM] Ubit Umarov: change this.. changed that.. tiny things here and there
[11:29 AM] Gavin.Hird @grid.xmir.org:8002: love your commit messages Ubit 😂
[11:29 AM] Ubit Umarov: added support for llgetobjectdetails LOCKED and VOLUME_DETECT
[11:30 AM] Ubit Umarov: that seems ll only has on beta grid for now
[11:30 AM] Michael Timeless: I had a few fixes for Linking and database update async clashing but couldn't get into Mantis at all.
[11:31 AM] Gavin.Hird @grid.xmir.org:8002: Are Mantis accounts actually OSG accounts?
[11:31 AM] Ubit Umarov: maybe andrew can take a look to what happened to yr mantis account
[11:31 AM] Vincent.Sylvester @hg.zetaworlds.com: I'm hearing some good news from FS that they are looking to hopefully put the recent fixes into their next release. We been tracking that ObjectUpdate packet issue for some time now. Still no smoking gun, some ideas on what it might be. Viewer side should guard against it and the few other crashes and bugs we found.
[11:31 AM] Vincent.Sylvester @hg.zetaworlds.com: There is a known mantis about the deleting of previously linked items causing database desync
[11:31 AM] Vincent.Sylvester @hg.zetaworlds.com: That's down to the deferred store of those changes
[11:32 AM] Ubit Umarov: like i was tellikg also added some code for OBJECT_PERMS
[11:32 AM] Ubit Umarov: OBJECT_PERMS_COMBINED is a bit more complex
[11:32 AM] Vincent.Sylvester @hg.zetaworlds.com: http://opensimulator.org/mantis/view.php?id=9228
[11:33 AM] Joe Magarac: SL-side object update problem was that there were multiple copies of zerocoding compression, and someone unified them and broke something. I saw that in Sharpview. It's fixed in SL now.
[11:33 AM] Michael Timeless: yeah I was trying to answer report 0009228
[11:33 AM] Ubit Umarov: we do have full scan of prim inventories perms with cache, but only for the permitions classes we need
[11:33 AM] Ubit Umarov: that support requires 2 more
[11:34 AM] Ubit Umarov: cosidering ppl put 1000 items per prim on links sets of more than 1000 prims..
[11:34 AM] Ubit Umarov: this is a issue
[11:34 AM] Vincent.Sylvester @hg.zetaworlds.com: Joe, not really, they did remove the guards against it and only now put them back. The question is where that comes from server side, cause I haven't found that yet
[11:34 AM] Vincent.Sylvester @hg.zetaworlds.com: Ubit had some idea on where it might be
[11:35 AM] Ubit Umarov: shh not telling
[11:35 AM] Vincent.Sylvester @hg.zetaworlds.com: If viewers guard against it, given how many object updates are sent the failure rate was low, but not zero
[11:35 AM] Ubit Umarov: >p
[11:37 AM] Ubit Umarov: anyway.. the viewer protection on that is light.. no idea why removed
[11:37 AM] Gavin.Hird @grid.xmir.org:8002: mobile viewer conciderations?
[11:37 AM] Ubit Umarov: no protections is our game.. not ll duhhh
[11:38 AM] Vincent.Sylvester @hg.zetaworlds.com: LL have been making quite a few weird decisions on that
[11:38 AM] Vincent.Sylvester @hg.zetaworlds.com: Turning brain off seems to be the new trend
[11:38 AM] Ubit Umarov: who do they dare to try to worse than us hmm??
[11:38 AM] Ubit Umarov: (how.. )
[11:38 AM] Vincent.Sylvester @hg.zetaworlds.com: At least Firestorm seem to care about the users and implement fixes for it
[11:39 AM] Vincent.Sylvester @hg.zetaworlds.com: Feels good to be cared about
[11:39 AM] Ubit Umarov: brain.. oh i know i had forgot something downstairs
[11:39 AM] Ubit Umarov: (i knew)
[11:40 AM] Ubit Umarov: i also added to robust a option to refuse automatic regions registration into the grid
[11:41 AM] Ubit Umarov: another case of how opensim was made for that other Woderfull magic world
[11:41 AM] Cuga.Rajal @rajal.org:9000: should refuse by default
[11:41 AM] Gavin.Hird @grid.xmir.org:8002: wasn't that configurable already?
[11:41 AM] Ubit Umarov: no
[11:41 AM] Gavin.Hird @grid.xmir.org:8002: yes should refuse by default
[11:41 AM] Ubit Umarov: well added a option now
[11:41 AM] Gavin.Hird @grid.xmir.org:8002: good
[11:42 AM] Ubit Umarov: ofc not a issue on normal grids
[11:42 AM] Ubit Umarov: where only gird owner can start regions
[11:42 AM] Ubit Umarov: but osgrid... well
[11:43 AM] Gavin.Hird @grid.xmir.org:8002: well
[11:43 AM] Ubit Umarov: https://www.youtube.com/watch?v=VqhCQZaH4Vs
[11:43 AM] Cuga.Rajal @rajal.org:9000: OSG website says nothing about closure... wtf?
[11:44 AM] Orbert.Tatham @hg.zetaworlds.com: Look under news
[11:44 AM] Ubit Umarov: it will only be public next sunday LOL
[11:44 AM] Michael Timeless: When you log in with an OSG ID it states there will be an announcement on the 18th and requires a checkbox
[11:44 AM] Ubit Umarov: there will be another meeting for every one
[11:44 AM] Gavin.Hird @grid.xmir.org:8002: I only knew because Maria messaged me for comment to HGB
[11:45 AM] Jagga.Meridith @hg.zetaworlds.com: err...won't we need to move meeting place? I suggest zeta
[11:45 AM] Cuga.Rajal @rajal.org:9000: Cant find link to news on homepage... doesn't matter
[11:45 AM] Ubit Umarov: shh shh no one notice that paela made it public in discord.. etc
[11:45 AM] Gavin.Hird @grid.xmir.org:8002: HGB has 3 stories now. Not that many read that palce anyway
[11:46 AM] Michael Timeless: That's why I keep this old ID after I've lost all my inventory multiple times.
[11:46 AM] Ubit Umarov: ohh i see yes
[11:46 AM] Vincent.Sylvester @hg.zetaworlds.com: From TJ, Firestorm dev: The guard was removed by Rye a while back to "optimize" the viewer
[11:46 AM] Ubit Umarov: big day for maria
[11:46 AM] Vincent.Sylvester @hg.zetaworlds.com: I can't find the rolls eye emoji
[11:46 AM] Cuga.Rajal @rajal.org:9000: lol
[11:46 AM] Gavin.Hird @grid.xmir.org:8002: 🤪
[11:46 AM] Joe Magarac: I heard about the OSgrid shutdown on Discord.
[11:46 AM] Jagga.Meridith @hg.zetaworlds.com: should be an animation
[11:47 AM] Ubit Umarov: yeap paela made it public there.. before time
[11:47 AM] Gavin.Hird @grid.xmir.org:8002: FS does not support complex emojis why youmight not find it
[11:47 AM] Cuga.Rajal @rajal.org:9000: Well that leak will help soften the sudden load of people getting their stuff
[11:47 AM] Ubit Umarov: Dan had told me it whould only be public sunday, like and said
[11:47 AM] Ubit Umarov: as soon dan leaves to rl work... poff public
[11:48 AM] Ubit Umarov: let me play that music again.. i need it :)
[11:48 AM] Joe Magarac: I still think this is a fixable problem. It's worth planning out a GC system. After the attacker gets done with OSGrid, they might attack others.
[11:49 AM] Ubit Umarov: yes joe... the tech id fixable.. the grid manages... well lets see
[11:49 AM] Jagga.Meridith @hg.zetaworlds.com: any competent DBA should be able to isolate the griefer inventory and nuke it, right?
[11:50 AM] Ubit Umarov: see the real problem on this joe?
[11:50 AM] Ubit Umarov: well it can be hard
[11:50 AM] Orbert.Tatham @hg.zetaworlds.com: https://www.osgrid.online/news/the-end-of-the-road/
[11:50 AM] Ubit Umarov: i think i know how, in this case
[11:50 AM] Gavin.Hird @grid.xmir.org:8002: it is hard because you have to unpack all the xml to find the references
[11:50 AM] Ubit Umarov: but will that hours
[11:50 AM] Vincent.Sylvester @hg.zetaworlds.com: Cleaning that stuff is a pain. I have had to delete millions of notecards before with no doubt some of them now being requested by regions causing all the 404 errors I'm seeing next to all the dynamic textures being requested. You need to be careful with deleting stuff not just for local things, but also in regards to HG. The bigger thing is monitoring and keeping backups so you can roll things back, which if things like inventory go awry isn't easy either
[11:51 AM] Vincent.Sylvester @hg.zetaworlds.com: Tools for all that technically exist and you can use existing code to fetch references, ask regions for what they "own" and so on, but you are chasing mice at that point
[11:51 AM] Vincent.Sylvester @hg.zetaworlds.com: Monitoring, backups, preventative maintenance
[11:51 AM] Joe Magarac: Right. First step is backup of everything. That will cost maybe $1000 of disk.
[11:52 AM] Vincent.Sylvester @hg.zetaworlds.com: Sticking your head in the sand, blaming the software and developers and at the point of failure hiding behind volunteer status is...
[11:52 AM] Ubit Umarov: we need to: https://www.youtube.com/watch?v=loTjRIL7qWo
[11:52 AM] Ubit Umarov: but sadly osgrid admin lost that..
[11:52 AM] Ubit Umarov: at least for now..
[11:52 AM] Ubit Umarov: lets see..
[11:53 AM] Vincent.Sylvester @hg.zetaworlds.com: Yeas ago I warned about the notecards piling up as some people can't write scripts properly, looking at you satyr, and that deletion is basically necessary even if that means dead references. No idea if anyone did, but I can only lead the horse to water, it has to drink it itself as they say
[11:53 AM] Cuga.Rajal @rajal.org:9000: I am curious whats going to happen to performance on other grids, when they try to get creator info for all the assets from OSG
[11:54 AM] Vincent.Sylvester @hg.zetaworlds.com: All that shit is known and the code is open, if there is concern, read it and see. All the options are there to run things properly
[11:54 AM] Joe Magarac: Basic concept: Back up, Set up database (probably Postgres, it scales), scan all inventory and world and build databas of UUIDs, Run scanner a second time to catch anything new since first scan. Make new asset store with only UUIDs in the database. Swtich to new asset store. Keep backup around for complaints.
[11:54 AM] Joe Magarac: Insist that if you want your assets, you must bring your regions online during a specified scan period.
[11:54 AM] Ubit Umarov: well im still trying to ask Dan to revert this decison
[11:55 AM] Vincent.Sylvester @hg.zetaworlds.com: I'm sick and tired of it being dismissed as something impossible when advise is not taken and even offers of hands on help are shoved aside because of some ego crusade while the incompetence is putting all of OpenSim in a bad light
[11:55 AM] Gavin.Hird @grid.xmir.org:8002: Tiny bit of news about my viewer (macOS only): We have a test build out where CEF has been completely removed and replaced with WKWebView, also for MOAP. Everything running in the viewer sandbox, no fony memory transfer. Istalled size on disk is down from 470 Mb to right over 200Mb
[11:55 AM] Ubit Umarov: put grid in quarentine mode... only with its core regions
[11:55 AM] Ubit Umarov: lbsa.. event.. wright
[11:55 AM] Joe Magarac: I agree. I'm not going to get into the politics.
[11:55 AM] Ubit Umarov: and keep all inventorues and assets..
[11:55 AM] Vincent.Sylvester @hg.zetaworlds.com: Gavin, does that improve moap performance as well?
[11:55 AM] Ubit Umarov: cleaning as possible
[11:56 AM] Ubit Umarov: of that is the hard part.. storage costs a lot
[11:56 AM] Ubit Umarov: but i think keeping that alive well get ppl to donate and help
[11:56 AM] Ubit Umarov: will
[11:56 AM] Ubit Umarov: that is what im trying...
[11:57 AM] Gavin.Hird @grid.xmir.org:8002: depends a bit what you do and which macOS version it is running on. Youtube was the big hurdle and on some macOS versions fps drops by 20% when a YT video is running on MOAP. If you run it in the internal browser there is no performance drop at all
[11:57 AM] Ubit Umarov: ofc keep the the social life..
[11:57 AM] Cuga.Rajal @rajal.org:9000: If OSG goes down will that clog up stuff on other grids? Like trying to get info for assets that came from OSG
[11:58 AM] Cuga.Rajal @rajal.org:9000: I know Vince has mentioned that issue
[11:58 AM] Michael Timeless: I just ran a check on my small grid it looks like this:How much of our content points at OSgrid creators Where References Notes CatGrid, Catlandia 60,956 47,927 of them are Nebadon Izumi; also Linda Kellie, James Beejay, Jim Jackson CatGrid, Storage World (ZS) 12,896 CatGrid, Gray Goo 4,400 TomCat 730 prims + 17 items
[11:58 AM] Joe Magarac: Seagate 20TB disk drive is now under $700.
[11:58 AM] Lyr.Lobo @cc.opensimulator.org:8002: good question on the ripple effect and impact, Cuga
[11:58 AM] Joe Magarac: I have to go; next meeting.
[11:58 AM] Ubit Umarov: worse.. it is a rented machine on a datacenter joe
[11:58 AM] Joe Magarac: Good luck. Don't give up the ship.
[11:58 AM] Vincent.Sylvester @hg.zetaworlds.com: My problem with moap was that I could never get it to refresh quickly and reliable to be useful for what I wanted to do with it. I had some workarounds for that, but it wasn't really ideal. I remember specifically moving items on the loaded pages would somehow not render the updated position for some reason. Though I wonder if that was just a bug in the library or something in rendering failing to update
[11:59 AM] Cuga.Rajal @rajal.org:9000: take care Joe
[11:59 AM] Lyr.Lobo @cc.opensimulator.org:8002: bye Joe
[11:59 AM] Michael Timeless: bye joe
[11:59 AM] Ubit Umarov: have fun joe
[11:59 AM] Gavin.Hird @grid.xmir.org:8002: what did you want to do wit it Vincent?
[11:59 AM] Cuga.Rajal @rajal.org:9000: Vince could you comment on ripple effect of OSG closing, ie clogging up network waiting for creator ino etc
[12:00 PM] Ubit Umarov: ( alse spent sometime with dan looking to some of the offending assets )
[12:00 PM] Andrew Hellershanks: I'm back. Had to step away for a few moments.
[12:00 PM] Michael Timeless: WB
[12:00 PM] Vincent.Sylvester @hg.zetaworlds.com: I had built something like a window manager, like one of those websites that emulate a desktop and moving the windows around would always freeze since it reloaded the page. I then did it via just changing the css right, but that somehow failed to update once in a while. Really weird things
[12:00 PM] Ubit Umarov: so .. yeah.. osgrid needs to go into quarentine.. mode
[12:01 PM] Ubit Umarov: does not need to close..
[12:01 PM] Orbert.Tatham @hg.zetaworlds.com: If/When OSG foes does, we are likely to have to put a hack on our own grids to block OSG requests.
[12:01 PM] Ubit Umarov: IMO
[12:01 PM] Gavin.Hird @grid.xmir.org:8002: OK, well test the build and give me feedback how it works
[12:01 PM] Vincent.Sylvester @hg.zetaworlds.com: osgrid or any larger grid closing will have detrimental effects on hypergrid data. Creator data, profiles, friends, those are requests made outbound to retrieve data from other grids it is linked to
[12:01 PM] Vincent.Sylvester @hg.zetaworlds.com: In the best case it just clogs things up with requests and you have to remove the reference
[12:02 PM] Ubit Umarov: it only depends on osgrid admin to honour theis mission or not
[12:02 PM] Orbert.Tatham @hg.zetaworlds.com: That is why I said some sort of filter
[12:02 PM] Vincent.Sylvester @hg.zetaworlds.com: Worst case, which has already happened once or twice, is the url pointing to an IP that responds with headers, but no data. This can cause re-requesting and error spam
[12:02 PM] Gavin.Hird @grid.xmir.org:8002: we deliberately dropped some support like open in new tab or oven in new window (which would have to duplicate the prim)
[12:02 PM] Ubit Umarov: and yes.. creators checks may burn some cpu
[12:02 PM] Andrew Hellershanks: Vincent, I've wondered about scripts and notecards too. If I create a script or NC, them later modify it and resave it should update the one asset. The only time it is a problem is if the script or NC has been referenced in an object, or given to others.
[12:03 PM] Vincent.Sylvester @hg.zetaworlds.com: Back on mono the difference behavior of the http client would cause regions to entirely lock up. Have 500+ friends on dead grids = no more teleporting for you
[12:03 PM] Ubit Umarov: ohh andrew.. forgetig Melanie rules on immutable assets
[12:03 PM] Ubit Umarov: any change you make is a NEW asset
[12:04 PM] Vincent.Sylvester @hg.zetaworlds.com: Thankfully dotnet is less retarded in that case and does work better, but the servers responding with headers and garbled data is still a problem. Worst case this can crash stuff as Cuga knows just getting random bad data used to do that before we put some checks in place
[12:04 PM] Ubit Umarov: ( think others bf here had same rule..)
[12:04 PM] Andrew Hellershanks: I know. That is why we get all this bloat in assets. If I'm the one that made it and haven't given it to anyone else making a new asset every time I update it just wastes space.
[12:04 PM] Vincent.Sylvester @hg.zetaworlds.com: And you cannot or should not change assets to remove such references either, because that could be seen as removing copyright notice
[12:05 PM] Vincent.Sylvester @hg.zetaworlds.com: If you leave it in and that asset gets copied, you now are a spreader of bad data
[12:05 PM] Ubit Umarov: giving to other is just sharing the referene
[12:05 PM] Cuga.Rajal @rajal.org:9000: Andrew that was going to be one of my next questions
[12:05 PM] Vincent.Sylvester @hg.zetaworlds.com: So in every case you lose
[12:05 PM] Ubit Umarov: just any change on text is a new asset
[12:05 PM] Vincent.Sylvester @hg.zetaworlds.com: Fsassets and similar things deduplicate so you have at least 20% less space usage if you go that route, but even that has to create a new asset in the db for each even if the data is the same, simply because the metadata is different
[12:06 PM] Gavin.Hird @grid.xmir.org:8002: copy on write and delete the old version after?
[12:06 PM] Andrew Hellershanks: I know. It bugs me a bit. I often wonder if there is a better way of handling updates to scripts or NCs.
[12:06 PM] Vincent.Sylvester @hg.zetaworlds.com: We have some code now to prevent too much of that when you create new empty scripts or notecards
[12:06 PM] Ubit Umarov: oyu cant delete the old version ..
[12:06 PM] Lyr.Lobo @cc.opensimulator.org:8002: good thought
[12:06 PM] Ubit Umarov: it can already in use elsewhere
[12:06 PM] Gavin.Hird @grid.xmir.org:8002: why not Ubit?
[12:06 PM] Ubit Umarov: just answered :P
[12:06 PM] Vincent.Sylvester @hg.zetaworlds.com: Once a user has logged in once you shouldn't delete them either, or change their username
[12:07 PM] Gavin.Hird @grid.xmir.org:8002: I can see that no being the case always, but I digress
[12:07 PM] Orbert.Tatham @hg.zetaworlds.com: Once an asset has been created, there is no way to delete it because it could have been moved or transferred, so the UUID is still needed
[12:07 PM] Ubit Umarov: yeap
[12:07 PM] Cuga.Rajal @rajal.org:9000: Can OSG be DNS-aliased so it stays alive in DNS for lookuos? Does that help?
[12:08 PM] Orbert.Tatham @hg.zetaworlds.com: Still need server for the lookups
[12:08 PM] Ubit Umarov: well legally a 501 association as rules about the company assets
[12:08 PM] Cuga.Rajal @rajal.org:9000: I see.. right
[12:08 PM] Andrew Hellershanks: Orbert, there would need to be a flag to say if it was used elsewhere but I know there would be problems with trying to track that.
[12:08 PM] Ubit Umarov: not sure who can legally own the dns domain
[12:09 PM] Cuga.Rajal @rajal.org:9000: If Dan specufucally allowed it...
[12:09 PM] Ubit Umarov: texas laws for example forbid members to keep anything
[12:09 PM] Ubit Umarov: all needs to be giving to texas gov LOL
[12:09 PM] Vincent.Sylvester @hg.zetaworlds.com: There are some options of local blacklists to prevent the requests going out if you know something is definitely dead. It means you are still retaining bad data and potentially spread that, but outside of altering the assets that's the only option
[12:10 PM] Orbert.Tatham @hg.zetaworlds.com: Local Blacklist is what I was thinking
[12:10 PM] Cuga.Rajal @rajal.org:9000: ok, so OSG could be added to the dead list
[12:10 PM] Ubit Umarov: man... we must Resist and Bite still
[12:10 PM] Ubit Umarov: call dan many names in ims like i do
[12:10 PM] Ubit Umarov: :)
[12:11 PM] Cuga.Rajal @rajal.org:9000: So best option is to purge everything and all friends from DB
[12:11 PM] Vincent.Sylvester @hg.zetaworlds.com: I looked into that option and as a sort of service for a grid to tell regions what not to request from that can work, but then that still needs every grid to make sure they copy that information and coordinate and we know how well that usually goes
[12:11 PM] Lyr.Lobo @cc.opensimulator.org:8002: poor Dan
[12:11 PM] Cuga.Rajal @rajal.org:9000: fromOSG that is
[12:11 PM] Orbert.Tatham @hg.zetaworlds.com: It is even harder to get them to rethink than it was to get them to think the first time
[12:11 PM] Andrew Hellershanks: Orbert, only assets marked temporary could get removed (in theory). Not sure what the mechism is for having an asset marked as temp.
[12:11 PM] Orbert.Tatham @hg.zetaworlds.com: Don't remember there being one
[12:11 PM] Vincent.Sylvester @hg.zetaworlds.com: It thankfully is less problematic now that http isn't as blocking anymore, so the only big concern is bad actors returning bogus data rather than the request timeouts clogging stuff up
[12:12 PM] Lyr.Lobo @cc.opensimulator.org:8002: Thank you for a great meeting.My apologies for dashing back to life....and classes
[12:12 PM] Ubit Umarov: have fun Lyr.Lobo
[12:12 PM] Cuga.Rajal @rajal.org:9000: tc Lyr =)
[12:12 PM] Ubit Umarov: :)
[12:13 PM] Lyr.Lobo @cc.opensimulator.org:8002: Thank you
[12:13 PM] Cuga.Rajal @rajal.org:9000: Is the Dead List user-configurable or is it hard coded
[12:13 PM] Vincent.Sylvester @hg.zetaworlds.com: This shit has been running up and down my brain for years now, trying to figure out a way, but the way HG works assumes all endpoints remain existing and we don't have a way to federate that outside of providing such connectors that grids can then use to sync up to a central thing
[12:13 PM] Lyr.Lobo @cc.opensimulator.org:8002: OSCC is still in planning
[12:13 PM] Vincent.Sylvester @hg.zetaworlds.com: A lot of it requires just talking to each other to be good stewards of their grids and the hypergrid
[12:13 PM] Andrew Hellershanks: Thanks for being here, Lyr.
[12:13 PM] Ubit Umarov: good there is more Opensim outside osgrid
[12:14 PM] Vincent.Sylvester @hg.zetaworlds.com: This isn't just throw stuff up willy nilly and hope for the best and code can't fix this because it can never make those assumptions or at least it shouldn't
[12:14 PM] Gavin.Hird @grid.xmir.org:8002: will there be a stable release before they close?
[12:14 PM] Vincent.Sylvester @hg.zetaworlds.com: So throwing up a small grid or standalone without a care on what that means for everyone else well that's, to me, closest thing to hostile someone can be
[12:15 PM] Vincent.Sylvester @hg.zetaworlds.com: Operating a grid isn't trivial, there are a ton of things to consider
[12:15 PM] Vincent.Sylvester @hg.zetaworlds.com: From how things operate, to legal requirements just down to having users that you are responsible for
[12:15 PM] Michael Timeless: ...and then there is dreamgrid...
[12:16 PM] Vincent.Sylvester @hg.zetaworlds.com: Don't even get me started on that or the "bypass viewer limits on mesh uploads" which is going to cause all manners of region crashes when physics engines explode on the nonsense people upload
[12:16 PM] Vincent.Sylvester @hg.zetaworlds.com: There is a reason the viewer checks for physics
[12:16 PM] Orbert.Tatham @hg.zetaworlds.com: Along the line of Gavin's question: Has there been any organized attempt to size up what migrating to dotnet 11 will take? I had to downgrade a server because Micro$haft wouldn't give me the dotnet8 runtime for it
[12:17 PM] Gavin.Hird @grid.xmir.org:8002: It is probalby a good idea to make a release now that people will start maing their own grids or standalones
[12:17 PM] Gavin.Hird @grid.xmir.org:8002: making*
[12:18 PM] Ubit Umarov: hmm release?
[12:18 PM] Vincent.Sylvester @hg.zetaworlds.com: There is a ppa with versions still available and in both ubuntu and debian the versions don't have any large dependencies that would make them fail to run, so downgrading the distro shouldn't be necessary
[12:18 PM] Vincent.Sylvester @hg.zetaworlds.com: The problem with moving to newer version is the removed things we depend on
[12:18 PM] Anthropic.Claude @hg.zetaworlds.com: dotnet 11 isn’t LTS -- that’ll be dotnet 12 (even numbers are LTS).
[12:18 PM] Gavin.Hird @grid.xmir.org:8002: you know the meaning of the word Ubit 😉
[12:18 PM] Vincent.Sylvester @hg.zetaworlds.com: That needs to be exchanged for something else and the verdict is still out on what the best option is
[12:18 PM] Vincent.Sylvester @hg.zetaworlds.com: There are multiple, but they have their own caveats to handle
[12:18 PM] Cuga.Rajal @rajal.org:9000: You can still get Dotnet 8 on M$ website and install manually
[12:19 PM] Gavin.Hird @grid.xmir.org:8002: yup
[12:19 PM] Cuga.Rajal @rajal.org:9000: I did that recently
[12:19 PM] Jagga.Meridith @hg.zetaworlds.com: that's what I'm running
[12:20 PM] Vincent.Sylvester @hg.zetaworlds.com: It's being looked into, but there are a bunch of things that need doing still so give it some time
[12:20 PM] Vincent.Sylvester @hg.zetaworlds.com: With all the things submitted we are working on getting the fixes reviewed and released
[12:21 PM] Gavin.Hird @grid.xmir.org:8002: is anything broken that prevents a release?
[12:21 PM] Vincent.Sylvester @hg.zetaworlds.com: Gavin sent in a bunch a while back that I did mostly fix, I think, but I haven't had a chance to properly test yet
[12:21 PM] Gavin.Hird @grid.xmir.org:8002: more broken that usually I mean
[12:21 PM] Vincent.Sylvester @hg.zetaworlds.com: There are currently a few known bugs and things yeah, much rather fix those up before release, though that's up to Ubit, I can only throw in my own protest, he doesn't have to listen :)
[12:22 PM] Michael Timeless: Has anyone looked at https://tuxcare.com/blog/net-8-end-of-life/
[12:22 PM] Vincent.Sylvester @hg.zetaworlds.com: If need be I been meaning to revive an apt repo anyways so putting packages on there for a few dependencies including libgdiplus
[12:22 PM] Cuga.Rajal @rajal.org:9000: That EOL date has been at MS website for some time, not new news
[12:22 PM] Andrew Hellershanks: Not yet. Not a site I look at regularly.
[12:23 PM] Vincent.Sylvester @hg.zetaworlds.com: Which speaking of both ubuntu and debian are still sleeping on building new versions with the fixes
[12:23 PM] Lyr.Lobo @cc.opensimulator.org:8002: .NET 8 end of life is November 10, 2026
[12:23 PM] Cuga.Rajal @rajal.org:9000: Last meeting I asked if we should care =)
[12:23 PM] Michael Timeless: I know but they are offering to support it for $$$
[12:23 PM] Vincent.Sylvester @hg.zetaworlds.com: Win10 eol was last year, yet I ain't fucking moving to 11, cause fuck that junk
[12:23 PM] Gavin.Hird @grid.xmir.org:8002: Have Claude help you test. It is sometimes going nuts to make test plans, or decompile libraries to find why things are not working, or write a buch of extra code to test
[12:23 PM] Ubit Umarov: yeah you should not care
[12:23 PM] Andrew Hellershanks: I can't move to Win11. My laptop doesn't meet the requirements.
[12:24 PM] Vincent.Sylvester @hg.zetaworlds.com: Most of the fixes are db related so I just need to setup testing environments for them again and poke it
[12:24 PM] Michael Timeless: Neither did mine that's why I went to Kubuntu
[12:24 PM] Cuga.Rajal @rajal.org:9000: I tried 11 on a VM, and after a few days said "F this"
[12:24 PM] Ubit Umarov: .net 4,8 is long gone... and windows depends on it
[12:24 PM] Ubit Umarov: bahh
[12:24 PM] Vincent.Sylvester @hg.zetaworlds.com: Specifically for postgres that's a pain though so I been avoiding it cause I'm lazy
[12:24 PM] Andrew Hellershanks: I still need at least one machine in the house running Windows. The rest are Linux, not counting a couple Android tablets that can't be updated.
[12:25 PM] Anthropic.Claude @hg.zetaworlds.com: The decompile-and-analyze use case is exactly where cross-referencing implementation details pays off -- same approach as the viewer/server cross-reference he mentioned earlier.
[12:25 PM] Gavin.Hird @grid.xmir.org:8002: it is actaully awesome. at times I sit back floored almost
[12:26 PM] Vincent.Sylvester @hg.zetaworlds.com: We have enough brain power between all the people involved in development that I don't think there is anything not fixable, just coordination might be lacking a bit. Need to get more people onto discord or some other direct messaging to progress not just once a week during the meeting :)
[12:26 PM] Gavin.Hird @grid.xmir.org:8002: like when I threw it 22 lsl scripts and it just went out an pulled the entire lsl implementation from thje opensim server code and fixed the scripts to work with it
[12:27 PM] Cuga.Rajal @rajal.org:9000: I need to dash to RL... take care all, see you next week
[12:27 PM] Vincent.Sylvester @hg.zetaworlds.com: Communication is the human strong point or at least it is supposed to be
[12:28 PM] Jagga.Meridith @hg.zetaworlds.com: I don't have the discord link. I thought that was for oSG
[12:28 PM] Cuga.Rajal @rajal.org:9000: *smies and waves
[12:28 PM] Lyr.Lobo @cc.opensimulator.org:8002: /me waves
[12:28 PM] Vincent.Sylvester @hg.zetaworlds.com: There is an OpenSimulator discord, though most of the behind the scenes development tends to happen in various dms
[12:29 PM] Gavin.Hird @grid.xmir.org:8002: I am going to hop back to RL
[12:29 PM] Gavin.Hird @grid.xmir.org:8002: have a good week
[12:29 PM] Michael Timeless: You too
[12:29 PM] Orbert.Tatham @hg.zetaworlds.com: Peace, Gavin
[12:30 PM] Jagga.Meridith @hg.zetaworlds.com: apologies for osimdev.org being down, power issue. Need new UPS and generator. Also will have more time now that project is finished. https://www.audible.com/pd/B0HM66PLC9?source_code=ASSOR150021921000O
[12:32 PM] Lyr.Lobo @cc.opensimulator.org:8002: it returned an error, Jagga
[12:33 PM] Lyr.Lobo @cc.opensimulator.org:8002: I need an extraction team *laughs*
[12:33 PM] Lyr.Lobo @cc.opensimulator.org:8002: /me poofs
[12:34 PM] Michael Timeless: FWIW it worked for me, added it to my wish list. Thanks Jagga. Take care everyone.
[12:34 PM] Ubit Umarov: have fun
[12:34 PM] Andrew Hellershanks: We have been going for an hour and a half. Time to close this meeting.
[12:34 PM] Andrew Hellershanks: Thank you all for coming. See you again next week.
