|
|
|
Moon Metty made changes - 15/May/08 03:16 PM
Scouse Linden made changes - 18/May/08 05:50 AM
Hi Scouse,
Sure! I'm surprised you can't reproduce, it happens 10% - 15% of the times. -Rez a box At the office hour on friday 16th we showed it to Vektor, and he reproduced it within a minute. "Yumi Murakami: when I detach an attachment with a running mono script, sometimes the script goes back to non-running and non-mono." I've tried this. It took about 20 times of reattaching, but sure enough, the script stopped running, the Mono-box unchecked. ======= Second Life 1.20.6 (87323) May 12 2008 17:18:14 (Second Life Mono) You are at 255832.9, 255380.0, 21.5 in Sandbox Cordova MONO located at sim3001.aditi.lindenlab.com (8.2.33.227:13000) Aha, when you take a misbehaving object in inventory and rez it or attach it, the Mono-box is checked again. The script will run if you check the running-box. (Unless you get the attaching-bug again)
Scouse Linden made changes - 19/May/08 02:14 AM
What do you mean by "Recompile it"?
What are the exact steps that you take? Hello, by recompile it means to save the script...
when you alter it, you can re-save it, when you re-save it the running bit, ( the check of running, just goes away ) it started when we got the new viewer: "Second Life 1.20.6 (87323)" here's a way of reproducing it fast: Using the viewer: "Second Life 1.20.6 (87323)" default } Check the mono box, and save it, normally you should get the message "I am running" now re-open it. edit it ( add an space ) and save it again you should get a new "I am running" message, then do it again... and again... and again... at one time or another ( mostly sooner ) you will never get the "I am running" message. then if you re-open the script, you will see that the "Running" checkbox is unchecked ! now, this is not the same, and this was reported to vektor, by myself using a notecard ( he told me he was going to file a Jira about it, but I have not checked it ) this exact same behavior but randomly shows when you have an object with multiple scripts, and yo go to: Tools - Recompile objects in selection - Mono ( or LSL ) Some will end up running, others will loose the running bit, others where in mono now they will loose the mono check, etc, that part the running and the mono check just goes away, and it is becomming a real issue and problem when beta testing scripts. Vektor is well aware of the problem, he can do a much better explanation than me or anyone else. Seems to be related to the load on the sim. I've tried those steps but can't get it to happen on my sim. I've managed to reproduce this on a sandbox.
I will investigate some more and we will hopefully finally fix this elusive bug. oh, related to sim load, that's interesting, and could be completly true, because of the randomness of the pattern it displays, there's another bug that also displays a random pattern, llDetectedKey(0)=NULL_KEY when using a freshly rezzed sensor, maybe that one is also related to the load of the sim, and some other random behaviors that moon and me have seen on mono. this gives us a light, on what is happening, also another script I coded, that oscillates, by self feedback, stops randomly, but when I added a sleep the problem got solved, but I also noticed that the error was in direct relationship to dilation of the sim, makes you think about this errors.
thx Scouse
Scouse Linden made changes - 20/May/08 09:51 AM
Scouse Linden made changes - 20/May/08 09:51 AM
Scouse Linden made changes - 20/May/08 02:51 PM
Raising priority for this to showstopper, as it is a serious blocker for testing. I have managed to get this to happen reliably by:
Teleporting I added this to seven scripts in a prim for a simple repro, incrementing scriptnum for each new script: // <Code> default { timer() {
llOwnerSay ("Script " + (string)scriptnum + " still running");
} Setup: Repro: 1) Add the above script to a prim 7 times, increment the scriptnum and save to mono for each. Results: Scripts are frequently being set to Not Running, as well as having their Mono bit toggled off. LSL2 compiled scripts seem to be ok. Expected Results: Mono scripts should be able to go through the above steps and remain running Mono scripts.
Vektor Linden made changes - 20/May/08 05:08 PM
Vektor Linden made changes - 20/May/08 05:09 PM
lindenrobot made changes - 20/May/08 05:10 PM
Put th priority back down as its already fixed and will be in the next release.
Scouse Linden made changes - 21/May/08 12:10 AM
I have the same problem with products that have a lot of scripts (40+). There is no logic to it, it just randomly disables scripts when rezzing or even when using the recompile function (to mono) from the Second Life viewer.
After manually recompiling all scripts and activating them... testing and working fine... then take the object back to inventory... rez again and randomly scripts are disabled and mono turned off again. There was nobody in the sim (cordova) and therefore sim load seems to have nothing to do with it. The fact that the fix is pending does not change the fact that it is a showstopper in the eyes of QA. Please do not change the priority.
Vektor Linden made changes - 22/May/08 11:48 AM
Me too. I have a 32-prim linked object, each prim containing the same script. If I use "tools/recompile-in-selection", I find that the root prim in particular can end up with the mono flag off, and running off, and need recompiling.
I may be wrong but I have the impression this was NOT happening a month ago.
Periapse Linden made changes - 28/May/08 10:24 AM
Periapse Linden made changes - 28/May/08 10:25 AM
Great, the mono scripts are running again. I will mark this bug as fixed.
Phew
Moon Metty made changes - 31/May/08 04:25 AM
Sue Linden made changes - 13/Nov/08 12:10 PM
Sue Linden made changes - 13/Nov/08 04:44 PM
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Can you give more specific details like how many attempts you made to compile, the specific scripts you've had problems with.