96f0c6f074
More on mantis #7567 . Two things:
...
- Increase the inventory cache timeout to 20 secs, so that the items will still be there when they are needed by the sim for rezzing
- Before rezzing attachs, make a call to GetMultipleItems so to fetch them all at the same time
2015-06-05 10:07:50 -07:00
cc22d81b4d
The only purpose of this commit is to remove an erroneous comment. It turns out that rezzing the attachments is necessary in modern viewers too. Proof: commenting the lines below the comment results in no attachments seen in the avatar.
2015-06-05 09:47:36 -07:00
52d9ea1cee
Mantis #7603 -- bad folders in inventory could produce null pointer exception. Thanks for the line numbers in the exception trace.
2015-06-04 19:57:01 -07:00
83267e19f6
Mantis #7536 . This may alleviate the side effects of broken SceneObjects XML. It doesn't (can't) fix the cause.
2015-06-04 16:17:45 -07:00
51fccff235
Mantis #7600 . The warning message was not a symptom of a problem, it was just that the HG client was trying to fetch folders that aren't in the Suitcase, so that fails. This should make the warning message go away in that situation. (fetching those folders will still fail, because they aren't available over HG)
2015-06-04 15:04:36 -07:00
f63e7ec46f
Trivial: fix copy-paste typo in string
2015-06-04 09:30:41 -07:00
c4ecd1946b
Fixed order of projects, so that the solution builds in mono.
2015-06-04 09:20:34 -07:00
2fff376762
More assertions in inventory client tests to check for assorted combinations of cached/non-cached/existing/non-existing items
2015-06-04 09:17:12 -07:00
6f469ec0a5
Added tests that verify the absence of the bug fixed in the previous commits (cache issues).
2015-06-04 08:52:57 -07:00
a433a40980
Making the inventory cache static, so that there's less timers.
2015-06-04 07:10:06 -07:00
ea5aa82c2c
Mantis #7567 . Once again, avoiding prefetching linked items within linked folders. Also fixing the inventory connector GetMultipleItems, so that if everything is in the cache, it returns successfully rather than unsuccessfully.
2015-06-04 06:58:07 -07:00
bac53387a9
Mantis #7567 : added an 8-sec expiring item cache to the inventory network connector. This fixed the problem on my local test grid and generally made things faster. This cache has been needed for a while... there are many parts in the code where the sim gets an item multiple times in a short amount of time (rezzing attachs and objects, for example).
...
Other minor changes:
- added the scene as a parameter to the constructor od FetchInvDescHandler, so that I could see in which scene the handler was being called
- brought linked items in linked folders back to being prefetched
2015-06-03 17:42:41 -07:00
3141664d6b
I suspect the viewer doesn't need the target of linked items inside linked folders to go in the reply of the original request. At least my tests indicate that. Pushing this out, so that others who use linked folders a lot more can verify.
2015-06-02 16:31:56 -07:00
0531448664
New unit tests for FetchInventory2 cap.
2015-06-02 15:48:16 -07:00
be6fb22d69
Mantis #7567 . One of the reported log messages showed this:
...
09:38:40 - [LOGHTTP]: Slow handling of 15572 POST /CAPS/b12c7e98-8261-4953-b7d1-1c414c9893fc FetchInventory2 8acfbca3-13b5-434f-898c-5f4bbe8a76ff from 92.237.199.112:60083 took 62391ms
FetchInventory itself wasn't taking advantage of the new inventory API. This commit fixes that.
2015-06-02 15:40:13 -07:00
8de65a8fb3
Mantis #7594 : putting things as they were before regarding duplicate removal. Also added test to check that duplicates are being removed. The test passes. I have no idea how duplicates would not be removed, as reported in the mantis.
2015-06-01 20:05:51 -07:00
442c025335
Mantis #7594 . Fixing the broken code I just introduced.
2015-06-01 18:57:01 -07:00
7b5e97f04f
More on mantis #7594 , this a=time addressing the reported exception, which seems to be a separate issue from the duplicate folders.
2015-06-01 16:37:04 -07:00
7effd3b158
Mantis #7594 . This should be functionally equivalent to what it was, but just in case mono has a bug in List<T>.Find, here is the Linq equivalent of distinct-ness.
2015-06-01 15:11:30 -07:00
897c7e1bda
Improve configuration description for llCastRay V3.
...
Signed-off-by: Michael Cerquoni <nebadon2025@gmail.com>
2015-06-01 13:09:21 -04:00
dcf31ee762
Correct constant RCERR_CAST_TIME_EXCEEDED.
...
Signed-off-by: Michael Cerquoni <nebadon2025@gmail.com>
2015-06-01 13:09:05 -04:00
22e0367203
Add an entry for osGetGender to config-include/osslEnable.ini. Thanks to Austin Tate for the suggestion.
2015-05-30 15:10:21 -07:00
5d3c327e93
Add a new os function "osGetGender()"
2015-05-29 23:02:12 -07:00
0af17c9484
llListRandomize() wasn't very random
...
Signed-off-by: Diva Canto <diva@metaverseink.com>
2015-05-27 11:15:09 -07:00
f0533132eb
Merge branch 'master' of ssh://opensimulator.org/var/git/opensim
2015-05-26 15:06:39 -07:00
5e09e52ee9
Attachments Module. Fix detach event not being fired until the next time the object is attached.
...
Not an ideal fix but this allows scripts such as AOs to remove animations when detached etc. The pause added does not affect other avatars or the scene in general and only pauses the avatar performing the detach for an extra 2 milliseconds.
Signed-off-by: Diva Canto <diva@metaverseink.com>
2015-05-26 15:05:32 -07:00
d271151504
fix file perms and change AllowLightshareFunctions to AllowLightShareFunctions
2015-05-25 20:52:27 -04:00
22ee1c30e3
Merge branch 'master' of ssh://opensimulator.org/var/git/opensim
2015-05-25 15:19:27 -07:00
35d58fe9d4
Minor change of statement order in unit test
2015-05-25 15:19:00 -07:00
abb3bb60a6
Fix mesh cache race condition in llCastRay V3.
...
Signed-off-by: Michael Cerquoni <nebadon2025@gmail.com>
2015-05-23 08:14:28 -04:00
defc9c1d86
fix permmission issue on OpenSimDefaults.ini
2015-05-22 01:22:17 -04:00
fdc383a843
Use own mesh cache in llCastRay V3 after throttling.
...
Signed-off-by: Michael Cerquoni <nebadon2025@gmail.com>
2015-05-22 01:21:50 -04:00
7d26815d0e
Throttle llCastRay V3.
...
Signed-off-by: Michael Cerquoni <nebadon2025@gmail.com>
2015-05-22 01:21:18 -04:00
edcba95ab5
Merge branch 'master' into moses
...
Conflicts:
OpenSim/Region/Framework/Scenes/Scene.cs
2015-05-20 23:40:52 +02:00
d9afe354ac
New values returned by llGetEnv
...
Implements recent changes to llGetEnv
region_cpu_ratio hardcoded to 1
region_product_sku hardcoded to OpenSim
Signed-off-by: Michael Cerquoni <nebadon2025@gmail.com>
2015-05-20 14:36:54 -04:00
946a6622fe
Removing duplicate definition
2015-05-19 15:52:13 -04:00
140ea04b9d
Merging Opensim upstream before generating patch
2015-05-19 15:18:45 -04:00
714c082ca8
Merge pull request #7 from gamucf/moses.metricsPhase2
...
Moses.metrics phase2
2015-05-19 14:40:49 -04:00
91a55c0794
Merge branch 'master' of https://github.com/M-O-S-E-S/opensim into moses.metricsPhase2
...
Fixed conflicts.
Conflicts:
OpenSim/Region/Framework/Scenes/Scene.cs
2015-05-19 14:07:44 -04:00
3efcef7d7e
Merge branch 'moses.metricsPhase2' of git.irl.ucf.edu:/irl/git/opensim into moses.metricsPhase2
2015-05-18 16:36:06 -04:00
6a48c87cd2
resolve possible nullref when sending appearance packet. Thanks to zadark for pointing this out.
2015-05-13 15:59:58 -07:00
3c2886cd75
Adding back command to clear region's user cache
2015-05-13 14:31:28 -05:00
cb586d1158
Enable grab feature (Ctrl+Drag) for non-physical link-sets and add code to handle spin (Ctrl+Shift+Drag)
...
This patch fixes permission issues with dragging scripted objects. As on LL grid, scripted prims can not be dragged. Also after dragging, the group was not marked as updated. This meant that after the region was restarted the group would revert to its original position. This version fixes issues where scripts in child prims were not detected and also blocks grab for attachments.
Signed-off-by: BlueWall <jamesh@bluewallgroup.com>
2015-05-13 12:16:14 -05:00
7566d22a18
update libomv to 6b8afb2b61d3
2015-05-12 22:54:53 -07:00
d348f87161
Patch llCastRay fully simplified to V3.
2015-05-12 15:40:42 -07:00
d52f58a733
Disable the "show threadpool calls active" console command for now.
...
Whilst this can be very useful to identify threads held too long by incoming packet processing,
it can also be very misleading as the count updates are not atomic (race conditions lead to growth in numbers over time).
This could be solved by making the counts volatile but I don't currently have time to assess the performance impact, though it's almost certainly negligible.
If anybody else wants to do that and re-enable this command then please feel free.
2015-05-12 19:54:25 +01:00
aefed57158
One more test in UserAccountsClient.
2015-05-12 08:55:32 -07:00
9366165b34
Again, forgot to update Prebuild.xml regarding the deleted test client project. Done now.
2015-05-11 20:04:01 -07:00
923a57a91f
Added tests for UserAccountService in Robust. In the process fixed a couple of bugs in the network connectors. For some reason the robust-bound code had a CreateUser method, while the client-bound code had no such method. I assume someone is extending the client-side code with their own connectors. I added the missing method, but didn't add it to the service interface.
2015-05-11 19:55:23 -07:00
d0292b0731
Forgot to delete an unused project from Prebuild. Done now.
2015-05-11 09:12:55 -07:00
Diva Canto