Server Bugs

• Use concise, precise descriptions
• Do not include sensitive information.
• Create a support ticket at https://support.secondlife.com for individual account issues or sensitive information.
terrain export is badly encoded/quantised leading to significant resolution loss
See blog post https://www.blogger.com/blog/posts/8863246996882687669 for a longer write up. tl;dr version is... The terrrain files produced by the SL server when exporting terrain are not properly encoding the height, resulting in coarse quantised terrain. It would appear to choose to encode only the height value (R channel) leaving the G channel untouched. This means that the resolution of the land exported is limited to whatever value happened to be in the G channel at the time of creation/import. Untested repro steps, as I have no access to SL regions with terrain rights. My terrain tools should be usable but they do try to minimise the G value for any given height. The attached 20m-Q4 file is encoded to R=80, G=32, meaning that the minimum G step is in their 1/4 metre, so it should still be visible with a sharp gradient. 1 start with an empty region 2 load a flat terrain RAW file, if possible use one with a G channel suitably high to demonstrate the issue, 120 would be a good starting point. 3 in SL manipulate the terrain using the inworld tools, adjus the strength of the tools. 4 create a mount with a slow smoother gradient that is shallow enough and wide enough that the stepping should be obvious. For a G value of 120 a smooth mound of diamater 10m and height 10m should give ample evidence 5 see the smooth land in SL 6 bake and export 7 re-import explected results a stepped pyramid instead of a smooth mound.
6
·

tracked

Scripted Agents (bots) kept relogging on SL even when they are logged off from Client side + LL servers not checking if an account is really online at intervals adding to the weight of the servers. and affecting performance
Client side crashes are not reported to the LL servers. This means users that crash from client side are still showing as online to everyone on SL. This is the same case with the scripted agent (bots) run with Corrade as I have experienced. So what kind of ripple effect repercussions does this have? For an account (scripted agent/bot or not) that crashes from a viewer once your try to relog from viewer it will send update. First attempt to relog back in may be stuck as trying to connect to region but may not be successful. But 2nd attempt to relog via viewer is usually successful. My assumption as to why the first time fails is as it tries to relog back in the LL servers realise that the account has been offline so cut off its first connection/links to SL but fails to reconnect it back on first relog attempt. This needs refining and retuning done server side. On top of that viewer side issue the other issue comes when attempting to relog as a scripted agent(bot). So you might try to logout from command shell (not service) and during the logout process the bot crashes from LL server side without my knowledge. I on the client side am not aware of this and would just turn wifi off and shut down computer. But on the LL server side the bot has not properly logged out so after some seconds LL servers auto logs the scripted agent (bot) online from server side. Everyone on SL see this bot as online but it is not suppose to be online. Atleast not from client side. But to everyone on SL this bot is legit online and SL LL servers are sending all the data related to this bot. And obviously the region the bot is logged back onto is now having to host the bots data as well. Servers keep sending this data to every client that connects adding to server lag. Imagine how many accounts like this could be floating ghost accounts online on server side taking up valuable server resources. This is very bad for performance. This is dead weight. Needs fixing. EDITED : Going back to when the bot crashed while logoff, the bot was being logged back online by the LL servers. I tried to re-run Corrade from the command shell and now the command shell running gave errors. So I log the bot from viewer. First attempt failed. This was probably what I explained earlier. Could be another reason. But the 2nd attempt at relog from viewer worked. So I logged the bot back offline from viewer and I observe though the bot was logged offline from viewer and SL also shows and even sends out messages to ones in bots friends list that the bot is logged offline then suddenly back again LL servers log the bot back online all auto when its been logged offline successfully via viewer. I try this from viewer several times though viewer shows successful logouts and SL also sends friends that the bot went offline then seconds later LL servers log it back online. However then I try with command shell. Command shell no longer works. So I log the bot back on from viewer and log it off again in an attempt to try several times to get the LL servers to update the bot account to stay logged off when its logged off but though its logged off from client side viewer, still LL servers log it auto back online. So I try logging the bot via windows "Services" and it worked from services. Now the LL servers no longer logged bot back online after I log the bot offline from "Services". It keeps bot offline until I log the bot back online via Services or viewer. The Windows "Services" does something that both the viewers or the Corrade.Application command shell does not do to fix the issue where the LL servers auto logs bot back online whether the bot is logged offline from viewer or command shell. In Windows "Services" the bots startup type is set to Automatic. Would this give this type of behaviour on crash of a bot run via the command shell (Corrade - Corrade.Application)? The command shell (Corrade - Corrade.Application) still not worked. For about 12 hours or more the relog from command shell failed. Then suddenly the command shell logins also worked. So something is up there. I cannot pin point to why this happens via command shell. I dont know full backend details ins and outs to properly do some tests. And I am new to SL bots as I have worked with SL bots (Corrade) only 2 days. Plus recreating a crash issue is also there.
0
Load More