Chat log from the meeting on 2012-09-11

From OpenSimulator

(Difference between revisions)
Jump to: navigation, search
(Created page with "<pre style="white-space: pre-wrap; white-space: -moz-pre-wrap; white-space: -pre-wrap; white-space: -o-pre-wrap; word-wrap: break-word"> [11:03] Franziska Bossi: eep!! [11:03] ...")
 
m
 
Line 1: Line 1:
<pre style="white-space: pre-wrap;
+
<pre>
white-space: -moz-pre-wrap;
+
white-space: -pre-wrap;
+
white-space: -o-pre-wrap;
+
word-wrap: break-word">
+
 
[11:03] Franziska Bossi: eep!!
 
[11:03] Franziska Bossi: eep!!
 
[11:03] Franziska Bossi: who eeps me
 
[11:03] Franziska Bossi: who eeps me
Line 379: Line 375:
 
[12:02] Richardus Raymaker: Thats maby best option to get less confusion
 
[12:02] Richardus Raymaker: Thats maby best option to get less confusion
 
</pre>
 
</pre>
 +
 +
 +
[[Category:Office Hour Logs]]

Latest revision as of 15:47, 24 October 2015

[11:03] Franziska Bossi: eep!!
[11:03] Franziska Bossi: who eeps me
[11:03] Richardus Raymaker: there comes the misisng part :)
[11:03] Franziska Bossi: :P
[11:03] Key Gruin: the ones must rebake slower than the zeroes
[11:04] Franziska Bossi: yayaya
[11:04] Sarah Kline: Yes I say auto return the lot
[11:04] Richardus Raymaker: strange, zero is bigger then 1
[11:04] Franziska Bossi: -*- loOl -*-
[11:04] Tiffany Sicling: Which one is faster, the zero or the one ?
[11:04] Franziska Bossi: Jake followed me
[11:04] mr.core @tenebrae.odosys.net:8002: just don't divide by zero
[11:04] Key Gruin: hello Justin
[11:04] Franziska Bossi: duno where he is
[11:04] tx Oh: hi justin
[11:04] Richardus Raymaker: i would say the 1 bakes faster then the 0
[11:04] Richardus Raymaker: hi justin
[11:04] Justin Clark-Casey: hello folks
[11:05] Franziska Bossi: hi Justin
[11:05] mr.core @tenebrae.odosys.net:8002: hi justin
[11:05] Sarah Kline: hi Justin
[11:05] Spike Miles: hiya :)
[11:05] Richardus Raymaker: hi spike
[11:05] Key Gruin: well, zero is alwasy first
[11:05] Richardus Raymaker: did we lost justin ?
[11:05] Tiffany Sicling: when you take one away, zero is still there.
[11:05] Justin Clark-Casey: lost?
[11:05] Franziska Bossi: there he is
[11:05] Franziska Bossi: -*- loOl -*-
[11:06] Richardus Raymaker: i dont see you anymore justin
[11:06] Key Gruin: geez I said that 45 seconds ago
[11:06] Tiffany Sicling: hi Justin
[11:06] Richardus Raymaker: oh tehre you are. needed to move my camera
[11:06] Justin Clark-Casey: I wish the Linden Lab viewer wouldn't constantly reset its preferences
[11:06] mr.core @tenebrae.odosys.net:8002: lat chag
[11:06] Franziska Bossi: `*•.¸(`*•.¸ lool ¸.•*´)¸.•*´
[11:06] tx Oh: i use teapot
[11:06] Key Gruin: chag
[11:07] Key Gruin: for short
[11:07] Justin Clark-Casey: ha
[11:07] Richardus Raymaker: think bluewall is not comming... ?
[11:07] Key Gruin: well, it's getting faster and faster
[11:07] Tiffany Sicling: Bluewall has been idle almost 3 hours on irc
[11:08] Justin Clark-Casey: perhaps he just happens not to be around right no
[11:08] Justin Clark-Casey: now
[11:08] Richardus Raymaker: for justin (and others) i did lost of testing sicne last week, and the conclusion i now have is that ScheduleGroupForTerseUpdate (and llsetpos to) is leaking memory. or the backend. made a simple test script. http://opensimulator.org/mantis/view.php?id=6279
[11:09] Richardus Raymaker: unless there's other explenation why memory is growing with that example
[11:09] Justin Clark-Casey: you are Digi Fly?
[11:09] Richardus Raymaker: Yes
[11:09] Richardus Raymaker: mistake from the begin.
[11:09] Richardus Raymaker: but you cant change name
[11:09] tx Oh: i like to know one thing, i use dynamic textures a lot but because they are only stored in the region cache (flotsam) they got regulary expired. the prims look stupid when the textures can't get loaded from the cache anymore. can we tweak the cache to allow a 'sticky bit' on dynamic texures?
[11:10] Richardus Raymaker: did someone where not looking in the dynamic textures to improve it ?
[11:10] Justin Clark-Casey: tx: you're trying the new dynamic texture reuse setting?
[11:10] Nebadon Izumi: hey Douglas
[11:11] Douglas Maxwell: How it going?
[11:11] Richardus Raymaker: Wish i could change my name on mantis.. without createing new account.
[11:11] Douglas Maxwell: Hi Justin, long time no see
[11:11] Richardus Raymaker: hello douglas
[11:11] tx Oh: reuse setting? how is that?
[11:11] Justin Clark-Casey: hi douglas
[11:11] tx Oh: something undocumented?
[11:11] Nebadon Izumi: I am not familiar with the reuse option either, I have been pretty much completely disconnected from everything for the past 10 days or so
[11:12] Justin Clark-Casey: tx: experimental, but from your response I assume you are not
[11:13] tx Oh: no, but i will experiment
[11:13] Justin Clark-Casey: tx: so you have cache settings that expire contents after 24 hours?
[11:13] Nebadon Izumi: hello Andrew
[11:13] Richardus Raymaker: hi andrew
[11:13] Andrew Hellershanks: Hello, everyone
[11:13] Justin Clark-Casey: richardus: there's no obvious reason why llSetLinkPrimitveParamsFasst() would leak meory
[11:13] Justin Clark-Casey: hi andrew:
[11:13] tx Oh: i can try expire time of 2 minutes
[11:14] Richardus Raymaker: Yes justin, but then why does it go up , up and up ?
[11:14] Justin Clark-Casey: richardus: has this always been the case?
[11:14] Nebadon Izumi: I have heard some people complain about it increasing memory
[11:14] Justin Clark-Casey: ricardus: I don't know - there are could be any numbe rof reasons
[11:14] Richardus Raymaker: Hard to say, i discoveryed this since a few weeks.
[11:14] Nebadon Izumi: but I have never been able to recreate it myself
[11:14] Richardus Raymaker: and since this week i get the same result with that script
[11:14] Andrew Hellershanks: Richardus, do you have a test script that demonstrates the issue?
[11:14] Nebadon Izumi: at one point i put like 100+ panels out
[11:14] Nebadon Izumi: and let them run for 24+ hours
[11:14] Nebadon Izumi: and memory never increased
[11:14] Justin Clark-Casey: tx: well, the workaround would really be to set expiry time to never
[11:14] Richardus Raymaker: see mantis http://opensimulator.org/mantis/view.php?id=6279
[11:14] Andrew HellershanksAndrew Hellershanks looks
[11:15] Nebadon Izumi: that is what I said to Justin, but he had a good point of then nothing ever expires
[11:15] Nebadon Izumi: so it bloats up a lot
[11:15] Richardus Raymaker: i did the last test with that, and it still did go up. also on standalone. to rule search and neighborns out
[11:15] tx Oh: justin: just tell me about the reuse option, how can i try it with that?
[11:15] Justin Clark-Casey: tx: but in general I can imagine this is an issue
[11:15] Justin Clark-Casey: tx: reuse won't help you here - it just reuses existing generated dynamic textures if the input command s& params are the same
[11:16] Justin Clark-Casey: but if the textures are being expired anyway from the cache after 24 hours you will see the same problem
[11:16] Andrew Hellershanks: There are quite a few different calls in that script.
[11:17] Richardus Raymaker: well, yu need to move the prim , and this the smallest example i could make to let it run endless andrew
[11:17] tx Oh: yeah, but in the end i need to have a script in each panel.. well.. i think a good option would be a tweak on the flotsam cache. if i could determine if it is a dynamic texture then i could exclude them from standard expire
[11:18] Andrew Hellershanks: Richardus, just as a check, what happens to memory if you use that script but comment out the set prim params call?
[11:19] Richardus Raymaker: Not tested that...
[11:19] Nebadon Izumi: ya I guess that is the trick
[11:19] Nebadon Izumi: I think right now there is probably no way to determine that
[11:19] Richardus Raymaker: i can start that test on standalone..
[11:19] Andrew Hellershanks: That would tell you if the problem is related to that call or something else.
[11:19] Nebadon Izumi: to flotsam its just another texture
[11:20] Justin Clark-Casey: nebadon: yes. As tx says, one might want to stop expiry of textures that are only in the cache but not in the asset service (because they are temporary, or perhaps local)
[11:20] Nebadon Izumi: yea
[11:20] Nebadon Izumi: it makes sense for sure
[11:20] Richardus Raymaker: andrew you say use the scipt withoy llSetLinkPrimitiveParamsFast but keep the otehr part ?
[11:21] Hiro Protagonist: thats essentially forming an alternative local asset service
[11:21] Andrew Hellershanks: Richardus, um... right. Comment out llsetlink...
[11:22] Nebadon Izumi: ya i guess the other option is give the dynamic OSSL texture stuff its own storage location
[11:22] Nebadon Izumi: not sure how tricky that would be though
[11:22] tx Oh: hiro: that was the other way i thought
[11:22] Andrew Hellershanks: I've started a project using C# so I'm understanding the language a bit more but I haven't had to deal with object ref counting yet
[11:22] Justin Clark-Casey: nebadon: no, it really would be best done in the same channels as everything else
[11:22] Nebadon Izumi: ya
[11:22] Nebadon Izumi: that would be more sensible I think too
[11:23] Nebadon Izumi: question is how do you flag them?
[11:23] Nebadon Izumi: some kind of list?
[11:23] Nebadon Izumi: or modify the UUID?
[11:23] Justin Clark-Casey: well, there's already a local flag for each asset
[11:23] Nebadon Izumi: ah
[11:23] Richardus Raymaker: andrew its running.
[11:23] Justin Clark-Casey: or one could simply not flag them as temporary
[11:24] Nebadon Izumi: ya
[11:24] Nebadon Izumi: that should be optional though too I think
[11:24] Justin Clark-Casey: although then you have to stop them getting to the permanent asset cache....
[11:24] Nebadon Izumi: cause that could get messy too
[11:24] Andrew Hellershanks: Richardus, ok. I'm looking at the code behind that call
[11:24] Justin Clark-Casey: though at the moment if that doesn't happen you would have problems moving anything using dynamnic textures to another simulator
[11:24] Nebadon Izumi: i dont think you can anyway can you?
[11:24] tx Oh: no, they should stay in the cache only
[11:24] Nebadon Izumi: since its local temp only?
[11:25] Justin Clark-Casey: you can move the object, but the texture will didsappaer until you regenerate it manually in script
[11:25] Nebadon Izumi: right
[11:25] Richardus Raymaker: tried to look in C# code to andrew. but i got stuck after the call seems to go to other function :O
[11:25] Nebadon Izumi: there is no avoiding that
[11:25] Justin Clark-Casey: well, unless you were to upload it to the asset service but that's another kettle of fish
[11:25] Andrew Hellershanks: I'm at the setLinkPrimitiveParams call
[11:25] Justin Clark-Casey: or do something else that is tricky
[11:26] Nebadon Izumi: ya
[11:26] Andrew Hellershanks: um... setLinkPrimParams
[11:26] Nebadon Izumi: if that is vital for your own simulaotrs
[11:26] Nebadon Izumi: you can just set flotsam to share cache location?
[11:26] Nebadon Izumi: then it actually would work
[11:26] Nebadon Izumi: in your own sims
[11:26] Justin Clark-Casey: share cache location?
[11:26] Nebadon Izumi: yes
[11:26] Nebadon Izumi: its configurable now
[11:26] Justin Clark-Casey: oh I see, point all simulators to a single location
[11:26] Nebadon Izumi: yes
[11:26] Richardus Raymaker: Andrew, in this case test is on linux standalone. not have soemthing else handy. but i know how it works. just need to wait 1 day to see something. (only run 1 script now
[11:27] Justin Clark-Casey: yes, you could do that, though I'm not completely sure that there aren't some edge cases if, for instance, two simulators try to write to the same asset cache location at once
[11:27] Justin Clark-Casey: I think some people are doing it though
[11:27] Nebadon Izumi: ya true
[11:27] Nebadon Izumi: i suspect that would be rare
[11:27] Justin Clark-Casey: yes, but rare sooner or later ends up causing trouble :)
[11:27] Nebadon Izumi: heh
[11:27] Richardus Raymaker: Justin, i stopped with sharing assets cahce. i now have for every simulator its own.
[11:28] Justin Clark-Casey: anyway, possibly the local flag should really signal that an asset should not be removed from cache on expiory
[11:28] tx Oh: ooops
[11:28] Justin Clark-Casey: I mean, should not be expired like other assets if the assret cache is doing expiry
[11:28] Simulator Version v0.5 shouts: OpenSim 0.7.5 Dev          15d5f3d: 2012-09-04 00:11:14 +0100 (Unix/Mono)
[11:28] tx Oh: the bench disappeeared
[11:28] Richardus Raymaker: yes tx. here to
[11:28] Arielle.Popstar @bound2.dyndns.org:9005: oh good, not just me
[11:29] Justin Clark-Casey: not me!
[11:29] mr.core @tenebrae.odosys.net:8002: i thought things felt a little flimsy
[11:29] Sarah Kline: looks at Key
[11:29] Justin Clark-Casey: I mean, I 'm not responsible
[11:29] Taarna Welles: O-o
[11:29] Richardus Raymaker: its just gone complete on my viewer. wireframe to
[11:29] Richardus Raymaker: bedbugs ?
[11:29] Taarna Welles: yep same for me
[11:29] Sarah Kline: somebody deleted it by accident
[11:29] Taarna Welles: floating pillow :)
[11:29] tx Ohtx Oh sighs
[11:30] Justin Clark-Casey: anyway, until someone owns up - any other opensimulator topics today?
[11:30] tx Oh: the expire routine just didn't examine what kind of asset gets deleted
[11:30] Key Gruin: sorry - I did a restore to position but seems to no be working
[11:31] Key Gruin: it was me
[11:31] tx Oh: and if it would it might be a bit slow
[11:31] Tiffany Sicling: does osMessageObject still work?
[11:31] Justin Clark-Casey: heh
[11:31] Richardus Raymaker: i did mine for today.. thats pretty bugger to see your memory go up with 1GB/day
[11:31] Justin Clark-Casey: tiffany: It should...
[11:31] Richardus Raymaker: in case i run my train..
[11:31] mr.core @tenebrae.odosys.net:8002: key was looking for loose change again
[11:31] Tiffany Sicling: what OS level does it want, I cannot find that on the wiki
[11:31] Justin Clark-Casey: tiffany: os level?
[11:31] Sarah Kline: classic Key your a legend
[11:31] Richardus Raymaker: threadlevel i guess
[11:32] Tiffany Sicling: yes, none, verylow, low, etc.
[11:32] Nebadon Izumi: huh, is it me or did the couch disappear?
[11:32] Richardus Raymaker: lol
[11:32] Nebadon Izumi: hope its just my viewer
[11:32] mr.core @tenebrae.odosys.net:8002: lol
[11:32] Nebadon Izumi: lol
[11:32] tx Oh: well, there is another 'bug' with dynamic texutres
[11:32] Justin Clark-Casey: you're slow, neb :)
[11:32] Sarah Kline: lols
[11:32] Key Gruin: lol
[11:32] Nebadon Izumi: i minimzed viewer and came back
[11:32] Tiffany Sicling: couch disappeared to me also
[11:32] Richardus Raymaker: no its not you. thats gone for at least 1 minute
[11:32] Nebadon Izumi: dang what happened?
[11:32] Justin Clark-Casey: tiffany: it is on the wiki
[11:33] Key Gruin: it was an accidental right click
[11:33] Justin Clark-Casey: http://opensimulator.org/wiki/OsMessageObject
[11:33] Tiffany Sicling: I tried search and it asked me if I wanted to make a page for it hehe
[11:33] Nebadon Izumi: k
[11:33] Nebadon Izumi: i'll fix
[11:33] Nebadon Izumi: and lock it
[11:33] Nebadon Izumi: lol
[11:33] tx Oh: osSetPrimitiveParams didn't work with dynamic texures
[11:33] Dahlia Trimble: nice couch :) I have one just like that
[11:33] Key Gruin: lock it down lol
[11:33] Richardus Raymaker: you needed one key ?
[11:33] Key Gruin: I see it in inv and tried to restore it
[11:34] Justin Clark-Casey: tx: what aspect of osSetPrimitiveParmas() doesn't work?
[11:34] tx Oh: set a dynamic texture with osSetPrimitiveParams
[11:34] Justin Clark-Casey: you mean setting texture?
[11:34] tx Oh: yes
[11:34] mr.core @tenebrae.odosys.net:8002: ok, i guess i won't get you that 5 button mouse for xmas key - we don't need space-time to collapse ;)
[11:34] Richardus Raymaker: whats the osSetPrimitiveParmas() have extra compared to LSL one ?
[11:35] Key Gruin: hehe
[11:35] Justin Clark-Casey: tx: that's odd. Does llSetTExture(0 work?
[11:35] tx Oh: richardus: you can manipulate prims out of a linkset
[11:35] tx Oh: justin: didn't tried that
[11:35] Richardus Raymaker: aha, try to find the page...
[11:36] Richardus Raymaker: seems not on wiki ?
[11:36] tx Oh: http://opensimulator.org/wiki/OsSetPrimitiveParams
[11:36] Nebadon Izumi: eh close enough
[11:36] Nebadon Izumi: hehe
[11:36] Richardus Raymaker: http://opensimulator.org/wiki/OsSetPrimitiveParams
[11:36] Richardus Raymaker: lol
[11:36] Justin Clark-Casey: tx: That would show whether it's a general issue or some parcitular problem with osSetPrimitiveParams()
[11:37] Richardus Raymaker: curious if that one leaks also memory ..
[11:37] Richardus Raymaker: hmm its not the fast one
[11:38] tx Oh: justin: i made a picasa album rezzer and wanted to make the rezzer object to set the texture but that didn't worked
[11:38] Richardus Raymaker: is the OsSetPrimitiveParams have a delay to ?
[11:38] tx Oh: so each panel has a script now :-(
[11:39] tx Oh: isn't delay something the lindens invented?
[11:39] tx Oh: i mean for political reasons
[11:39] Dahlia Trimble: ya but they call it "lag"
[11:39] Justin Clark-Casey: tx: political? I'm sure delays are essential for perf reasons
[11:40] Richardus Raymaker: i see that with llSetLinkPrimitiveParamsFast and llSLeep to, if you really go to fast you see in the ctrl-shift-1 stats..
[11:41] Nebadon Izumi: llSleep suuuuuucks
[11:41] Richardus Raymaker: and my server is not short on cpu power
[11:41] Nebadon Izumi: avoid using it at all costs
[11:41] Richardus Raymaker: but there is no replacement nebadon
[11:41] Nebadon Izumi: yes there is
[11:41] Nebadon Izumi: timer
[11:41] Richardus Raymaker: no timer cant go below 0.1 and thats to slow
[11:41] Nebadon Izumi: use states
[11:41] Nebadon Izumi: well use a counter then
[11:41] Richardus Raymaker: states screw everything
[11:41] Nebadon Izumi: hehe
[11:41] Nebadon Izumi: not as much as llSleep does
[11:41] tx Oh: i use llSleep only for debug purposes
[11:42] Richardus Raymaker: sofar i have no problem with the llSleep(0.025);
[11:42] Richardus Raymaker: but i take the suggestion for later to replkace it by counter
[11:42] tx Oh: hehe..
[11:42] Richardus Raymaker: or at least test with it, but keep thinsg simple use for now llsleep
[11:43] tx Oh: i'm not sure if llSLeep(0.025) will not sleep for 2 seconds
[11:43] Richardus Raymaker: no tx it dont sleep for 2 seconds.
[11:43] Richardus Raymaker: thats easy spottable on moveing prim. thats why timer is useless.
[11:44] Daniel Allen: hi all
[11:44] Richardus Raymaker: timer is default to slow. and that make things more complex
[11:44] Ubit Umarov: 25ms is not asking a bit 2 much err or less ?
[11:44] Justin Clark-Casey: Hi dnaiel
[11:44] Justin Clark-Casey: daniel
[11:45] Richardus Raymaker: ubit
[11:45] tx Oh: did we come to a conclusion about the flotsam and dynamic texture handling? i'm not sure if i missed something
[11:45] Richardus Raymaker: ??
[11:45] Justin Clark-Casey: llSleep translates into a c# sleep, which I generally read does not have an accurancy better than 10ms
[11:45] Justin Clark-Casey: tx: my tentatuive conclusion is that this is an issue. One workardoun is to set the cache to never expire. The other might be to stop the caching expiring if the asset is flagged as local
[11:45] Richardus Raymaker: oh well 0.020 or 0.030 its both good. but 0.1 for timer is to slow for all things i use it
[11:45] Justin Clark-Casey: But this would require some investigatio nand code work
[11:46] tx Oh: i would like the second option. but that means that i need to check each asset before it gets expired, right?
[11:47] Justin Clark-Casey: no, they would simply not actually be expired
[11:47] Nebadon Izumi: just flag em so they are ignored
[11:47] Justin Clark-Casey: howver, that would lead to cache bloat so a cleverer option may be to still delete those assets but only do it on restart of th esimulator
[11:48] tx Oh: yepp
[11:48] Nebadon Izumi: couldnt we just check the DB
[11:48] tx Oh: or a expire-locals timers
[11:48] Nebadon Izumi: and say if this UUID exists in DB
[11:48] Nebadon Izumi: dont expire it
[11:49] Justin Clark-Casey: nebadon: that doesn't help since dynamic textues aren't in the db
[11:49] Nebadon Izumi: oh
[11:49] Nebadon Izumi: only in memory?
[11:49] Justin Clark-Casey: no, in cache
[11:49] Nebadon Izumi: i mean the prim db
[11:49] Justin Clark-Casey: unless that's configuered to be memory, of course
[11:49] Nebadon Izumi: for the surface its applied to
[11:49] Justin Clark-Casey: nebadon: I'm confused
[11:49] Justin Clark-Casey: you mean the uuid?
[11:49] Nebadon Izumi: yes the temp dynamic texture uuid
[11:50] Nebadon Izumi: if its still applied to any surface
[11:50] Nebadon Izumi: ignore it
[11:50] Justin Clark-Casey: that would require a scan of all region objects, and potentially all those inside other objects
[11:50] Nebadon Izumi: hmm ya i guess that could get nasty
[11:50] tx Oh: everytime i log in my message board is blanked... :-(
[11:50] Justin Clark-Casey: expensive and intricate
[11:50] Nebadon Izumi: ya
[11:50] Richardus Raymaker: hmm, auto refresh TX ?
[11:51] Justin Clark-Casey: tx: well, that would actually be quite odd becaus even if the texture is gone from the simulator, it should still be in your cache
[11:51] Justin Clark-Casey: client cache, that is
[11:51] tx Oh: yepp, client cache is another issue
[11:52] Richardus Raymaker: bye andrew
[11:52] Nebadon Izumi: hmm i wonder what the asset cache for this sim is set to expire for
[11:52] Andrew Hellershanks: I've got a few things going on where I am so I'm going to head out a little early. Almost top of the hour anyway
[11:52] Andrew HellershanksAndrew Hellershanks waves
[11:52] Nebadon Izumi: because that board with the OSgrid.org website on it
[11:52] Nebadon Izumi: in front of us
[11:52] Nebadon Izumi: is a dynamic texture
[11:52] Nebadon Izumi: and it rarely turns white
[11:52] Justin Clark-Casey: bye andrew
[11:52] Sarah Kline: bye andrew
[11:53] tx Oh: byr andrew
[11:53] Nebadon Izumi: its 48 hours here
[11:53] Richardus Raymaker: think thats the default setting
[11:53] Nebadon Izumi: i guess my viewer is caching it
[11:53] Justin Clark-Casey: ther are some issues related to pixel size and data size of textures
[11:53] Nebadon Izumi: probably why i still see it
[11:54] Nebadon Izumi: then occasionally it poofs
[11:54] Nebadon Izumi: i never thought of that
[11:54] Justin Clark-Casey: possibly it is getting expired
[11:54] tx Oh: i tried with expire times of 2 minutes and client cache clearing
[11:54] Nebadon Izumi: ya it probably is
[11:54] Nebadon Izumi: and i still see it until i clear viewer cache
[11:54] Nebadon Izumi: but noobs would just see white screen
[11:54] Justin Clark-Casey: yeah, that might be happening
[11:55] tx Oh: yes, thats the bad thing. the white screens
[11:55] tx Ohtx Oh sighs
[11:55] Nebadon Izumi: thats why it was blank today
[11:55] Nebadon Izumi: when i logged in
[11:55] Nebadon Izumi: i had not logged in with viewer for like 10 days
[11:56] Nebadon Izumi: the weather boards update every 15 minutes or so
[11:56] Nebadon Izumi: thats why i never noticed this before
[11:56] Justin Clark-Casey: so another workaround would be to regenerate the tedxture every hour or so
[11:56] tx Oh: ok, i will try to find out what the local flags are, where they get set and how i can figure out which asset has it set in the flotsam routines
[11:56] Nebadon Izumi: I just assumed it was randomly occuring
[11:56] Nebadon Izumi: heh
[11:57] Justin Clark-Casey: tx: this is just an idea - I'm not 100% sure it is feasible to do it this wasy though it probably is
[11:57] Nebadon Izumi: well if you do come up with patch, please post it on mantis
[11:57] Nebadon Izumi: and let me or justin know
[11:57] Justin Clark-Casey: yes, please
[11:57] Nebadon Izumi: and we'll go from there
[11:58] tx Oh: the thing is, i even don't want to have a script in the panels and now i shall reload the url every hour in each panel :-(
[11:58] Justin Clark-Casey: I may end up looking at this issue myself though it might not be soon
[11:58] Nebadon Izumi: cool
[11:58] Nebadon Izumi: ya i would say this is really worthy of a fix
[11:58] Richardus Raymaker: well as long moap is not co9mmon used yes
[11:59] tx Oh: that would be sooooo cool.
[11:59] Nebadon Izumi: wouldnt effect that
[11:59] tx Oh: moap always starts a webkit. i think thats very intensive to the client with 50 images or so
[11:59] Justin Clark-Casey: well, one could use moap instead but it's very heavyweight solution
[12:00] Nebadon Izumi: ya
[12:00] Richardus Raymaker: hmm, right
[12:00] Justin Clark-Casey: tx: exactly - in fact I believe it's more like 6 or 7, epending on the client machine
[12:00] Dahlia TrimbleDahlia Trimble waves bye :)
[12:00] Justin Clark-Casey: bye dahlia
[12:00] tx Oh: bye dahlia
[12:00] Justin Clark-Casey: ok, I need to go and do some other work as well
[12:00] Richardus Raymaker: bye dahlia
[12:00] Richardus Raymaker: bye justin
[12:00] Justin Clark-Casey: see you folks later
[12:01] Sarah Kline: bye justin
[12:01] Hiro Protagonist: tc Justin
[12:01] VivK Lowlag: bye Justin
[12:01] tx Oh: the next thing i like to work on, is, adding some os camera control functions
[12:01] Richardus Raymaker: there's no way to change name on mantis ??
[12:01] Justin Clark-Casey: ricahrdus: just create a new acct - it's not a big deal
[12:02] Richardus Raymaker: Thats maby best option to get less confusion
Personal tools
General
About This Wiki