[11:02 AM] Andrew Hellershanks: Hello, everyone
[11:02 AM] Gavin.Hird @grid.xmir.org:8002: Hi Andrew
[11:02 AM] Vincent.Sylvester @hg.zetaworlds.com: I am not even sure if they have a character limit
[11:02 AM] Vincent.Sylvester @hg.zetaworlds.com: They are assets so its all converted to bytes
[11:02 AM] Gavin.Hird @grid.xmir.org:8002: I can try and paste one of the report into one and we shall find out
[11:03 AM] Vincent.Sylvester @hg.zetaworlds.com: Probably like integer limit of chars like 2 billion something
[11:03 AM] Vincent.Sylvester @hg.zetaworlds.com: Wouldn't surprise if the limit is the viewer crashing trying to display it xD
[11:04 AM] Vincent.Sylvester @hg.zetaworlds.com: Already got scripts that take a long time to display locking up the viewer cause people leave that preprocessor crap in there
[11:05 AM] Andrew Hellershanks: There is certainly a practical limit. What does the LSL wiki state? I'm sure we handle more than that. The limit may be imposed by how the data is stored in the database.
[11:08 AM] Vincent.Sylvester @hg.zetaworlds.com: It probably is a byte limit of sorts, but it could also be enforced on the viewer end alone. Possibly osmakenotecard would have something to check for that, but then again maybe not since that would hit heap limits probably
[11:09 AM] Vincent.Sylvester @hg.zetaworlds.com: If limits are being hit something is wrong imo, most cases you'd probably be better off doing something different than trying to squeeze it that much
[11:09 AM] Vincent.Sylvester @hg.zetaworlds.com: Not that I haven't raised heap to fiddle with stuff, but the limits usually exist for a reason
[11:10 AM] Fiona.Sweet @hg.zetaworlds.com: hi. sorry somehow i ended up at 0,0 or something
[11:10 AM] Andrew Hellershanks: np, Fiona. You found your way back here.
[11:10 AM] Fiona.Sweet @hg.zetaworlds.com: yes ty
[11:13 AM] Andrew Hellershanks: The log of OS indicates no new bugs in the past week. ;)
[11:13 AM] Ubit Umarov: Ohh yes.. isn-t that great?
[11:13 AM] Andrew Hellershanks: It is indeed.
[11:14 AM] Vincent.Sylvester @hg.zetaworlds.com: Lazy devs been finding bugs elsewhere xD
[11:14 AM] Andrew Hellershanks: It means fewer things to talk about.
[11:14 AM] Ubit Umarov: did look a bot to the pqsql ai report...
[11:14 AM] Fiona.Sweet @hg.zetaworlds.com: hehe
[11:15 AM] Gavin.Hird @grid.xmir.org:8002: look at the mysql one first
[11:15 AM] Ubit Umarov: was just telling gavin about on report
[11:15 AM] Ubit Umarov: one..
[11:15 AM] Ubit Umarov: (CreateParameterValue(value, "boolean"/"bit")): does value.ToString() == "true". A C# bool's ToString() is "True"/"False" (capitalized)
[11:16 AM] Vincent.Sylvester @hg.zetaworlds.com: case sensitivity is so annoying right xD
[11:16 AM] Ubit Umarov: the silly thing missed that boolean is a pqslq bool not c# so its tostring is in fact lower case
[11:16 AM] Ubit Umarov: pfff
[11:16 AM] Ubit Umarov: ofc several things about injection...
[11:16 AM] Vincent.Sylvester @hg.zetaworlds.com: Should just do caseinvariant blah on the compare really for when they get funny ideas like sqlite did
[11:17 AM] Ubit Umarov: who cares if those dbs should be closed to humans
[11:17 AM] Vincent.Sylvester @hg.zetaworlds.com: Some of it is on private port yes, but you know how some are "firewalls too complicated, open all the things"
[11:17 AM] Ubit Umarov: if any human other than owner gets there, then sql injection the the least of problems
[11:18 AM] Ubit Umarov: ...is..
[11:18 AM] Fiona.Sweet @hg.zetaworlds.com: oh yes the tricky nftables
[11:18 AM] Gavin.Hird @grid.xmir.org:8002: it is a problem because some grids have economy, some grids store private information about members and some people wants to be malicious.
[11:19 AM] Gavin.Hird @grid.xmir.org:8002: not sure why I shouted?
[11:19 AM] Ubit Umarov: well ill try to read all that and make new worse bugsd
[11:19 AM] Vincent.Sylvester @hg.zetaworlds.com: Best not look at wifi code then, the holes are craters lol
[11:19 AM] Ubit Umarov: err new fixes
[11:19 AM] Gavin.Hird @grid.xmir.org:8002: do I shout now too?
[11:19 AM] Fiona.Sweet @hg.zetaworlds.com: drooled on shift key and it stuck
[11:20 AM] Ubit Umarov: yeah.. happens when we dont chat much.. when we do we shout :)
[11:20 AM] Ubit Umarov: (rl also )
[11:20 AM] Gavin.Hird @grid.xmir.org:8002: we can pick some candidates and let claude develop patches and see how it does
[11:20 AM] Fiona.Sweet @hg.zetaworlds.com: hopefully claude likes us
[11:20 AM] Vincent.Sylvester @hg.zetaworlds.com: Most of the things in the reports there are easy fixes, don't waste water on that
[11:21 AM] Ubit Umarov: bahh those things are blind to reality
[11:21 AM] Andrew Hellershanks: Sure. Just don't skip reviewing its code before trying out anything it suggests
[11:21 AM] Fiona.Sweet @hg.zetaworlds.com: what to do when AI hates you
[11:21 AM] Fiona.Sweet @hg.zetaworlds.com: hehe
[11:21 AM] Ubit Umarov: know all on blogs etc.. not real code
[11:21 AM] Andrew Hellershanks: Might be better of using a lint tool for C# (if there is one). Can't think of one off hand.
[11:22 AM] Cuga.Rajal @rajal.org:9000: Were these from a Mantis? Who wrote them?
[11:22 AM] Ubit Umarov: gavin played with his ai
[11:22 AM] Gavin.Hird @grid.xmir.org:8002: who wrote what?
[11:22 AM] Cuga.Rajal @rajal.org:9000: The report... ok I got answer. AI bot
[11:23 AM] Gavin.Hird @grid.xmir.org:8002: Claude wrote them. Each were a 15-20 min run with Sonnet 5 model
[11:23 AM] Cuga.Rajal @rajal.org:9000: ahhh ok
[11:24 AM] Lyr.Lobo @vharmonyartshub.com:8002: /me smiles
[11:25 AM] Cuga.Rajal @rajal.org:9000: In perfect world, we would have a test suite to check for each of these exploits and validate code
[11:25 AM] Gavin.Hird @grid.xmir.org:8002: what exploits?
[11:26 AM] Vincent.Sylvester @hg.zetaworlds.com: Speaking of. I did get a bit further with libgdiplus. The flaw on the font collection being only metadata I could find, but the test still failed. Apparently the fonts need to register functions, which is supposed to be done by the create function as well, but for some reason that's not happening. Adding an explicit function register works and test passes, but obviously it's not supposed to be needed. Escalated it to the harfbuzz folks to get some insight as to whether that's a bug or just misunderstanding. Pending response.
[11:26 AM] Gavin.Hird @grid.xmir.org:8002: there are just issues that needs cosmetic changes
[11:26 AM] Cuga.Rajal @rajal.org:9000: ah ok. I meand, vulnerablities
[11:27 AM] Vincent.Sylvester @hg.zetaworlds.com: There is no end to probing open http endpoints for things. If you want to do damage you can just overload it good ole script kiddy style
[11:27 AM] Fiona.Sweet @hg.zetaworlds.com: I got a bunch of NonFinite Avatar position errors and crashed the other day haven't tracked down that problem but it says report to Mantis hehe
[11:27 AM] Cuga.Rajal @rajal.org:9000: Oh thats hopeful news about libgdiplus. fingers crossed
[11:28 AM] Vincent.Sylvester @hg.zetaworlds.com: It says that because if it continues to happen it could be an engine bug, but usually it is caused by bad mesh interaction or just overloading the engine with overly complex things
[11:28 AM] Vincent.Sylvester @hg.zetaworlds.com: People uploading mesh with broken physics etc. or apparently straight up bypassing all sanity checks and dumping raw data in until everything blows up
[11:28 AM] Andrew Hellershanks: Vincent, I read there may be an order of operations issue and/or that something was only done once but never again that leads to some of the font related tests dying.
[11:30 AM] Andrew Hellershanks: There are about three ways to run the libgdiplus tests. What I found is that the order in which they run is partly responsible for whether the tests succeed or not.
[11:30 AM] Vincent.Sylvester @hg.zetaworlds.com: It could be, but I haven't gotten a clear answer yet, so I don't really know how to proceed from there, especially given the clean code requirement wine has regarding llm code, which I did technically look at for the last part
[11:30 AM] Vincent.Sylvester @hg.zetaworlds.com: I couldn't get that gtest thing to work, I tried compiling it but it failed
[11:30 AM] Andrew Hellershanks: Compiled fine for me.
[11:31 AM] Vincent.Sylvester @hg.zetaworlds.com: So just using the make check and tests from the folder they all pass with the changes. Much like our tests if something isn't cleaned up properly they cannot be re-run on the same thing
[11:31 AM] Vincent.Sylvester @hg.zetaworlds.com: gcc would give me a compile error on gtest.cpp
[11:31 AM] Andrew Hellershanks: Might be that I missed a config option and don't have all features compiled in and ready for testing.
[11:32 AM] Ubit Umarov: (site is up again )
[11:32 AM] Vincent.Sylvester @hg.zetaworlds.com: Think wine also only looks at the make check passing, at least that's all MS ever cared about and even then they were happy to release a broken version to Canonical and others
[11:32 AM] Vincent.Sylvester @hg.zetaworlds.com: So now all upstream libgdiplus is tainted
[11:32 AM] Vincent.Sylvester @hg.zetaworlds.com: Raising my blood pressure again
[11:33 AM] Vincent.Sylvester @hg.zetaworlds.com: I'll give it another poke when it gets a little colder and then maybe can try and get a fix pushed through
[11:34 AM] Ubit Umarov: "ai can be clever .. "... this one method looks copy-pasted from MySQLAssetData.cs, which legitima..."
[11:34 AM] Andrew Hellershanks: I just looked again. I didn't add --with-pango during my first builds.
[11:34 AM] Ubit Umarov: seems all ppl do the same c@p mistake :)
[11:34 AM] Ubit Umarov: even c&p
[11:35 AM] Vincent.Sylvester @hg.zetaworlds.com: Attempted commonality between the connectors to make it easier
[11:35 AM] Vincent.Sylvester @hg.zetaworlds.com: Over ten people worked on this over the years some of it is all over the place
[11:36 AM] Vincent.Sylvester @hg.zetaworlds.com: Sometimes easier just to rip something out, replace with a working thing and adjust than to try and salvage what's there
[11:36 AM] Andrew Hellershanks: Especially true if the existing thing is poorly written and/or documented.
[11:37 AM] Vincent.Sylvester @hg.zetaworlds.com: sqlite changes are worse cause of the different versions changing major parts of the behavior so you have to be extra cautious to write compatible code
[11:37 AM] Fiona.Sweet @hg.zetaworlds.com: well somebody has a broken asset server i think, about 18% of assets had an \0 null character at the end of data. so MS XML parser barfs. stripping out the null byte made the assets work again
[11:38 AM] Fiona.Sweet @hg.zetaworlds.com: not sure where they are coming from but maybe using C and sending an extra character hehe
[11:38 AM] Vincent.Sylvester @hg.zetaworlds.com: Should be much less bad stuff floating around after osg deleted most of it
[11:39 AM] Fiona.Sweet @hg.zetaworlds.com: well removing the null char seems to make the assets good again
[11:39 AM] Vincent.Sylvester @hg.zetaworlds.com: Old aurora stuff is not compatible either and got mixed in with some surviving to this day
[11:39 AM] Fiona.Sweet @hg.zetaworlds.com: yes could be some legacy assets in the wild
[11:40 AM] Gavin.Hird @grid.xmir.org:8002: some?
[11:40 AM] Vincent.Sylvester @hg.zetaworlds.com: New bad stuff is not impossible either given the lengths some go to to import broken crap
[11:41 AM] Vincent.Sylvester @hg.zetaworlds.com: A bad alpha breaking baking isn't visible to the user, they don't see the log file blowing up from all the messages
[11:42 AM] Vincent.Sylvester @hg.zetaworlds.com: Can't feel the network card melt from sending a mesh body of 200+ megabytes from a bad upload including ten copies of the same mesh for no reason
[11:42 AM] Vincent.Sylvester @hg.zetaworlds.com: Can't test for everything in something this complex. The users will find a way to break it yet still
[11:43 AM] Fiona.Sweet @hg.zetaworlds.com: hehe yes
[11:44 AM] Fiona.Sweet @hg.zetaworlds.com: i did a frequency distribution on asset sizes and most are 1-10mb as i recall. not many huge ones but they do exist
[11:44 AM] Andrew Hellershanks: Yes, they will.
[11:44 AM] Gavin.Hird @grid.xmir.org:8002: Can let Claude look at the code involved and see what it says?
[11:45 AM] Vincent.Sylvester @hg.zetaworlds.com: You can, but it's unlikely to be merged one to one or at all. Becomes a problem of intellectual property
[11:46 AM] Vincent.Sylvester @hg.zetaworlds.com: Ai assisted and human reviewed, but best to just let it state what's wrong, cause usually the fixes aren't that hard to implement
[11:46 AM] Gavin.Hird @grid.xmir.org:8002: why a problem
[11:47 AM] Vincent.Sylvester @hg.zetaworlds.com: Cause ai code cannot be copyrighted and thus no license can be applied, the authorship isn't clear
[11:48 AM] Vincent.Sylvester @hg.zetaworlds.com: For the problems in those reports, just from skimming through, they are a couple minutes of work each, so no point in wasting water on ai when someone can just fix them
[11:48 AM] Lyr.Lobo @vharmonyartshub.com:8002: "Intellectual property issues arise because works predominantly generated by AI without meaningful human authorship are not eligible for copyright protection. This means that if Claude generates code, it may not be protected under copyright law, leading to potential legal challenges regarding ownership and use of that code in open-source projects." From news.ycombinator.com
[11:49 AM] Fiona.Sweet @hg.zetaworlds.com: and if the no fakes act passes will possibly have trouble with our heads in os hehe
[11:50 AM] Lyr.Lobo @vharmonyartshub.com:8002: /me smiles and nods
[11:50 AM] Fiona.Sweet @hg.zetaworlds.com: get sued by celebrities for being look-alike
[11:50 AM] Andrew Hellershanks: Vincent, I just configured and rebuilt the libgdiplus library and tests. tests/gtest usually dies when doing "5 tests from MatrixTests" with the messages -> gtest: fccache.c:795: FcCacheFini: Assertion `fcCacheChains[i] == NULL' failed. Aborted (core dumped)
[11:50 AM] Gavin.Hird @grid.xmir.org:8002: The patches are mine if they are developed for me. The only thing it has done a few times is adding a note to the commit it was developed in cooperation with Claude
[11:51 AM] Vincent.Sylvester @hg.zetaworlds.com: Andrew, dump me the commands used in discord, I'd like to try that myself and see if I can find the fix for that. I gave you the ones I used for pulling libgdi from wine and build, they seem different from your approach
[11:52 AM] Vincent.Sylvester @hg.zetaworlds.com: Need to get on the same page there to figure out what's going on or we'll go in circles heh
[11:53 AM] Andrew Hellershanks: The few times I used an AI for coding help I used what was shown to me as a guide. I usually did better than the AI in the final code but it showed me a path to take.
[11:53 AM] Andrew Hellershanks: Vincent, will do.
[11:54 AM] Gavin.Hird @grid.xmir.org:8002: Claude totally removed all pathfinding, marketplace and experience code from the viewer + we also did a very difficutl revert of code that started in 2018
[11:54 AM] Vincent.Sylvester @hg.zetaworlds.com: We much prefer ai simply stating a problem and possible solution for human evaluation rather than dumping code into a commits or branches that need days to review. There are more considerations to be done in a lot of cases because of how things interact or what the usage environment actually is. It can guess, but unless you critically review it then it becomes Schrödingers code
[11:54 AM] Gavin.Hird @grid.xmir.org:8002: marketplace was a difficult removal as it fanned all over
[11:55 AM] Vincent.Sylvester @hg.zetaworlds.com: The viewer code is a much bigger mess, thankfully os code is usually not that convoluted
[11:55 AM] Gavin.Hird @grid.xmir.org:8002: it is indeed
[11:56 AM] Vincent.Sylvester @hg.zetaworlds.com: If it can remove or rather replace the "Linden Parcel" or whatever in the parcel category in an SL compatible manner I might be impressed, cause that's one big can of worms to open
[11:56 AM] Gavin.Hird @grid.xmir.org:8002: this is what it have added to some commits. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
[11:56 AM] Gavin.Hird @grid.xmir.org:8002: not a pip about copyright
[11:57 AM] Lyr.Lobo @vharmonyartshub.com:8002: That is because the copyright is not defensible at this time
[11:57 AM] Vincent.Sylvester @hg.zetaworlds.com: The copyright is with the millions of stackoverflow posts, official docs and whatever else it ingested or is it with the company that set the model up and trained it, or maybe it even is the ai itself, but then what license would it want applied... that's just a whole conundrum to avoid by just not having it author code directly
[11:58 AM] Vincent.Sylvester @hg.zetaworlds.com: Advisory is much simpler as consultants don't get code credit
[11:58 AM] Gavin.Hird @grid.xmir.org:8002: also I could just amend the commit and remove the information and nobody would know
[11:58 AM] Vincent.Sylvester @hg.zetaworlds.com: Until someone manages to replicate the code with a prompt to a point even the comments are the same
[11:58 AM] Lyr.Lobo @vharmonyartshub.com:8002: A summary of the problem on who owns Claude-generated code https://aiproductivity.ai/news/who-owns-code-written-by-claude-code/
[11:59 AM] Lyr.Lobo @vharmonyartshub.com:8002: The safest practical position right now: treat AI-generated code as a first draft.
[11:59 AM] Lyr.Lobo @vharmonyartshub.com:8002: The Copyright Office is expected to publish further guidance on AI authorship in 2026.
[11:59 AM] Lyr.Lobo @vharmonyartshub.com:8002: Summary
[11:59 AM] Lyr.Lobo @vharmonyartshub.com:8002: I teach AI research classes at the doctoral level
[12:00 PM] Gavin.Hird @grid.xmir.org:8002: good thing I am no Norway so US copyright law does not apply here
[12:00 PM] Fiona.Sweet @hg.zetaworlds.com: well the other thing is no safe harbor unless you are registered with the copyright office
[12:00 PM] Vincent.Sylvester @hg.zetaworlds.com: This is a current hot topic and a problem so most want to avoid having to either take a stance or to have code in their projects they may have to rip out later. I mean what do you do if the fix is a single capital letter, but ai found it and now all ai code has to be removed. Can you then no longer fix this typo, because it would be under the ai copyright
[12:00 PM] Vincent.Sylvester @hg.zetaworlds.com: Do you then change it to a slower function just so the code is different. What if it compiles to the same code after optimization, does that count?
[12:01 PM] Vincent.Sylvester @hg.zetaworlds.com: This goes so far if you start spinning that it just ends up taking up a lot of time to consider
[12:01 PM] Lyr.Lobo @vharmonyartshub.com:8002: for idea generation and personal development, it is a great aide. For final code generation, uncertain
[12:02 PM] Vincent.Sylvester @hg.zetaworlds.com: Not to mention the whole bias of prompt fulfillment over adhering to code standards or sum such. Like I recall asking it for a solution and it kept going at it by brute force when really it was just easier to rewrite the whole thing proper
[12:02 PM] Vincent.Sylvester @hg.zetaworlds.com: It, much like a human, may or may not get the bigger picture correctly
[12:02 PM] Fiona.Sweet @hg.zetaworlds.com: it maybe doesn't apply but what _can_ happen is you get sued in the US and lose by default and probably won;t lose assets but they can do stuff with visa/american express/master card and paypal possibly making it hard to collect money from customers
[12:03 PM] Gavin.Hird @grid.xmir.org:8002: good luck getting the bigger picture correctly on the viewer code :-))
[12:03 PM] Vincent.Sylvester @hg.zetaworlds.com: Going at it critical trying to find flaws much like you would a junior dev might be the best approach you can currently take, cause you cannot know if it didn't just read the stackoverflow question instead of the answer
[12:04 PM] Vincent.Sylvester @hg.zetaworlds.com: At least claude is less braindead than gpt, but not by much. Just enough to not be annoying
[12:04 PM] Fiona.Sweet @hg.zetaworlds.com: anyway i need to run take care everyone
[12:04 PM] Lyr.Lobo @vharmonyartshub.com:8002: bye Fiona
[12:04 PM] Lyr.Lobo @vharmonyartshub.com:8002: have a great week
[12:04 PM] Ubit Umarov: have fun
[12:04 PM] Cuga.Rajal @rajal.org:9000: I need to head out as well, have a great week
[12:04 PM] Fiona.Sweet @hg.zetaworlds.com: thank you too
[12:04 PM] Vincent.Sylvester @hg.zetaworlds.com: Even so with the instructions it has to not deal with security topics and reject any attempt to dig into those when it does come back with recommendations for that it gets muddy
[12:04 PM] Cuga.Rajal @rajal.org:9000: waves and poofs
[12:05 PM] Lyr.Lobo @vharmonyartshub.com:8002: /me waves
[12:05 PM] Vincent.Sylvester @hg.zetaworlds.com: They went pretty hard on the side of not letting it deal with cybersecurity fearing it might expose something or aid hackers
[12:05 PM] Lyr.Lobo @vharmonyartshub.com:8002: you are pioneering it for the viewer, Gavin, and we are glad to hear of your discoveries
[12:05 PM] Gavin.Hird @grid.xmir.org:8002: In any case you have a bunch of findings. You can write the fix yourself, or we can generate fixes for review as a starting point
[12:06 PM] Vincent.Sylvester @hg.zetaworlds.com: When I'm not melting I will take a look
[12:06 PM] Lyr.Lobo @vharmonyartshub.com:8002: It is very hot here too
[12:06 PM] Vincent.Sylvester @hg.zetaworlds.com: Most of the stuff from skimming through is easy to fix
[12:06 PM] Andrew Hellershanks: D*mn hot here too
[12:07 PM] Gavin.Hird @grid.xmir.org:8002: It has been very helpful Lyr till now, but we have not really generated new functionality for now. Removed and fixed existing. I plan to use it to generate a lot of Mac specific code. Will see how far we get
[12:07 PM] Lyr.Lobo @vharmonyartshub.com:8002: yes
[12:08 PM] Lyr.Lobo @vharmonyartshub.com:8002: that's great, let us know how it goes
[12:08 PM] Gavin.Hird @grid.xmir.org:8002: sure will
[12:08 PM] Lyr.Lobo @vharmonyartshub.com:8002: exploring newer territory is important
[12:08 PM] Lyr.Lobo @vharmonyartshub.com:8002: /me waves and grins
[12:08 PM] Lyr.Lobo @vharmonyartshub.com:8002: I must dash, take care
[12:08 PM] Vincent.Sylvester @hg.zetaworlds.com: Especially with more and more viewer code from LL seemingly exploring the depths of human stupidity lol
[12:08 PM] Gavin.Hird @grid.xmir.org:8002: would be interesting to have it generate an admin frontend for opensim, and also look at search functionality
[12:09 PM] Vincent.Sylvester @hg.zetaworlds.com: Someone has been writing one for a while
[12:09 PM] Gavin.Hird @grid.xmir.org:8002: who is someone
[12:09 PM] Vincent.Sylvester @hg.zetaworlds.com: It looks promising albeit a bit basic still
[12:09 PM] Vincent.Sylvester @hg.zetaworlds.com: I'll send you a link
[12:10 PM] Gavin.Hird @grid.xmir.org:8002: thanks
[12:10 PM] Vincent.Sylvester @hg.zetaworlds.com: Search actually has major code flaws on both server and region side
[12:10 PM] Andrew Hellershanks: Tampa, I sent you the information you requested via Discord.
[12:10 PM] Vincent.Sylvester @hg.zetaworlds.com: The heat kinda killed the effort getting that fixed, cause my brain cannot process that much stupid at once
[12:11 PM] Gavin.Hird @grid.xmir.org:8002: as I havge said before I really cannot see why pretty much all search can be canned in the db for easy access, but that is just me
[12:11 PM] Gavin.Hird @grid.xmir.org:8002: postgres has more functionality to support it
[12:12 PM] Vincent.Sylvester @hg.zetaworlds.com: The flaws exist with the way the data is queried from regions. Then there are bugs in parsing on the server end and some fairly idiot concepts used to store it. Also some cross references that aren't done because it attempts to be self-contained which just causes further problems
[12:13 PM] Andrew Hellershanks: I will have to take another look at the search module some day. It worked the last time I used. Would be nice to have an enhanced version that can search across multiple categories of things at one time.
[12:13 PM] Vincent.Sylvester @hg.zetaworlds.com: There is still a conceptual issue even in the fixed version given the way data is retained when the source becomes unavailable
[12:13 PM] Gavin.Hird @grid.xmir.org:8002: where is the code Andrew?
[12:13 PM] Andrew Hellershanks: In my case it would be the add on OpenSimSearch module in gitlab.
[12:14 PM] Gavin.Hird @grid.xmir.org:8002: link?
[12:14 PM] Vincent.Sylvester @hg.zetaworlds.com: Problem is making changes on os side to fix this means the server end needs to pull along with it and that's harder to accomplish
[12:15 PM] Vincent.Sylvester @hg.zetaworlds.com: I have started the rewrite for that, but it's a lot of stuff to get done
[12:15 PM] Vincent.Sylvester @hg.zetaworlds.com: Whole xmlrpc handling and managing databases
[12:15 PM] Andrew Hellershanks: Gavin, I'm looking
[12:15 PM] Gavin.Hird @grid.xmir.org:8002: ok
[12:17 PM] Vincent.Sylvester @hg.zetaworlds.com: Could probably have claude rewrite the ci version into plain php, but they don't give enough tokens for the free users to do something that complex
[12:17 PM] Andrew Hellershanks: Firestorm is causing a high CPU load today and Firefox does the same when it is running. The two combined make it slow to look something up in Firefox. :P
[12:17 PM] Gavin.Hird @grid.xmir.org:8002: I have $20 per month plan and it has not been anywhere close to consume the limits yet
[12:18 PM] Vincent.Sylvester @hg.zetaworlds.com: I'll take you up on that when I get around to it, when it gets colder. Need brain function to make sense of it :)
[12:18 PM] Gavin.Hird @grid.xmir.org:8002: when it analyzed MySQL it generated 340k tokens
[12:22 PM] Andrew Hellershanks: Gavin, gitlab wants to send me a code so I can complete the log in. I need to start my email program to get the code. I now have more load on my system while I wait for it to download 227 email messages. (been a few days since I picked up mail on my desktop).
[12:22 PM] Gavin.Hird @grid.xmir.org:8002: don't worry right now Andrew
[12:24 PM] Andrew Hellershanks: Gavin, https://gitlab.com/kcozens/OpenSimSearch.git
[12:24 PM] Gavin.Hird @grid.xmir.org:8002: brilliant! bookmarked
[12:25 PM] Andrew Hellershanks: I have some other OpenSim modules there too.
[12:25 PM] Gavin.Hird @grid.xmir.org:8002: I'll check it out Andrew.
[12:25 PM] Gavin.Hird @grid.xmir.org:8002: I'll have to leave for now. See you next week I suppose
[12:25 PM] Vincent.Sylvester @hg.zetaworlds.com: Your profile is private, they can't see other projects
[12:26 PM] Gavin.Hird @grid.xmir.org:8002: Cheers
[12:26 PM] Vincent.Sylvester @hg.zetaworlds.com: Search really just needs fixes in the module and support in os base to function correctly. Then just ship the server code as a zip. At least that's the plan I have for it for now
[12:26 PM] Andrew Hellershanks: I didn't think the projects were private.
[12:27 PM] Vincent.Sylvester @hg.zetaworlds.com: They aren't, but the list is. Only direct links to them are visible
[12:27 PM] Andrew Hellershanks: I'll have to look at that some other time. Don't know why gitlab did that.
[12:28 PM] Andrew Hellershanks: That's the meeting done for another week with several people already having departed.
[12:29 PM] Andrew Hellershanks: I need to get back to some RL work. See you again next week.
