Chat log from the meeting on 2021-06-01

From OpenSimulator

Jump to: navigation, search

The meeting was held in Hurliman Plaza because the Dev Outreach region was not accessible.

[11:08] Ubit Umarov: no that is not a thing for ini files
[11:08] Kayaker Magic: I've heard complaints about CPU bound tasks being alowed to run for SECONDS
[11:09] Ubit Umarov: yes some may do that
[11:09] Ada.Radius Hi Andrew
[11:09] Andrew Hellershanks: Hello, everyone.
[11:09] Ada.Radius Hi Jamie, just saw you there
[11:09] Gavin.Hird Hi everyone
[11:09] Ubit Umarov: always a issue with non coop multitaks
[11:09] Jamie.Jordan hi everybody
[11:10] Ubit Umarov: i mean with co op  multitask
[11:10] Andrew Hellershanks: Here we are at our alternate location. I wonder how many people might miss us today. Some of the newer people may not know about this alternate meeting location.
[11:10] Ada.Radius I sent a couple of lms
[11:11] Gavin.Hird yeah, I just experienced non coop multitask when no meshes loaded here. Viewer probably have a "might do" loop
[11:11] Ubit Umarov: guess outreach did not like my commits
[11:11] Ubit Umarov: fun region is up
[11:11] Ubit Umarov: but tp fails
[11:11] Ada.Radius yup, with some inconsistent error messages. Mostly the failure to connect one
[11:13] Ubit Umarov: well lmost all usual ppl is here already
[11:13] Ubit Umarov: except selby
[11:13] Andrew Hellershanks: The only changes this past week seem to be mainly OSSL related. Not sure what may have broken something.
[11:13] Ada.Radius Owl is on his way
[11:13] Andrew Hellershanks: I don't see Selby online.
[11:13] Ada.Radius hi Owl!
[11:14] Jamie.Jordan Selby is afk today
[11:14] Kayaker Magic: Hi there Owl!
[11:14] Andrew Hellershanks: Hello, Owl.
[11:14] Gavin.Hird Hi Owl
[11:14] Ubit Umarov: yeah no changes related to tps
[11:14] Owl.Eyes ello, everyone *smiles around*
[11:14] Owl.Eyes *+H
[11:14] Andrew Hellershanks: Gavin, you are still a cloud.
[11:15] Ubit Umarov: well and about those code changes...
[11:15] Ada.Radius i see Gavin's av
[11:15] Ubit Umarov: they where mostly on lsl List support things
[11:15] Andrew Hellershanks: Gavin just popped in.
[11:15] Gavin.Hird I rebaked for you
[11:15] Andrew Hellershanks: Gavin, ty. It worked.
[11:15] Gavin.Hird great :-)
[11:15] Andrew Hellershanks: Ubit, yes I noticed that.
[11:15] Ubit Umarov: tried to save some us here and there
[11:16] Ubit Umarov: but, behvoiur of a funtion did change
[11:16] Andrew Hellershanks: us as in microseconds (in case someone reading this later gets confused by what Ubit just said).
[11:17] Ubit Umarov: ofc don't remember the fucntion name :p
[11:17] Andrew Hellershanks: I've pointed out some things that could save time in the past and was told they didn't matter.
[11:17] Ubit Umarov: llDeleteSubList( list src, integer start, integer end );
[11:18] Ubit Umarov: this one
[11:18] Andrew Hellershanks: Ubit, If you updated the wiki page it should be ok.
[11:18] Ubit Umarov: the  case start > end is now diferent
[11:18] Andrew Hellershanks: What behavioural change did you make?
[11:18] Ubit Umarov: it is like SL now
[11:19] Andrew Hellershanks: The sart > end should have been its normal use case.
[11:19] Ubit Umarov: our code was returning:  [0, end] + [start, -1]
[11:19] Ubit Umarov: where -1 means to end of string
[11:19] Andrew Hellershanks: oh, sorry. The backwards case.
[11:19] Ubit Umarov: now it returnts [end +\, start -1]
[11:19] Ubit Umarov: ie the oposite range
[11:20] Ubit Umarov: err
[11:20] Ubit Umarov: now it returnts [end +1, start -1]
[11:20] Ubit Umarov: start > end is not the normal case :p
[11:22] Andrew Hellershanks: Definitely not. MIght be useful in some situations. I haven't found a need for it yet.
[11:22] Ubit Umarov: also added void osListSortInPlace(LSL_List src, LSL_Integer stride, LSL_Integer ascending);
[11:22] Ubit Umarov: yeah well the SL wiki is confusing on that
[11:22] Andrew Hellershanks: yes, it is.
[11:22] Ubit Umarov: so we where not doing it right
[11:22] Ubit Umarov whispers: guess no one actually uses that.. :)
[11:23] Ubit Umarov: it does melt neurons :)
[11:23] Andrew Hellershanks: You will find out about the one person who is using it when they complain you broke their script. ;)
[11:23] Ubit Umarov: well the new function is just like the ll one
[11:23] Ubit Umarov: but does not copy the list
[11:23] Ubit Umarov: to saving a bit memory
[11:23] Ubit Umarov: so..
[11:23] Ubit Umarov: it just does the sort on the list itself
[11:24] Ubit Umarov: i also changed the sort method a big, getting like a 4x speed up on some case
[11:24] Ubit Umarov: note that is disturbing..
[11:24] Andrew Hellershanks: That is definitely worth having.
[11:25] Andrew Hellershanks: Can you apply the sort method change to the normal ll version?
[11:25] Ubit Umarov: if the type of things on the list is all the same, and stride is 1
[11:26] Ubit Umarov: the native .net Array.Sort optimized thing is used
[11:26] Ubit Umarov: if not a simple crude ( but needed ) bubble sort is used
[11:27] Ubit Umarov: that is like 200x slower on a 50000 integers list
[11:27] Ubit Umarov: ( forced by hadn to use one or the other )
[11:28] Gavin.Hird Hi Bill
[11:28] Andrew Hellershanks: Would quicksort also expect all items to be of same type?
[11:28] Ubit Umarov: well that is bc Array sort is now optimized
[11:28] Bill Blight: howdy
[11:28] Andrew Hellershanks: Hello, Bill.
[11:28] Ubit Umarov: even uses one of 3 methods acording to the case
[11:28] Owl.Eyes Hi Bill
[11:28] Bill Blight: greetings all
[11:29] Ubit Umarov: quincksort one of those, ofc
[11:29] Bill Blight wipes of residual server chunks off his boots
[11:29] Ubit Umarov: quick..
[11:29] Ubit Umarov: Bill :)
[11:29] Jamie.Jordan hey Bill
[11:29] Andrew Hellershanks grins at Bill
[11:29] Bill Blight: Had to kick the server hard ..
[11:29] Owl.Eyes chuckles
[11:29] Ubit Umarov: well so a small speed up on lists
[11:30] Ubit Umarov: and that fix(?) on DeleteSublist
[11:30] Gavin.Hird It's like the old TV sets, you had to smack hard to get in Sweden
[11:30] Ubit Umarov: btw replaceList was like DeleteSublist is now
[11:30] Ubit Umarov: on the considered range is start > end
[11:31] Ubit Umarov: so both are coerent now
[11:31] Andrew Hellershanks: ok
[11:31] Andrew Hellershanks: Good to know.
[11:31] Ubit Umarov: ahh and delete is also like 10x faster now
[11:31] Ubit Umarov: well X times faster :)
[11:32] Andrew Hellershanks: X is 10 in roman numerals. :)
[11:32] Gavin.Hird :-))
[11:32] Ubit Umarov: yes i played with that :p
[11:32] Gavin.Hird Mac OS X
[11:33] Ubit Umarov: also a incognite ( also on that mac thing :p )
[11:33] Gavin.Hird now that Apple has switched ot macOS 11, I hear Microsoft is working on Windows 11
[11:34] Ubit Umarov: oh not XI ?
[11:34] Gavin.Hird 11
[11:34] Owl.Eyes we should call Windows 10 Windows X as well … they eXaggerate it a lot
[11:34] Andrew Hellershanks: I recall MS saying something about 10 being the last version. They would just keep updating it.
[11:35] Gavin.Hird the one to be presented on Monday will be numbered 12
[11:35] Ubit Umarov: yeha ms said a mess like that
[11:35] Jamie.Jordan they say all kinds of things
[11:36] Andrew Hellershanks: True. :)
[11:37] Kayaker Magic: I have a question: I tried to use save IAR from the console, but find that it only works if I know the password of the avatar I'm saving for, which even the admin does not know. Makes IAR useless for anyone but the grid owner. Any interest in changing this to make it useful?
[11:37] Gavin.Hird will be interesting to see if macOS 12 will kill OpenGL for good
[11:37] Jamie.Jordan i'd pay for Mac os if apple wanted to release on pc
[11:38] Gavin.Hird they are all ARM now
[11:38] Jamie.Jordan I know just sayin
[11:38] Gavin.Hird yeah
[11:38] Jamie.Jordan such a great OS for work stuff
[11:38] Andrew Hellershanks: Kayaker, I have run in to that. I've had to ask the user for their password if/when I needed to save an IAR for them.
[11:39] Ubit Umarov: inv owner password will go on a requirement
[11:39] Gavin.Hird agreed Jamie
[11:40] Kayaker Magic: I noticed NaniSim changed that to work on an admin password instead of the user password.
[11:41] Gavin.Hird not sure that is a good thing
[11:41] Ada.Radius why?
[11:41] Ubit Umarov: possible both should be needed :p
[11:42] Gavin.Hird usually you desing a system so that even an admin cannot (easily) acquire user content
[11:42] Kayaker Magic: Because only people who run private grids can save their own IARs, unless people are willing to give their passwords to the admins to get an IAR.
[11:43] Kayaker Magic: As it is, only grid owners can get their own IAR, nobody else can.
[11:43] Bill Blight: yeah but you can always change their password,  copy/past hash/salt .. does not really stop anybody that really wants to steal
[11:43] Kayaker Magic: Only people with console access can use the save IAR command anyway!
[11:44] Gavin.Hird also for a user to save their own IAR making them able to iport to another grid could make them break all kinds of copyright laws etc
[11:44] Andrew Hellershanks: Right
[11:44] Bill Blight: yeah but that also goes with attached regions
[11:44] Gavin.Hird import*
[11:45] Andrew Hellershanks: That was "Right, Kayaker"
[11:45] Bill Blight: so anybody with an attached region to a grid, could pull inventory if there was lesser password protection.
[11:46] Kayaker Magic: Attached region is that option where a region in one grid appears in another?
[11:46] Ubit Umarov: its like this grid
[11:46] Bill Blight: just like osgrid
[11:46] Ubit Umarov: or metro.. and others
[11:47] Ubit Umarov: you run you own region and connect to a grdi
[11:47] Bill Blight: there are a few , that I won't mention that you could attach to , that don't realize you can ..
[11:47] Ubit Umarov: :)
[11:47] Andrew Hellershanks: Needing to ask the user for their password also helps reduce the chance that a person can pretend to be a given user and asking for an IAR of an inventory they shouldn't have.
[11:47] Ubit Umarov: think some do send you a "cooked" bin
[11:48] Bill Blight: (grids need to know not to publish all their connection info, it can be used for nefarious purposes)
[11:48] Ubit Umarov: only issue of using the user password is that is should be private
[11:48] Jamie.Jordan that's funny
[11:48] Ubit Umarov: but.. hard to avoid
[11:48] Bill Blight: back when I ran my grid, I had a cooked binary that had the groups, and all the connection info hard coded, was not even in the INIs
[11:49] Gavin.Hird the user can always set a trivial password for the export and set a secret one when the export is done
[11:49] Ubit Umarov: correct is ppl change the password... ask for the operation and change password again
[11:49] Ubit Umarov whispers: ( not going into the details lke the one Bill did mention on that ;) )
[11:49] Andrew Hellershanks: Yes, Ubit. If someone wants me to save an IAR and they don't want to give me their password I just tell them to change it temporarily then they can change it back after I have made the IAR.
[11:50] Ubit Umarov: IAR should be a grid side operation
[11:50] Bill Blight: I had a special region that I would spin up with an object that would create their iar for them, and they specified their own password, I never knew it ..
[11:50] Ubit Umarov: well for "normal" grids
[11:51] Ubit Umarov: but yeah a grid can also have a special region for those things
[11:51] Andrew Hellershanks: Bill, that is a good way of doing it.
[11:51] Andrew Hellershanks: I don't particularly want to know anyone elses password. I have too many of my own. I never write down any user password that may be given to me in order to do something. If they asked again for an IAR they would have to give me their password again.
[11:51] Ubit Umarov: well IAR can't also be a inworld thing
[11:52] Ubit Umarov: it is just 2 heavy to allow it
[11:52] Ubit Umarov: ( plus the file delivery )
[11:52] Bill Blight: yep
[11:52] Andrew Hellershanks: Even more so when someone has a very large inventory.
[11:53] Bill Blight: better when it has a limited set of options, that is what I did with my special region , NOTHING was allowed there except, click box, enter password, wait for link ...
[11:53] Ubit Umarov: well can be a inworld controled thing, like Bills one
[11:54] Bill Blight: worked well, but just like anything, give a person access they'll screw it up .. LOL
[11:54] Ubit Umarov: yeah but a grid operation approval is good
[11:54] Andrew Hellershanks: I could see them messing up by saying their password in open chat instead of privately while someone else is nearby. That can be avoided by only allowing the one user in to the special region
[11:55] Gavin.Hird agent limit = 1
[11:56] Andrew Hellershanks: We are near the top of the hour. I'm looking at Gavin. Not trying to put you on the spot. Just want to always be sure to give you the opportunity to give everyone an update (if there is one) about your viewer.
[11:56] Bill Blight: well it was a dialog box, not in chat ... so would have to be extraordinary dumb
[11:56] Gavin.Hird There was an update of the mac version a couple days back
[11:56] Bill Blight: but only allowing the "Get My IAR" group was how I did it ..
[11:56] Gavin.Hird rendering updates from EEP
[11:57] Gavin.Hird plus a bunch of other behind the hoods
[11:57] Andrew Hellershanks: Bill, Good idea.
[11:57] Gavin.Hird hood
[11:57] Owl.Eyes Just a request, Ubit:  There are only functions, which gives the ability to identify, in which grid an avatar is (necessary esp. after a teleport) – it's "osGetGridName", "osGetLoginURI" and "osGetGatekeeperURI", where the result of the second one is not consistent with the name tag of grid-foreign avatars, whereas the third one is (it's only undefined for private grids, whose avatars can't go otherwhere, and can't be entered by foreign avatars) – so "osGetGatekeeperURI" is the only function which really works as a key in the object model, top level and thus needed of course –– and it can be found out manually in the internet as well, in all cases. –– I think, everyone (resp. his/her scripts) have a (human) right to find out, where the avatar is (e.g. region names may be exist in several grids). But all three function are set to "Threatlevel=high" … I can't find any reason for this, because the gatekeeper URI is public and necessary for e.g. the viewer – and btw, if they can only be us
[11:57] Owl.Eyes  an admin or owner at that place, the make NOOO sense. But every region, where the allowed threat level is set to medium or yet lower gives an ugly debug message by calling those functions (and for "osGetGridName" as well).  –– The function(s) are really necessary for hypergrid>grid and object model consistence, for those avatars who jump around between the grids.  The threat level for those three pure informational functions should really be set to "none" or "very low" in the whole OpenSim software.  And in opposite to the current topic, we talked about, it's NO security issue !!!  And it's a pure default setting, no code changes necessary.  I and several of our roleplay members don't want to be bullied any longer, when I come to OSGrid wih my roleplay HUD worn.
[11:57] Gavin.Hird Win version will be a few more weeks now
[11:57] Ubit Umarov: oh adding EEP or just the rendering engine?
[11:57] Gavin.Hird taking rendering updates from the EEP viewer
[11:57] Andrew Hellershanks: Bill, I like the saying -> If you design something to be idiot proof, the universe will design a better idiot.
[11:57] Gavin.Hird so for now Windlight looks better than ever :-)
[11:58] Owl.Eyes *be used by an admin …
[11:59] Ubit Umarov: hmm osGetGatekeeperURI is not where avatar is
[11:59] Gavin.Hird There was no minutes / video from the TPV meeting that was supposed to happen last Friday, so not even sure there was a meeting
[11:59] Owl.Eyes then the name tag is not a key, Ubit – what are teh semantics behind?
[12:00] Andrew Hellershanks: Gavin, ty for the update.
[12:00] Gavin.Hird yw
[12:00] Owl.Eyes *the name tag +addition for foreign avatars
[12:00] Ubit Umarov: hmm and we do have osGetGatekeeperURI ??
[12:00] Owl.Eyes *resp. how are the name tag addition built?
[12:01] Kayaker Magic: Owl: I think some of those functions are THREAT HIGH because they were poorly written and stop the CPU while waiting for the answers from an HTTP request. Ubit: I thought you fixed that for osGetGridName, can it get a lower threat level now?
[12:01] Ubit Umarov: we do
[12:02] Ubit Umarov: that is only local grid information
[12:02] Owl.Eyes ok, then the name tag additions should be adjusted accordingly
[12:02] Ubit Umarov: public string osGetGridGatekeeperURI()
            CheckThreatLevel(ThreatLevel.Moderate, "osGetGridGatekeeperURI");

            return World.SceneGridInfo.GateKeeperURLNoEndSlash;
[12:02] Owl.Eyes is the grid name really a key value? Or cann it appear duplicate?
[12:03] Owl.Eyes *-n
[12:03] Ubit Umarov: what key value?
[12:03] Owl.Eyes in the object modeö
[12:03] Owl.Eyes *model
[12:03] Ubit Umarov: that is a URI ( or url )
[12:03] Bill Blight: Those are some of the info I was speaking of earlier that allows people to attach a region to a grid, without the owners permission ..
[12:04] Ubit Umarov: well yeah
[12:04] Bill Blight: unless you use authentication, publishing the main URL/URIs  for your grid can just be inviting trouble
[12:05] Ubit Umarov: but methods that allow regions to register should be on a private firewall protected thing, if nor using auth
[12:05] Owl.Eyes But the gatekeeper (or login URI) is always know, if people have an avatar of that grid.
[12:05] Bill Blight: should be, does not mean actually is .... LOL
[12:05] Owl.Eyes otherwise the viewer wouldn't work
[12:06] Bill Blight: well they don't have to be the actual, you can have one for internal and one for external, so you don't have to give the internal user url to HG  users ..
[12:06] Ubit Umarov: login is not same as .GateKeeperURL
[12:06] Owl.Eyes I know
[12:07] Ubit Umarov: dsome grids ( like osg) have a login url and a hg url
[12:07] Owl.Eyes but if the extended name tags show the gatekeeper URL (I tested it with more than 10 grids) … how are they computed?
[12:08] Ubit Umarov: you type them on ini files :p
[12:09] Owl.Eyes I have already read in the resp. xml file.  That's where I got all this info, for a consistent object model (for my system) it's inevitable.
[12:09] Owl.Eyes The Roleplays have to work cross-grid
[12:10] Ubit Umarov: what resp xml?
[12:10] Owl.Eyes let me have a look, it's in the config files of the firestorm
[12:11] Owl.Eyes ande it's given for every published (in the internet) grid, at least one of those infos.
[12:11] Ubit Umarov: well handling of hg is still not perfect on viewers
[12:11] Ubit Umarov: ( posisble bc its spec is not that clear either )
[12:12] Owl.Eyes let's not talk around secondary topics … the main thing is to get the information, in which grid my avatar is.  It's a human right, and the functions make otherwise NO sense.
[12:13] Ubit Umarov: so issue is the CheckThreatLevel(ThreatLevel.Moderate, "osGetGridGatekeeperURI"); ?
[12:14] Owl.Eyes It's ok, if only osGetGridName is given free, IF (and only IF) it can not be duplicate
[12:14] Ubit Umarov: even  gridname
[12:14] Owl.Eyes yep
[12:14] Ubit Umarov: public string osGetGridName()
            CheckThreatLevel(ThreatLevel.Moderate, "osGetGridName");
[12:14] Ubit Umarov: don't see it free
[12:15] Owl.Eyes But the threat level should be none or at least very low.  Else trying it on arriving at OSGrid welcome Region gives the ugly debug message
[12:15] Ubit Umarov: well that is since original..
[12:15] Ubit Umarov: regions owners should set that
[12:16] Ubit Umarov: no idea why origical coders considered that a issue
[12:16] Owl.Eyes no, that restrict the human right to get this information
[12:16] Owl.Eyes *+s
[12:16] Ubit Umarov: ossl is not for humans,, its for scripts
[12:17] Owl.Eyes but the script are used under human control and concern humans.  I could be given a TP to a Region, where something bad is done on my AV
[12:18] Owl.Eyes anbd as I said, the function make NO sense otherwise
[12:19] Ubit Umarov: well i was the one to introduce "free" ossl functions
[12:19] Ubit Umarov: doutrine was to always check a threat level
[12:20] Kayaker Magic: There is no way to check threat level from a script.
[12:21] Kayaker Magic: Or to check if the region owner set any OSSL function to free.
[12:21] Ubit Umarov: now there is
[12:21] Ubit Umarov: kinda :p
[12:21] Andrew Hellershanks: Even if one could get the currently set threat level it still wouldn't tell you which functions are allowed to be used in a given grid.
[12:21] Kayaker Magic: The try ... catch statements...
[12:21] Owl.Eyes if you consider the hypergrid as an integrated environment, you SHOULD give the grid key information free.  SL is one grid, Opensim is another grid, and there are other hypergrids.  A hypergrid is defined as a collection of "internal" grids without access to other hypergrids.  That's the object model.  And every grid needs to have a respective key
[12:22] Ubit Umarov: hmm pls file a mantis on that
[12:22] Owl.Eyes ok, then I need to get the permission there, how to?
[12:23] Andrew Hellershanks: Owl.Eyes, Open Sim isn't a grid.
[12:23] Bill Blight: I'd think that is something that should be left to grid owners to decide, hence in the OSSL perms
[12:23] Owl.Eyes in fact, this topic breaks my whole consistent object model for the RP system, and I used already around 2 weeks to workaround this fuzzy
[12:23] Owl.Eyes I said hypergrid
[12:24] Owl.Eyes a hypergrid is a collection of grids connected by a central instance, which allows to TP between the grid.
[12:24] Owl.Eyes *grids
[12:25] Bill Blight: but not all grids are on the hypergrid
[12:25] Owl.Eyes so they have their own hypergrid.  They are in the same one and only metaverse
[12:26] Andrew Hellershanks: Owl.Eyes: Why would an RP system need to care about the grid origin of a user?
[12:27] Owl.Eyes this is the object structure:  World [RealWorld,Metaverse] … Metaverse -> n x Hypergrid … Hypergrid -> n x Grid … Grid -> n x Region ……… Gatekeepers are between the Hypergrids
[12:27] Bill Blight: You can always do a local lookup if the users is local, exists in user db ..
[12:28] Bill Blight: If they don't exist, they are hypergrid
[12:28] Bill Blight: or even parse their username to see if it has an @ symbol in it ..
[12:28] Owl.Eyes Andrew, because AV-GUIDs may be duplicate between grids.
[12:29] Kayaker Magic: Gavin: But I have the problem, I TP into a grid, I want to know where I am, each region in each grid has threat level and OSSL allow set different. I have to start over on every TP, I cannot keep a table of functions allowed, I have to check new each time.
[12:29] Bill Blight: the info after the @ symbol gives you the grid
[12:29] Ubit Umarov: NO they should NOT be duplicated
[12:29] Owl.Eyes I don't want to talk about workarounds, but about a consisten object model
[12:29] Ubit Umarov: a avatar uuid MUST be uinque in the universe
[12:29] Owl.Eyes *t
[12:29] Bill Blight: uuid paired with   should be a pretty unique ID
[12:29] Andrew Hellershanks: Owl.Eyes, it is possible that a GUID could be the same but it is extremely unlikely. I would go to the effort of worry about it.
[12:29] Owl.Eyes oh, how is that done, Ubit?
[12:29] Gavin.Hird well, I suppose that is the inherit nature of hypergrid
[12:30] Andrew Hellershanks: Bill, Right.
[12:30] Ubit Umarov: it must be unique ALONE
[12:30] Gavin.Hird a hypergrid transfer is somewhat like going to a differnet country
[12:30] Kayaker Magic: Avatar UUIDs are not unique: That home grid installer has a fixed one it uses on every copy. Then it lets them HG TP around the metaverse.
[12:30] Ubit Umarov: it is a UUID
[12:30] Gavin.Hird you have to check the laws of each country
[12:30] Ubit Umarov: uiids must be uinique
[12:30] Bill Blight: two uers would have to be made on the same system at the exact same millisecond to have a duplicate ID, since UUID is based on time/date/hardware
[12:30] Gavin.Hird and in reality, a HG transfer might indeed place you ina differnet jurisdiction
[12:31] Owl.Eyes ok, then I can find out, where an AV comes from.  But it doesn't solve the problem to get the information about where an AV just arrived and it breaks the consistent object model
[12:31] Owl.Eyes yep, Gavin, that's a reason as well, thanks
[12:31] Ubit Umarov: as i said, file a mantis asking for a review of those functions threat levels
[12:31] Owl.Eyes one reason more to get this info
[12:31] Bill Blight: tracking previous grid movements , could very well violate local privacy laws
[12:32] Ubit Umarov: wel yeah
[12:32] Ubit Umarov: it can be a can of worms
[12:32] Andrew Hellershanks: Yes, thank the Europeans for making us have to deal with the GDPR.
[12:32] Owl.Eyes but that's already possible from the AV name  addition
[12:32] Owl.Eyes *extension
[12:33] Gavin.Hird GDRP is only for 37 of the 50 European countries
[12:34] Andrew Hellershanks: Gavin, perhaps but it has still caused an impact outside of those countries.
[12:34] Bill Blight: getting  home grid is not the same as tracking where they just landed from, which you should never actually need, but if you got it, you could be intruding on personal information
[12:34] Gavin.Hird well, yeah
[12:35] Owl.Eyes you see, I'm software architect since half a century, and I'm Asperger autist.  This object model consistence is ESSENTIALLY for me and all of my work
[12:35] Bill Blight: (and there are already ways to get it, if you are savvy enough)
[12:35] Gavin.Hird info about to what country your HG transfer is to should probably be part of any HG request/TP exchange
[12:36] Owl.Eyes Bill, THAT information is NOT necessary (where s.o. came from) – and his "home grid" is already well-known
[12:37] Andrew Hellershanks: Gavin, I think it could be determined from the users IP address.
[12:37] Gavin.Hird not a good way
[12:37] Gavin.Hird it can be approximated
[12:37] Owl.Eyes pleasse don't let us talk any longer about populistic matters.  This is purely a matter of a consistent object model
[12:38] Gavin.Hird we have a pan Nordic ISP, which use one class A address for multiple countries
[12:38] Owl.Eyes how do I connect to opensim's mantis?
[12:39] Andrew Hellershanks:
[12:40] Owl.Eyes every hypergrid should have it's GUID as well, retrievable everywhere and by anyone.  Then the object model would be fine.
[12:41] Bill Blight: would love an GUID that I could spoof to make people think I was from another grid .... LOL
[12:41] Gavin.Hird :-)
[12:41] Owl.Eyes Thx Andrew
[12:41] Andrew Hellershanks: Owl.Eyes, yw
[12:41] Owl.Eyes hope to be able to report
[12:42] Andrew Hellershanks: Any other last minute items for today? I need to get going soon.
[12:42] Bill Blight: name/ip address is pretty reasonable , because in the internet of things two things cannot be in the same place so it is pretty unique , unless you spoof it ..
[12:42] Bill Blight: name=hg.address
[12:43] Owl.Eyes yep, this adress would be fine as well, if it doesn't change
[12:43] Bill Blight: gotta bail, outreach is back up ..
[12:43] Owl.Eyes I workarounded it by the gatekeeperUR for foreign AVs
[12:46] Owl.Eyes Case:  we want to play our RP in OSGrid AND in Kitely.  But if my OSGrid regions are not up (or it's inworld RP server is down), the Players land on the welcome SIM of OSGrid and make debug noise, for which they are bullied.  And therir script has to be manually resetted in case of my RP server down.
[12:47] Andrew Hellershanks: If your RP server was down they wouldn't be able to RP.
[12:47] Owl.Eyes and my database work for the system runs into brainfuck over and avoer, becaus the grid has no key
[12:47] Owl.Eyes Andrew, that abvious, but it doesn't change anything on havong to manually reset
[12:48] Owl.Eyes *obvious *having
[12:49] Owl.Eyes as I said, I don't want to argue, but have a severe bug be corrected.   KEY for OBJECT !!!!!
[12:49] Owl.Eyes (where "OBJECT" is "Grid" here
[12:50] Owl.Eyes to be precise:  Object-TYPE is "Grid"
[12:51] Owl.Eyes … and it's a relative top-level object, so it's really necessary … it's a FAULT in the hitherto object model
[12:52] Jamie.Jordan i'm out be well yall
[12:52] Andrew Hellershanks: ok, Jamie. Thanks for dropping by.
[12:53] Owl.Eyes The name can be replaced by the key itself (for display), if it's not known, but a unique key is INEVITABLE
[12:53] Andrew Hellershanks: Owl.Eyes, I don't think there is a severe design problem. I don't know how you wrote you scripts and what they do to know if you have found a problem or if it is a problem with your scripts.
[12:54] Owl.Eyes NO, it IS a severe design problem for every expert, esp. autistic ones and software architects
[12:54] Andrew Hellershanks: For a vendor system I have I have the scripts provide a private code that is unique to each grid when it interacts with my off-world database.
[12:54] Owl.Eyes I don't talk about applications, but the underlying object model
[12:55] Andrew Hellershanks: Owl.Eyes, If it is a real problem as you are saying it is then file a mantis. I don't follow this enough to understand the problem.
[12:55] Owl.Eyes the problem is a basic one.  Not one of applications
[12:55] Andrew Hellershanks: It is almost the top of another hour. I need to get going. I will close the meeting for now.
[12:55] Owl.Eyes It concerns the database structur for example
[12:56] Andrew Hellershanks: Thank you all for coming. See you again next week.
Personal tools
About This Wiki