Chat log from the meeting on 2023-12-12

From OpenSimulator

Jump to: navigation, search
[12:00 PM PT]  Andrew Hellershanks: Hello, everyone.
[12:00 PM PT]  Mike.Chase Way easier than managing the other maps.  Unless your making your own materials
[12:00 PM PT]  Ubit Umarov: guess it is a nice thign to trade objects..
[12:00 PM PT]  Joe Magarac: A big advantage is that in time you should be able to upload a whole linkset.
[12:00 PM PT]  Andrew Hellershanks: What is done by the same people that made collada?
[12:00 PM PT]  Joe Magarac: glTF.
[12:01 PM PT]  Andrew Hellershanks: ah. ok
[12:01 PM PT]  Joe Magarac: Really, it's now the de facto standard for shipping content around during creation.
[12:01 PM PT]  Ubit Umarov: well as i said only a few pbr assets make sense to be in marked
[12:01 PM PT]  Joe Magarac: You want to use .obj files?
[12:01 PM PT]  Ubit Umarov: since all the good ones are specifec to the mesh/prim
[12:01 PM PT]  Joe Magarac: Right.
[12:01 PM PT]  Web Rain: Hi :)
[12:02 PM PT]  Andrew Hellershanks: Hello, Web.
[12:02 PM PT]  Ubit Umarov: and ll had to make a new formal in llsd to actually broadcast it
[12:02 PM PT]  Andrew Hellershanks: I hope everyone enjoyed the OSCC this past weekend.
[12:02 PM PT]  Lyr.Lobo yes *grins*
[12:02 PM PT]  Mike.Chase I guess we'll just have to see what content creators make of it.  Pesky users :)
[12:02 PM PT]  Andrew Hellershanks grins at Mike.
[12:02 PM PT]  Joe Magarac: Well, I gave my OSSC talk, but then couldn't get back in, so I gave up.
[12:02 PM PT]  Web Rain: congratulations on the success of the conference :) very good job Lyr :)
[12:02 PM PT]  Ubit Umarov: they jsut want to have things to put on market
[12:02 PM PT]  Vincent.Sylvester That's the great thing about open standards, everyone messes with them and adds their own glitter to it. Same with collada. The amount of pain that creates when working with others. Things not compatible of importers/exporters dropping parts you then have to re-create. How long did it take for LL to properly obey collada standard if they even do that now
[12:03 PM PT]  Andrew Hellershanks: Lyr, I have one minor bit of feedback to the organizers.
[12:03 PM PT]  Ubit Umarov: this assets make little sense really
[12:03 PM PT]  Mike.Chase Yes it was very nice as always, Great job all around
[12:03 PM PT]  Ubit Umarov: unless if just setting a face to look like simple bronze etc
[12:03 PM PT]  Andrew Hellershanks: I haven't finished watching all of the panels I am interested in. I have a few more to go.
[12:03 PM PT]  Joe Magarac: glTF has reference scenes that should work everywhere. find the one called "bistro" and import it into Blender.
[12:03 PM PT]  Lyr.Lobo Sure *smiles*
[12:04 PM PT]  Ubit Umarov: yea like collada "works" everywhere
[12:04 PM PT]  Lyr.Lobo I'm testing two viewers at the moment and here for feedback
[12:04 PM PT]  Vincent.Sylvester All the breaking changes because of that and fundamental issues still being ignored, that's what bugs me the most
[12:04 PM PT]  Andrew Hellershanks: I can't use Collada in my main 3D modelling program.
[12:04 PM PT]  Joe Magarac: There really has been progress since Collada.
[12:04 PM PT]  Ubit Umarov: those without fs 7 will see the helmet grey :)
[12:05 PM PT]  Andrew Hellershanks: Hello, Cuga
[12:05 PM PT]  Lyr.Lobo Hi Cuga!
[12:05 PM PT]  Cuga.Rajal Hi!
[12:05 PM PT]  Joe Magarac: I'd tale glTF over Collada, fbx, obj, 3ds, etc.
[12:05 PM PT]  Joe Magarac: take
[12:05 PM PT]  Mike.Chase amen, me too
[12:05 PM PT]  Vincent.Sylvester Could be implementing a lot of cool new features and actually solve some long standing issues breaking other protocols, instead the bum around with fancy materials and shove all the stuff into http and web-loaded floaters. Polishing a turd doesn't turn it into gold
[12:06 PM PT]  Andrew Hellershanks: hehe
[12:06 PM PT]  Ubit Umarov: and as is said.. this helmet only uses basic asset .. that is working  here since firstt fs alpha
[12:06 PM PT]  Andrew Hellershanks: Ah, Now that I look at Ubit I see he is a headless gray avvatar.
[12:07 PM PT]  Ubit Umarov: the material overrides, ( changes per face) had a protocol change, as i said..  that is now working here also ( and on fs 7-0-1 )
[12:07 PM PT]  Joe Magarac: Look, at least there's a spec for glTF. After years of reverse-engineering LL code, that's a win.
[12:07 PM PT]  Andrew Hellershanks: Um... Make that an invisible avatar with a gray helmet.
[12:07 PM PT]  Ubit Umarov: im headless?
[12:07 PM PT]  Ubit Umarov: well happens :)
[12:07 PM PT]  Mike.Chase lol
[12:08 PM PT]  Mike.Chase been there
[12:08 PM PT]  Ubit Umarov: so.. about last week code changes.. hmm think none :)
[12:09 PM PT]  Jagga Meredith: Is there a viewer other than FS that runs on Mac?
[12:09 PM PT]  Andrew Hellershanks: Before we get other discussions I want to talk about one of the upcoming meetings.
[12:09 PM PT]  Cuga.Rajal Dayturn
[12:09 PM PT]  Ubit Umarov: dayturn should run on mac
[12:09 PM PT]  Jagga Meredith: ty
[12:09 PM PT]  Ubit Umarov: hope gavin is well ..  he is missing this meetings for someweeks now :(
[12:09 PM PT]  Cuga.Rajal Singularity also, but it has issues with latest server
[12:10 PM PT]  Ubit Umarov: and missed the panel abotu viewers at oscc :(
[12:10 PM PT]  Joe Magarac: A test indicated that we can make Sharpview run on Mac, but you have to jump through all those Apple hoops to get "approval", and it's not worth the trouble.
[12:10 PM PT]  Ubit Umarov: think singularity is dead
[12:10 PM PT]  Ubit Umarov: replaced by alchemy
[12:10 PM PT]  Cuga.Rajal I haven't had problems with those hoops
[12:10 PM PT]  Joe Magarac: If Apple loses their play store antitrust case, though, we might get app freedom back.
[12:10 PM PT]  Lyr.Lobo if anyone cannot see the helmet's PBR reflections, IM me and I'll send an image
[12:11 PM PT]  Andrew Hellershanks: The helmet is all gray to me. I'm on Firestorm 6.6.14
[12:11 PM PT]  Cuga.Rajal I need the new FS beta to see that, correst?
[12:11 PM PT]  Vincent.Sylvester Heh Joe you should just pawn off the mac work to Cuga, I'm sure he'll happily help *runs and hides*
[12:12 PM PT]  Ubit Umarov: hmm seems material overrides did not pass well into inventory.. oops :)
[12:12 PM PT]  Cuga.Rajal I don't know Rust lol
[12:13 PM PT]  Andrew Hellershanks: If we can talk about an upcoming meeting for a moment...
[12:13 PM PT]  Andrew Hellershanks: In two weeks time the meeting lands on the 26th. For those of us in North America that is boxing day. My feeling is that we should take a break that day.
[12:13 PM PT]  Orbert.Tatham No arguments here
[12:13 PM PT]  Lyr.Lobo I'll share images of it
[12:13 PM PT]  Lyr.Lobo ignore them if you do not want to see
[12:13 PM PT]  Joe Magarac: OK.
[12:13 PM PT]  Cuga.Rajal ok with me to skip
[12:14 PM PT]  Mike.Chase makes sense yes
[12:14 PM PT]  Andrew Hellershanks: Lyr, you could rez a prim and apply the image.
[12:14 PM PT]  Andrew Hellershanks: The helmet looks a lot better in the images than how I see it.
[12:15 PM PT]  Lyr.Lobo yes
[12:15 PM PT]  Lyr.Lobo i could... hmm...
[12:15 PM PT]  Lyr.Lobo no, i cannot
[12:15 PM PT]  Lyr Lobo: sorry, i can't either, Andrew
[12:15 PM PT]  Andrew Hellershanks: Depending on my day I might be here on the 26th in case anyone shows up but for the record we won't have an official meeting on the 26th. The next one after that date will be Jan 2 in the new year.
[12:16 PM PT]  Web Rain: gray for me but i have a bad internet :( maybe still loading
[12:17 PM PT]  Web Rain: FS 6.6.14
[12:17 PM PT]  Andrew Hellershanks: I used this viewer to attend the OSCC. It has messed up my cache. Part of the bar ismissing.
[12:17 PM PT]  Vincent.Sylvester Hm yeah that's been happening a lot with this version on all platforms apparently
[12:18 PM PT]  Cuga.Rajal I'm usoing the latest release 6.6.16, but I donlt think it has PBR support... I think you need the even newer beta
[12:18 PM PT]  Vincent.Sylvester You need to manually delete the caches. Also, for some reason, helps to start a second viewer instance and login, really odd bug
[12:18 PM PT]  Andrew Hellershanks: With 6.6.14 I don't seem to have voice any more. I was unable to attend the VIP gathering on Saturday because I could neither hear nor talk during that event.
[12:18 PM PT]  Cuga.Rajal Voice worked for me at the VIP gathering with 6.6.16
[12:18 PM PT]  Lyr Lobo: So very sorry, Andrew
[12:19 PM PT]  Kayaker Magic: I heard voice on my 6.6.14 version.
[12:19 PM PT]  Andrew Hellershanks: I should have checked for an update before the event.
[12:19 PM PT]  Lyr.Lobo I did too, but it is unfortunate that yours was not working
[12:19 PM PT]  Ubit Umarov: can see the picture?
[12:19 PM PT]  Andrew Hellershanks: Here, the voice button is still grayed out.
[12:19 PM PT]  Mike.Chase Well they've pulled .16 down to fix some bugs
[12:19 PM PT]  Lyr.Lobo Fs had an issue and pulled their viewer until a fix.
[12:19 PM PT]  Ubit Umarov: yeah broke AO
[12:20 PM PT]  Andrew Hellershanks: Perhaps it is the Linux client that has a problem with voice.
[12:20 PM PT]  Ubit Umarov: possible same issue on 7.x
[12:20 PM PT]  Lyr.Lobo They have 6.6.16 back online now
[12:20 PM PT]  Lyr.Lobo then again, maybe not
[12:20 PM PT]  Andrew Hellershanks: Personally it doesn't matter to me if AO is broken as I never use it but I know others do use it and would want it fixed.
[12:20 PM PT]  Lyr.Lobo right, still down
[12:20 PM PT]  Lyr.Lobo I used 6.6.16 during the conference
[12:21 PM PT]  Orbert.Tatham I would hope that they would release the fix with a point increment
[12:21 PM PT]  Cuga.Rajal PBR is not supported on any of the FS releases yet, only betas
[12:21 PM PT]  Lyr.Lobo as I have a few versions and needed stability *grins*
[12:21 PM PT]  Ubit Umarov: not even betas only alpha :)
[12:21 PM PT]  Cuga.Rajal ahh
[12:21 PM PT]  Andrew Hellershanks: Orbert, yes. Not good fix a version and keep the same number. People wouldn't know if they have the fixed or broken version.
[12:21 PM PT]  Ubit Umarov: guess pbr will be on betas in Jan
[12:22 PM PT]  Ubit Umarov: well no idea abotu other viewers :(
[12:24 PM PT]  Cuga.Rajal I was hoping to make mirror balls to hang from the Xmas tree but that will have to wait until next year :)
[12:25 PM PT]  Andrew Hellershanks: :)
[12:25 PM PT]  Ubit Umarov: well the prim on floor is supposed to have some reflection
[12:25 PM PT]  Andrew Hellershanks: I am interested to find out if PBR would allow for mirrors.
[12:25 PM PT]  Ubit Umarov: i do see .. well things.. on it :)
[12:25 PM PT]  Vincent.Sylvester "Oh what fun it is to lag while looking at a tree"~
[12:25 PM PT]  Joe Magarac: LL is putting in mirrors.
[12:26 PM PT]  Joe Magarac: Only one mirror will work at a time.
[12:26 PM PT]  Joe Magarac: Because overhead, of course,
[12:26 PM PT]  Ubit Umarov: think thr blue from sky..
[12:26 PM PT]  Joe Magarac: That's low-rez ambient reflection. You can have more of that.
[12:26 PM PT]  Lyr Lobo chuckles
[12:26 PM PT]  Andrew Hellershanks: hehe... so no putting one mirror directly across from another to see infinite reflections? :)
[12:26 PM PT]  Ubit Umarov: guess to show the roof it needs a probe here
[12:26 PM PT]  Ubit Umarov: they are fake mirrors
[12:26 PM PT]  Ubit Umarov: worse than the ones we see in games now
[12:27 PM PT]  Vincent.Sylvester Now? We had that 10  years ago
[12:27 PM PT]  Ubit Umarov: well ll is working on this for sometime.. still  well whatever..
[12:27 PM PT]  Ubit Umarov: and this form of pbr is now outdated i guess
[12:28 PM PT]  Vincent.Sylvester Someone made a Portal port on the N64, which has to render a second scene like a mirror for each portal so that's how long that's been possible
[12:28 PM PT]  Vincent.Sylvester Was possible back when LL was founded even, only took em 20 years to catch up huh
[12:29 PM PT]  Vincent.Sylvester So viewers with raytracing in 2055 I guess
[12:29 PM PT]  Andrew Hellershanks: :)
[12:29 PM PT]  Joe Magarac: Mirrors are easy to do, expensive to render. You have to render the scene an extra time for each mirror.
[12:29 PM PT]  Cuga.Rajal Mirror graphics not client side?
[12:29 PM PT]  Joe Magarac: Mirrors are definitely client side.
[12:29 PM PT]  Ubit Umarov: this is all client side
[12:30 PM PT]  Ubit Umarov: opensim side is just moving data around, as usual
[12:30 PM PT]  Orbert.Tatham I know of a couple of places in SL that had a mirror-image scene underneath a transparent floor to give the illusion of a reflective floor, but we were all vampires because we didn't see ourselves in the mirror :)
[12:30 PM PT]  Ubit Umarov: 99.999% viewer side things
[12:30 PM PT]  Cuga.Rajal ah ok
[12:30 PM PT]  Andrew Hellershanks: Orbert, I have seen a place or two like that.
[12:30 PM PT]  Ubit Umarov: water did reflections ..
[12:31 PM PT]  Ubit Umarov: back in 2007 :)
[12:31 PM PT]  Ubit Umarov: no idea now
[12:31 PM PT]  Kayaker Magic: There is a mirror trick using clever setup of the ocean. So you can dance on one mirror, if it is the water level.
[12:31 PM PT]  Web Rain: i remember i made one an era ago with 2 avatars and screen capture   moap
[12:32 PM PT]  Web Rain: fake mirror ^^
[12:32 PM PT]  Andrew Hellershanks: nice
[12:32 PM PT]  Ubit Umarov: well think this pbr implementation is a few years outdated relative to industry..
[12:33 PM PT]  Ubit Umarov: but well it does look nice :)
[12:33 PM PT]  Tess JL: It is the industry standard, Ubit. That's why LL is switching to it
[12:33 PM PT]  Ubit Umarov: no it is not
[12:33 PM PT]  Tess JL: It isn't actually more advanced than the existing materials, just easier to work with for people used to other platforms
[12:33 PM PT]  Ubit Umarov: there are several diferent implementations of pbr
[12:34 PM PT]  Ubit Umarov: even the gltf one now has more features that what ll added
[12:34 PM PT]  Tess JL: That's right but with the Kronos Group's support glTF PBR is effectively the standard
[12:34 PM PT]  Joe Magarac: Yes.
[12:34 PM PT]  Joe Magarac: Fortunately.
[12:35 PM PT]  Joe Magarac: There's an actual spec, for which there are multiple implementations.
[12:35 PM PT]  Mike.Chase yes
[12:35 PM PT]  Ubit Umarov: well for spec, it is not blender idea.. etc etc :)
[12:36 PM PT]  Ubit Umarov: a week free spec it seems :)
[12:36 PM PT]  Vincent.Sylvester LL implementing anything to spec without making any custom stuff that trips you up would be a new thing
[12:36 PM PT]  Joe Magarac: True.
[12:36 PM PT]  Tess JL: Spot on ;-)
[12:37 PM PT]  Joe Magarac: LL is, I think, sending binary LLSD for materials now.
[12:37 PM PT]  Ubit Umarov: no
[12:37 PM PT]  Joe Magarac: No?
[12:37 PM PT]  Joe Magarac: I need to look.
[12:37 PM PT]  Ubit Umarov: the overrides are sent in LLSD notation
[12:38 PM PT]  Ubit Umarov: llsd binary is really bad
[12:38 PM PT]  Ubit Umarov: very naive thing
[12:38 PM PT]  Ubit Umarov: they only use it on a few things like mesh
[12:39 PM PT]  Ubit Umarov: well in fact the mesh asset is nonsense JSON embeded into llSD binary
[12:40 PM PT]  Ubit Umarov: i added the default empty pbr asset to our lib
[12:41 PM PT]  Ubit Umarov: ? LLSD/Binary ?>
{   �k   �datas   -{"asset":{"version":"2.0"},"materials":[{}]}
k   �types   �GLTF 2.0k   �versions   �1.1}
[12:41 PM PT]  Ubit Umarov: that is it :p
[12:42 PM PT]  Ubit Umarov: the non ascii things are sizes of strings and array as per llsf binary..
[12:43 PM PT]  Ubit Umarov: {"asset":{"version":"2.0"},"materials":[{}]}  is gltf json directly embeded as string
[12:43 PM PT]  Ubit Umarov: as i told .. empty one :)
[12:44 PM PT]  Andrew Hellershanks nods
[12:45 PM PT]  Mike.Chase When its appropriate I have something I wanted to toss out related to our work on NGC. I dont want to derail a conversation though
[12:45 PM PT]  Andrew Hellershanks: If we are finished with PBR discussions I am going to ask Mike to take the floor.
[12:45 PM PT]  Ubit Umarov: guess now is a good time :)
[12:45 PM PT]  Mike.Chase I want to mention that we have a working implementation of linkset data in our latest release of NGC.
It may affect people running grids.  This is mostly only an issue if you HG from an NGC grid to core 
but most people choosing to run NGC know that.

Zontrek had submitted a patch to Mantis and I contacted her to see if she'd be willing to let us use it in NGC.
We collaborated on getting it into the code on a MR that we worked together. It came out really nice.
My users are really excited about it.
[12:46 PM PT]  Ubit Umarov: (bad users) ;)
[12:46 PM PT]  Andrew Hellershanks: Mike, For those reading the meetinglog who don't know what is NGC, please say what that is.
[12:46 PM PT]  Mike.Chase If there is any interest in the code its our master branch now
[12:47 PM PT]  Jagga Meredith: New Good Code?
[12:47 PM PT]  Lyr.Lobo Next Gen Compounding?
[12:47 PM PT]  Mike.Chase NGC is a derivative of OpenSim Core. We're focused on adding content protection and additional security. Plus tracking features from SL a but more closely
[12:47 PM PT]  Web Rain: XD
[12:47 PM PT]  Mike.Chase Next Gen Core
[12:48 PM PT]  Mike.Chase Or New Good Code. That works too
[12:48 PM PT]  Mike.Chase :)
[12:48 PM PT]  Andrew Hellershanks: :)
[12:48 PM PT]  Jagga Meredith: kewl
[12:48 PM PT]  Orbert.Tatham It was my understanding that what you are talking about is things that this group has discussed and decided we were better off without
[12:48 PM PT]  Andrew Hellershanks: Mike, Were do interested parties go to find the code for NGC?
[12:49 PM PT]  Mike.Chase
[12:49 PM PT]  Mike.Chase
[12:49 PM PT]  Andrew Hellershanks: ty again.
[12:49 PM PT]  Mike.Chase is the github
[12:49 PM PT]  Ubit Umarov: ty
[12:49 PM PT]  Web Rain: ty :)
[12:49 PM PT]  Ubit Umarov: i did look to it.. still don't like the sortedlist :)
[12:49 PM PT]  Mike.Chase I'm working on addressing the System.Image issues in dotnet core as well.
[12:50 PM PT]  Web Rain: you can push pull requests in opensim github... will make the things easy to the core to verify and merge
[12:50 PM PT]  Ubit Umarov: well think it is a bad ll thing really
[12:50 PM PT]  Andrew Hellershanks: Orbert, there wasn't a lot of interest in the feature. It wasn't high on the list of priorities but if someone else has written the code it is worth a look.
[12:50 PM PT]  Vincent.Sylvester If it wasn't flawed concept from the getgo
[12:50 PM PT]  Mike.Chase Well just as a comparison its about 13 times faster than list access.  So if you need a KVP list its very handy. Keeps people from writing tons of notecards too
[12:50 PM PT]  Lyr.Lobo smiles
[12:51 PM PT]  Vincent.Sylvester Reading how it is meant to work is just horrid. They did change some things for the better, but it is still flawed
[12:51 PM PT]  Mike.Chase So far our users are excited.   I have people bringing some projects over from SL because we have it
[12:52 PM PT]  Ubit Umarov: that 13x is a questionable thing.. and depends if you mean read or insert ( and what list)
[12:52 PM PT]  Mike.Chase Anyway I just wanted to raise awareness. It might not be an issue for some grids but it doesnt hurt to be aware of it
[12:52 PM PT]  Andrew Hellershanks: Just for discussion, if we didn't follow SL closely on how they implemented it is there room for improvement on the idea and its implementation?
[12:52 PM PT]  Ubit Umarov: but ill look again.. not "closing doors"
[12:52 PM PT]  Mike.Chase Inserts are about 13x faster
[12:52 PM PT]  Vincent.Sylvester Already have a perfectly usable data store on prims, their descriptions and names, which especially in linksets give you a lot of data to work with. We fixed notecards creating tons of assets a while back, so that should be reduced a lot. Writing scripts that don't spam notecards or leak memory and cpu is a skill. Patching that up with more things added on top is a bandaid not a fix
[12:52 PM PT]  Ubit Umarov: may add your implementation and eventually break it in ossl :)
[12:53 PM PT]  Mike.Chase As you wish. The code is there :)
[12:53 PM PT]  Mike.Chase Vincent again for some of my users it was an important feature
[12:54 PM PT]  Ubit Umarov: Mike all that Array.Copy needed to insert plus tjeh sort..   hadly any faster that simple list
[12:54 PM PT]  Lyr.Lobo how does it impact performance, Mike?
[12:54 PM PT]  Andrew Hellershanks: Mike, Does any of the changes translate to making standard lists faster?
[12:54 PM PT]  Ubit Umarov: and infact there was no need for sortedlist
[12:55 PM PT]  Mike.Chase I had some users run tests. 10000 kvp inserts in about 5 seconds.  Its not doing I/O so the overhead is small
[12:55 PM PT]  Ubit Umarov: if ll had not added those nonsense inbdex based functions
[12:55 PM PT]  Mike.Chase And no, Andrew nothing that mpacts list performance in this
[12:55 PM PT]  Vincent.Sylvester Users scream for a lot of things, that doesn't make an implementation great conceptually. There were multiple attempts to add script data or prim data in the past and at the end of the day doing it on prim level and randomly combining data on link, regex, so many things that are flawed about that. There are even patches on mantis for other implementations of such things that are even worse
[12:55 PM PT]  Ubit Umarov: ( we wehre talking abotu c# list, not lsl list.. well i was.. )
[12:56 PM PT]  Vincent.Sylvester Never mind the overhead it adds, incompatibility across versions etc. so it's a much bigger consideration
[12:56 PM PT]  Mike.Chase Its basically a persisted dictionary on the SOG. Which addresses merges when you link things and is persisted with the prims
[12:56 PM PT]  Andrew Hellershanks: Mike, I was just wondering if the methods used to handle the KVP type lists would translate to changing how regular lists work. Just wondering if you can leverage the technique.
[12:57 PM PT]  Web Rain: i don't use lists in my scripts... i use json and the json store
[12:57 PM PT]  Ubit Umarov: as mike tells is is basicly a KVP( with optional password) thing stored on the root prim of a linkset and seen by all scripts on the linkset
[12:58 PM PT]  Mike.Chase Well an LSL list is a different animal.  I dont think there is any low hanging fruit here. Its possible but I cant think of anything obvious
[12:58 PM PT]  Ubit Umarov: on sl the password is a hask thing, guess SHA256
[12:58 PM PT]  Andrew Hellershanks: At any rate, ty for the information, Mike. I have pulled a copy of the source and will have a look at it later.
[12:58 PM PT]  Mike.Chase This perisists to JSON in the database.
[12:58 PM PT]  Ubit Umarov: on -ngc is plain text
[12:59 PM PT]  Mike.Chase well its plain text but they reserver a certain amount of space if you provide one to accomodate a SHA hash
[12:59 PM PT]  Ubit Umarov: yeah 32 bytes.. and that is their pass
[12:59 PM PT]  Mike.Chase Thanks Andrew.  If I can answer any questions let me know.
[1:00 PM PT]  Ubit Umarov: anohter nice feature.. do sha256 per each access :)
[1:00 PM PT]  Ubit Umarov: ( at sl )
[1:00 PM PT]  Orbert.Tatham lol
[1:00 PM PT]  Mike.Chase I was actually really proud of how the implementation done. Collaborating with Zontrek was alot of fun
[1:00 PM PT]  Web Rain: about hashes! there is news about llHMAC ?
[1:01 PM PT]  Ubit Umarov: well seems ok...   json embeded in xlm etc
[1:01 PM PT]  Web Rain: i readed in a previous chat that Ubit was woRking in it or think to implement...
[1:01 PM PT]  Ubit Umarov: :p
[1:02 PM PT]  Ubit Umarov: yeah  i did ( do) consider just adding siimple dictionary based KVP stored at linkset not prim
[1:02 PM PT]  Andrew Hellershanks: During the OSCC weekend I also thought about the OpenSimEvents module I started a while back. Some ideas popped in to my head regarding implementing handling reminders. I'm now thinking of writing down my ideas, comparing it to a conversation I had with Melanie about an event reminder module, and getting back to doing some coding on the feature.
[1:02 PM PT]  Ubit Umarov: for us linkset is a entity
[1:03 PM PT]  Ubit Umarov: unlike ll that keeps addressing it via root prim and storing linkset only thing on the root prim
[1:03 PM PT]  Ubit Umarov: we do store KFM on the linkset already.. etc
[1:03 PM PT]  Ubit Umarov: ( hmm i think )
[1:04 PM PT]  Vincent.Sylvester Problem with events is that it is not clear what the future of search in viewers will be. It's supposed to be legacy and replaced all by web search, but that hasn't happened yet
[1:04 PM PT]  Joe Magarac: Gotta go, next meeting. Thanks.
[1:04 PM PT]  Web Rain: @  
[1:04 PM PT]  Andrew Hellershanks: ok, Joe. Thanks for dropping by. See you next week.
[1:04 PM PT]  Lyr.Lobo take care
[1:05 PM PT]  Vincent.Sylvester If it ends up all being web then might as well just write it there, cron easier than doing C# timers
[1:05 PM PT]  Mike.Chase I need to run as well.  Happy Holidays all.  I hope your holiday is full of blessings
[1:05 PM PT]  Lyr.Lobo That would be great, Andrew
[1:05 PM PT]  Vincent.Sylvester Thing is that doesn't really do much for HG in any case unless you were to revive that thingy
[1:05 PM PT]  Ubit Umarov: thanks Mike.Chase
[1:05 PM PT]  Andrew Hellershanks: ok, Mike. See you again next week.
[1:05 PM PT]  Web Rain: ty for the infos mike ^^
[1:05 PM PT]  Lyr.Lobo Right, @Vincent
[1:05 PM PT]  Vincent.Sylvester And with hg you then have the issue of hop links or what have you
[1:06 PM PT]  Lyr.Lobo Have a great week
[1:07 PM PT]  Andrew Hellershanks: Vincenet, it isn't web based. A web component is needed for the creation of events. The module I have now allows people to add/remove events of interest but there is currently no means to be reminded of an event.
[1:07 PM PT]  Lyr Lobo: Have fun
[1:07 PM PT]  Andrew Hellershanks: A timer would be used but it wouldn't have to be triggered that often.
[1:07 PM PT]  Ubit Umarov: oops RL also calls
[1:07 PM PT]  Ubit Umarov: cya later
[1:07 PM PT]  Vincent.Sylvester There is protocol for the viewer reminders I think, somewhere
[1:07 PM PT]  Andrew Hellershanks: ok, Ubit.
[1:08 PM PT]  Andrew Hellershanks: I'll wrap up this meeting shortly.
[1:08 PM PT]  Vincent.Sylvester Thing is you can talk to that however you want if you know how, doesn't have to sit inside OpenSim itself
[1:08 PM PT]  Web Rain: @   :)
[1:08 PM PT]  Vincent.Sylvester Much like IM
[1:08 PM PT]  Andrew Hellershanks: Vincent, yes but nothing that will send an IM or pop up a dialog box to say "The event X you were interested in starts in 5 minutes", or some such thing.
[1:09 PM PT]  Lyr.Lobo Have a great week and thank you for all that  you do
[1:09 PM PT]  Web Rain: you too Lyr   
[1:09 PM PT]  Vincent.Sylvester I have a system like that, sending IM directly to a user, that's not that hard
[1:09 PM PT]  Andrew Hellershanks: I need to review the code I have already written to remind myself what I had done.
[1:10 PM PT]  Vincent.Sylvester It's all xmlrpc, you just need to know what to send and where
[1:10 PM PT]  Andrew Hellershanks: no, IMing a user from a module is easy. It is the rest of it that takes more work.
[1:11 PM PT]  Andrew Hellershanks: You have to check for incoming and departing root agents in a regions, see who has subscribed to events, add/remove them from the list of people to notify about an upcoming event, determine when the reminders need to go out, and send them as needed.
[1:11 PM PT]  Andrew Hellershanks: Simple. :)
[1:12 PM PT]  Andrew Hellershanks: On that note, and with some people already having had to leave, I'll draw this meeting to a close.
[1:12 PM PT]  Andrew Hellershanks: Thank you all for coming. See you again next week. After next week, we will skip the 26th and meet again in January.
Personal tools
About This Wiki