[11:00 AM] Andrew Hellershanks: Hello, everyone.
[11:00 AM] Gavin.Hird @grid.xmir.org:8002: Hi Andrew
[11:01 AM] Jagga.Meridith @hg.zetaworlds.com: hi
[11:01 AM] Anthropic.Claude @hg.zetaworlds.com: hi
[11:03 AM] Andrew Hellershanks: Hm... Ubit isn't here yet. I will see if he is in IRC.
[11:04 AM] Andrew Hellershanks: He isn't logged in to IRC. I wonder if he is having problems with his Internet provider.
[11:06 AM] Object: stage_display: total prims=1
[11:06 AM] Gavin.Hird @grid.xmir.org:8002: Didn't he say he lose power quite often?
[11:07 AM] Andrew Hellershanks: That might be it. IIRC, the provider of his Internet was having problems in the last week or two.
[11:07 AM] Gavin.Hird @grid.xmir.org:8002: ok
[11:08 AM] Vincent.Sylvester @hg.zetaworlds.com: Commits this week included some patches and pull requests
[11:08 AM] Vincent.Sylvester @hg.zetaworlds.com: One regarding material encoding, because apparently that's somehow overloading protocol
[11:08 AM] Lyr.Lobo @cc.opensimulator.org:8002: /me smiles
[11:09 AM] Vincent.Sylvester @hg.zetaworlds.com: There was also some updates to postgres things, which conveniently broke the changes I was working on
[11:09 AM] Vincent.Sylvester @hg.zetaworlds.com: A couple minor cosmetics and safety added as well
[11:10 AM] Vincent.Sylvester @hg.zetaworlds.com: An update to libomv for the csj2k which I haven't had the time to look at yet
[11:10 AM] Vincent.Sylvester @hg.zetaworlds.com: Some of it apparently ai suggested
[11:11 AM] Vincent.Sylvester @hg.zetaworlds.com: I finished refactoring all 3 database connectors, updating a bunch of syntax and basically rewriting most queries, escaping everything and changing to less heavy queries in certain cases
[11:12 AM] Vincent.Sylvester @hg.zetaworlds.com: Some testing so far is looking good, it compiles and runs, so now have to do some runtime testing on the problems that did exist prior
[11:13 AM] Vincent.Sylvester @hg.zetaworlds.com: 64 files, 12000 changes
[11:13 AM] Vincent.Sylvester @hg.zetaworlds.com: Gavin already has access to that stuff for testing as well
[11:13 AM] Andrew Hellershanks: Sounds good, and also ouch.
[11:13 AM] Vincent.Sylvester @hg.zetaworlds.com: So testing and review left to be done then it's ready for merge
[11:14 AM] Andrew Hellershanks: That's a lotof files and changes to get through.
[11:14 AM] Vincent.Sylvester @hg.zetaworlds.com: I did break it up as much as possible, but git is annoying when you wanna do that. I had more granular commits and then the core changes threw them on the pyre
[11:15 AM] Vincent.Sylvester @hg.zetaworlds.com: A lot of it is syntax sugar and updates to order of operations, so it's all over the place
[11:15 AM] Vincent.Sylvester @hg.zetaworlds.com: Rewriting and unifying the queries and how they are run, which might need even more updates down the line to add some more transactions and logical structure
[11:16 AM] Vincent.Sylvester @hg.zetaworlds.com: Doesn't help that seemingly no one can agree on the correct form of those either
[11:16 AM] Andrew Hellershanks: :)
[11:18 AM] Gavin.Hird @grid.xmir.org:8002: > Gavin already has access to that stuff for testing as well – yes but too busy building viewer for the time being to wenture into it this side of Christmas
[11:18 AM] Vincent.Sylvester @hg.zetaworlds.com: Most significant change is adding keys to some things and changing queries around to a simpler format that doesn't do internal readback which consumes a bunch of db time
[11:19 AM] Vincent.Sylvester @hg.zetaworlds.com: I did take another attempt at the prim inventory thing again, but not certain I like that approach any more than fully implementing delete and update routines based on the upper level data rather than reading things out of the db once over
[11:20 AM] Andrew Hellershanks: Hello, Cuga.
[11:20 AM] Cuga.Rajal @rajal.org:9000: Hi all =)
[11:22 AM] Andrew Hellershanks: Vincent, Always worth the effort when you can clean up things and make it code a bit simpler. Even better if it will reduce the time to do something.
[11:23 AM] Vincent.Sylvester @hg.zetaworlds.com: Beyond the actual bugs having the whole thing be as uniform as possible makes it easier to work on. Every service connector having different design choices is making you go code blind much sooner
[11:23 AM] Vincent.Sylvester @hg.zetaworlds.com: The sqlite especially has some weird stuff in there that makes little sense
[11:23 AM] Vincent.Sylvester @hg.zetaworlds.com: I updated all of them so if anyone wants to help testing I'll get them the link to the patches
[11:23 AM] Andrew Hellershanks: Not surprising. Some of that must be rather old legacy code.
[11:24 AM] Vincent.Sylvester @hg.zetaworlds.com: I'm not even certain all the spots have escapes now as the various handlers do all sorts of funky things to build queries
[11:25 AM] Vincent.Sylvester @hg.zetaworlds.com: Postgres had I'll say at least 5 bugs in there which broke functionality, though nothing critical to basic operation, mostly things rarely used
[11:25 AM] Vincent.Sylvester @hg.zetaworlds.com: Still if broken needs fixing
[11:26 AM] Gavin.Hird @grid.xmir.org:8002: Claude said a lot can be done to the queries but I wasn't going to go in to it at the time
[11:26 AM] Vincent.Sylvester @hg.zetaworlds.com: Yeah escaping things like name or whatever is needed if it someone gets on the train of reserved words cannot be used without escaping them
[11:27 AM] Vincent.Sylvester @hg.zetaworlds.com: Through the syntax parser it does understand some of that, but why take chances
[11:27 AM] Gavin.Hird @grid.xmir.org:8002: it also said that all search needed can be done in the existing tables with a few extensions. No need to go off database
[11:28 AM] Vincent.Sylvester @hg.zetaworlds.com: You mean the search for places and events?
[11:28 AM] Gavin.Hird @grid.xmir.org:8002: which is an opinion I have held for a long time
[11:28 AM] Gavin.Hird @grid.xmir.org:8002: yues
[11:29 AM] Vincent.Sylvester @hg.zetaworlds.com: Well yes and no. There are some things that yeah work best with connections to grid db, but not all data can or should be stored there
[11:29 AM] Gavin.Hird @grid.xmir.org:8002: why not?
[11:29 AM] Vincent.Sylvester @hg.zetaworlds.com: I have a new system for that which is based on the old one, but reduces the complexity a lot
[11:29 AM] Gavin.Hird @grid.xmir.org:8002: not in my opinion
[11:29 AM] Gavin.Hird @grid.xmir.org:8002: just makes it all more convoluted
[11:29 AM] Vincent.Sylvester @hg.zetaworlds.com: Things like parcel data, the host register, additional tables needed for that or at least adding to already pretty hefty tables
[11:30 AM] Gavin.Hird @grid.xmir.org:8002: hefty - compared to the stored blobs and textures it is nothing
[11:30 AM] Vincent.Sylvester @hg.zetaworlds.com: Well that's only if you use assets in db and the few things like terrain
[11:31 AM] Gavin.Hird @grid.xmir.org:8002: pretty much all event data can be stored in one 8k record
[11:31 AM] Vincent.Sylvester @hg.zetaworlds.com: Sure, but without external tools they cannot be managed or administrated anyways
[11:32 AM] Gavin.Hird @grid.xmir.org:8002: maybe not managed, but accessedm by regular users they can
[11:33 AM] Vincent.Sylvester @hg.zetaworlds.com: They cannot create them from the viewer unless you add that function to it, but yeah I'm more concerned about moderation side of things there, because this is all public facing data that you may want to moderate so that functionality also needs to be there and that's completely lacking for all the search data
[11:33 AM] Gavin.Hird @grid.xmir.org:8002: so let's add it
[11:34 AM] Vincent.Sylvester @hg.zetaworlds.com: The datasnapshot also needs updates and fixes and the connection to other tables would need to be done in code because at the moment the tables are all different collations which means you cannot join them directly without a massive time penalty
[11:34 AM] Gavin.Hird @grid.xmir.org:8002: I totally disagree 😀
[11:34 AM] Vincent.Sylvester @hg.zetaworlds.com: I have the external things basically done, but in a format that requires a framework so need to transplant that into plain php then it can ship along with the updated search itself directly into core
[11:35 AM] Gavin.Hird @grid.xmir.org:8002: anyway I will not look at it this year
[11:36 AM] Andrew Hellershanks: Vincent, sounds like it would be simpler to migrate the tables to a common collation.
[11:36 AM] Gavin.Hird @grid.xmir.org:8002: it is Andrew
[11:36 AM] Vincent.Sylvester @hg.zetaworlds.com: That's what I did, cause honestly utf8mb4 is not that much more data compared to latin1, not sure why Ubit is so stressed about changing that
[11:36 AM] Vincent.Sylvester @hg.zetaworlds.com: Once all unified the joins are ms rather than seconds
[11:37 AM] Andrew Hellershanks: That's a big savings
[11:37 AM] Gavin.Hird @grid.xmir.org:8002: and if we are going to fix up Postgres, it will have to be done properly otherwise there is no point
[11:37 AM] Cuga.Rajal @rajal.org:9000: A while back that mb4 was causing crashes
[11:37 AM] Cuga.Rajal @rajal.org:9000: I think thats why Ubit was hesitant. Not sure if that was resolved
[11:38 AM] Vincent.Sylvester @hg.zetaworlds.com: I could never replicate that. The prior limits to table scale are no longer a thing
[11:38 AM] Vincent.Sylvester @hg.zetaworlds.com: Been running unified tables on ZetaWorlds for 3+ years now
[11:38 AM] Cuga.Rajal @rajal.org:9000: maybe related to the font rendering? just a guess
[11:38 AM] Andrew Hellershanks: For my local standalone test set up I use either latin1 or utf8 as I don't have to deal with any characters from foreign languages.
[11:39 AM] Cuga.Rajal @rajal.org:9000: Same
[11:39 AM] Vincent.Sylvester @hg.zetaworlds.com: The emojis didn't work with certain charsets, but they work fine in utf8mb4, it has enough space for those
[11:39 AM] Vincent.Sylvester @hg.zetaworlds.com: It makes the tables slightly bigger cause there is more encoding stored in there, but that's what like maybe 5%
[11:40 AM] Vincent.Sylvester @hg.zetaworlds.com: I take that over a reduction in query speed for search by a factor of 1000
[11:40 AM] Andrew Hellershanks: I agree.
[11:40 AM] Cuga.Rajal @rajal.org:9000: make sense
[11:40 AM] Cuga.Rajal @rajal.org:9000: makes*
[11:41 AM] Andrew Hellershanks: hehe... interesting how one letter can change how two words can be interpreted.
[11:41 AM] Vincent.Sylvester @hg.zetaworlds.com: Ubit is reluctant to merge anything that depends on external things, which I get, but while a lot of search could sit in core directly, the moderation is where things get complicated and I'm not about to add a webserver in there just to serve that stuff. Much like webrtc some things living outside make more sense
[11:42 AM] Vincent.Sylvester @hg.zetaworlds.com: It's easier for the user to navigate a web page than trying to put those moderation tools into viewers that then need to communicate with that on some protocol
[11:42 AM] Andrew Hellershanks: /me nods
[11:42 AM] Gavin.Hird @grid.xmir.org:8002: it is just like the complex emojis like the flags. If I type a flag here now you will most likely see it as two or tree characters or placeholders 🇳🇴
[11:42 AM] Gavin.Hird @grid.xmir.org:8002: I added proper viewer support for it
[11:43 AM] Andrew Hellershanks: Gavin, I see as two characters.
[11:43 AM] Vincent.Sylvester @hg.zetaworlds.com: Given that wifi is now completely unmaintained anyways it might be all the more reason to build external tools for basic functions and moderation
[11:43 AM] Gavin.Hird @grid.xmir.org:8002: right, the rainbow flag is as much as 4 I think
[11:43 AM] Gavin.Hird @grid.xmir.org:8002: 🏳️🌈
[11:44 AM] Andrew Hellershanks: I was given code for destinations that could be accessed on a website. I have been too busy with other projects to check that the DB queries are being done safely.
[11:45 AM] Andrew Hellershanks: We are getting close to the top of the hour. Lyr, did you want to update us on what is happening regarding the OSCC?
[11:46 AM] Andrew Hellershanks: I think Lyr stepped away from her computer.
[11:46 AM] Lyr.Lobo @cc.opensimulator.org:8002: Thank you, Andrew
[11:46 AM] Jagga.Meridith @hg.zetaworlds.com: I'm presenting building LL with Claude and using Claude to document code. Test robot outside the door.
[11:47 AM] Lyr.Lobo @cc.opensimulator.org:8002: We're setting the schedule for OSCC today. If anyone has a session that they want to add, let me know
[11:47 AM] Andrew Hellershanks: ok, Go ahead, Lyr.
[11:47 AM] Gavin.Hird @grid.xmir.org:8002: nice, even LL has started using Claude in the viewer code now
[11:47 AM] Lyr.Lobo @cc.opensimulator.org:8002: I'll be sending letters out today and tomorrow with the scheduled times, etc.
[11:47 AM] Jagga.Meridith @hg.zetaworlds.com: what little I've seen of the code it's ...er...interesting
[11:48 AM] Gavin.Hird @grid.xmir.org:8002: the LL code?
[11:48 AM] Lyr.Lobo @cc.opensimulator.org:8002: we made a few changes this year and the expos are at the landing zones, etc., Training is Saturday at 9 am and again the next week at 9 am at OSCC in the keynote 1 region
[11:48 AM] Jagga.Meridith @hg.zetaworlds.com: yes
[11:48 AM] Lyr.Lobo @cc.opensimulator.org:8002: Many thanks! More details later
[11:48 AM] Lyr.Lobo @cc.opensimulator.org:8002: Thank you, Andrew
[11:49 AM] Vincent.Sylvester @hg.zetaworlds.com: Speaking of viewer side, apparently LL's math was wrong for years returning -10 and 10 as being equal xD Bunch of fixes going into FS that were merged upstream cause god damn I thought my code was bad
[11:49 AM] Andrew Hellershanks: Thank you, Lyr. Time is passing quickly. The OSCC is fast approaching.
[11:49 AM] Gavin.Hird @grid.xmir.org:8002: We are writing a lot of ObjC to raplce LL interface code + using the C++ underlying where we can for the grid and viewer logic
[11:49 AM] Lyr.Lobo @cc.opensimulator.org:8002: Yes, Dec 5-6 is the conference and Dec 4th, the music
[11:50 AM] Vincent.Sylvester @hg.zetaworlds.com: TJ found a number of pretty critical bugs in there, some even causing multiple ms per frame slow downs, so I dread to think what else lurks in there or what timing issues will pop up once other things are unfucked
[11:51 AM] Andrew Hellershanks: Oh, nuts. I forgot the date of the OSCC. That is the day of another online event I was planning on attending.
[11:51 AM] Gavin.Hird @grid.xmir.org:8002: it is pretty much the same Vincent Only one bit differs
[11:51 AM] Vincent.Sylvester @hg.zetaworlds.com: Tell that to the guys crashing their moon landers cause of a similar bit lol
[11:51 AM] Andrew Hellershanks: Vincent, whatisTJ?
[11:52 AM] Vincent.Sylvester @hg.zetaworlds.com: Hecklezz on discord
[11:52 AM] Vincent.Sylvester @hg.zetaworlds.com: FS dev
[11:52 AM] Andrew Hellershanks: ok.
[11:53 AM] Gavin.Hird @grid.xmir.org:8002: He has started to use a tool to find such issues. Claud said it would be very noisy on the LL codebase 😀
[11:53 AM] Gavin.Hird @grid.xmir.org:8002: the use of S32 and U32 is all over the place
[11:54 AM] Gavin.Hird @grid.xmir.org:8002: which is what he discovered
[11:54 AM] Vincent.Sylvester @hg.zetaworlds.com: It used to take me almost questioning my sanity to find things, now claude just hallucinates for a while and sends you off to find bugs that don't exist, but I suppose the code quality over at LL it doesn't have to be creative to find problems
[11:54 AM] Vincent.Sylvester @hg.zetaworlds.com: Still takes a dev to actually think about whether what it says makes sense
[11:54 AM] Andrew Hellershanks: /me nods
[11:55 AM] Gavin.Hird @grid.xmir.org:8002: yes but the S32 and U32 with (maybe) implicit casts is all over their codebase
[11:55 AM] Vincent.Sylvester @hg.zetaworlds.com: Structural things it finds a bit better from good pattern matching, but still lacks context given the size of the codebase
[11:56 AM] Andrew Hellershanks: I have been using Google's AI (in the search engine page) to help guide me on writing some device tree files for the Linux kernel and for some code to handle GPIO pins on an embedded system. I don't often take the code as is but it does save some a bit of time.
[11:56 AM] Vincent.Sylvester @hg.zetaworlds.com: I threw some code at it to transcribe from one db connector to another and it got a bit confused with the task, broke the input
[11:56 AM] Gavin.Hird @grid.xmir.org:8002: not really and where it doesn't it writes a lot of test to find out what is actually going on
[11:57 AM] Andrew Hellershanks: Having an AI analyze existing code is one thing I haven't tried.
[11:57 AM] Gavin.Hird @grid.xmir.org:8002: scary at first Andrew
[11:57 AM] Andrew Hellershanks: I would throw the lint tool at it first.
[11:57 AM] Vincent.Sylvester @hg.zetaworlds.com: I don't give it local access so it only knows what I throw at it and what it parsed prior so those issues I expect
[11:58 AM] Vincent.Sylvester @hg.zetaworlds.com: Keeps me vigilant I suppose
[11:58 AM] Andrew Hellershanks: Yes, these tools can be good for that. I also run my code through valgrind regularly to check for memory related issues in code.
[11:58 AM] Gavin.Hird @grid.xmir.org:8002: I only give it local access on a dedicated machine. No cloud access at all
[11:59 AM] Jagga.Meridith @hg.zetaworlds.com: Claude is good at analyzing existing code
[12:00 PM] Gavin.Hird @grid.xmir.org:8002: yes, but it cannot read all from the code, so we have some pretty lengthy hour long test sesions and iterations to find out what is actually going on wiht the code and how it shall work
[12:01 PM] Vincent.Sylvester @hg.zetaworlds.com: Oh one thing. For those that read the logs. I just spent 5 hours yesterday cleaning a bunch of notecards and other junk from a grid that had a bunch of satyr crap on there. Almost filled their disks up and created tens of millions of notecards and objects on regions. So I'd advise anyone to remove that stuff or change the code in there to not do that
[12:01 PM] Gavin.Hird @grid.xmir.org:8002: satyr?
[12:02 PM] Andrew Hellershanks: Thanks for the heads up on that, Vincent.
[12:02 PM] Jagga.Meridith @hg.zetaworlds.com: farm sim
[12:02 PM] Gavin.Hird @grid.xmir.org:8002: ok
[12:02 PM] Vincent.Sylvester @hg.zetaworlds.com: This has come up a couple times over the years. Some people write scripts that store data in notecards but never check if they actually have to, so just create on loop until there are millions. Similarly they often also self replicate objects which if left unchecked will kill a simulator pretty quickly once that exponential growth hits
[12:03 PM] Vincent.Sylvester @hg.zetaworlds.com: You can monitor for "script generated notecard" in the assets db to check if the numbers are exploding
[12:03 PM] Vincent.Sylvester @hg.zetaworlds.com: I believe by default the enforcement for object number limits are turned off, so enabling that on simulators might be a good idea
[12:04 PM] Vincent.Sylvester @hg.zetaworlds.com: Monitoring and backups :)
[12:04 PM] Gavin.Hird @grid.xmir.org:8002: should be on by default with a reasonable setting
[12:05 PM] Andrew Hellershanks: That would be a good idea.
[12:07 PM] Gavin.Hird @grid.xmir.org:8002: I have posted a preview build where the entire toolbar, stats bar, navigation bar and favorites have been rewritten native macOS and is no longer rendere by OpenGL
[12:07 PM] Gavin.Hird @grid.xmir.org:8002: today we are reweriting the script editor more or less from scratch
[12:10 PM] Anthropic.Claude @hg.zetaworlds.com: ow
[12:11 PM] Gavin.Hird @grid.xmir.org:8002: with themes galore
[12:11 PM] Andrew Hellershanks: Ouch. The dinosaur got me good this time.
[12:11 PM] Gavin.Hird @grid.xmir.org:8002: and follow cursor lookups on the ossl website
[12:11 PM] Gavin.Hird @grid.xmir.org:8002: which is poorly formatted for the purpose
[12:13 PM] Gavin.Hird @grid.xmir.org:8002: why is that site so slow? Is it running on a 286?
[12:13 PM] Vincent.Sylvester @hg.zetaworlds.com: Flooded by ai crawlers
[12:13 PM] Gavin.Hird @grid.xmir.org:8002: hmmm
[12:14 PM] Jagga.Meridith @hg.zetaworlds.com: when writing the editor, can we have a dark mode? these old eyes cannot see dark text on white background easily
[12:14 PM] Gavin.Hird @grid.xmir.org:8002: you can install .tm and vscode themes to your own liking
[12:16 PM] Jagga.Meridith @hg.zetaworlds.com: .tm = TextMate theme file?
[12:16 PM] Gavin.Hird @grid.xmir.org:8002: yep
[12:16 PM] Gavin.Hird @grid.xmir.org:8002: so you literally have hundreds if not thousands of options
[12:17 PM] Lyr.Lobo @cc.opensimulator.org:8002: sorry, is something wrong with the website? Opensimulator.org is not opening for me
[12:17 PM] Lyr.Lobo @cc.opensimulator.org:8002: I may have been in class when you mentioned it
[12:17 PM] Vincent.Sylvester @hg.zetaworlds.com: That's why there is a .dev mirror
[12:17 PM] Andrew Hellershanks: The system looks ok, Lyr. Load average is low and memory using is at about 75%
[12:18 PM] Gavin.Hird @grid.xmir.org:8002: that was faster. Is that permanent?
[12:18 PM] Lyr.Lobo @cc.opensimulator.org:8002: ooh that took a long time to load
[12:18 PM] Vincent.Sylvester @hg.zetaworlds.com: I maintain that, properly, so I been having some fun with trying to do anti flood on the crawlers
[12:19 PM] Gavin.Hird @grid.xmir.org:8002: I might pint the follow cursor lookup to that site then and gain some speed
[12:19 PM] Andrew Hellershanks: Lyr, I'm restarting the webserver.
[12:19 PM] Lyr.Lobo @cc.opensimulator.org:8002: thanks
[12:21 PM] Gavin.Hird @grid.xmir.org:8002: I think I need to leave. Thanks for a good meeting
[12:21 PM] Andrew Hellershanks: ok, Gavin.
[12:21 PM] Gavin.Hird @grid.xmir.org:8002: next week. cheers
[12:21 PM] Andrew Hellershanks: Any last items for today?
[12:22 PM] Andrew Hellershanks: Alrighty, then. Thank you all for coming. See you again next week.
