Test of 0.6.7 post-fixes

classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|

Test of 0.6.7 post-fixes

Ai Austin
I thought I would test the current 0.6.7 post-fixes and tool r/11301
from the GIT web interface zip download.

It prebuillt and compiled fie (the usual warnings on used code
elements - 44 I note).

I already had a 0.6.6 setup (r/9903) with Hypergrid, UGRM (R=AI)
setup with 14 .xml specified region files

I brought over the previous changes to the config files into the new
version as usual and stayed with the UGRM (R=AI) setup.

I just had to alter references in my startup scripts from
Opensim.Services.exe to Opensim.Server.exe and of course change
incorporate the previous data base connect string changes into the
similarly renamed Opensim.Server.ini

Good news is that the system came up fine.

Problems that I spot in initial tests:

a) My avatar is stood with its ankles about 0.1m to 0.2m underneath
surfaces (whether the object surfaces are 0.5m or 2m deep its still
the same depth of penetration. If I go on solid terrain, the avatar
starts failing, drops through the terrain surface, etc. Or if you
move it a bit, it has ankles below ground with bent knees. If I go
back on to an object, the avatar straightens up, but still has anjles
below ground.

b) Using the new region .ini style files, the
http://virtual.aiai.ed.ac.uk:8002/index.php?method=regionImage1111111...
style of call to show region maps on a web page does not work .  This
is same as 0.6.6.  I think Melanie may have contributed a patch to
fix that in later versions than 0.6.7?  I wonder if someone has the
knowledge to incorporate that change in the 0.6.7 postfixes before
its stable... as it is a visible issue.  I reverted to .xml region
files and it works fine.

c) Having now disabled DotNetEngine to be ready for the change later
in 0.6.8, I note one minor extra error I have not seen before...
indicating System.Invocation.TargetInvocationException saying
llRemoveVehicleFlags is not implemented.  Its just in a script in a
vehicle brought over from SL that is not working anyway, so I
disabled it for now.#

d) I have had one issue when the regions went down while a Viewer
logged in but could not repeat that, but will do more testing now.

Hope these short notes are useful, Ai


_______________________________________________
Opensim-users mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/opensim-users
Reply | Threaded
Open this post in threaded view
|

Re: Test of 0.6.7 post-fixes

Ai Austin
At 10:49 30/10/2009, Ai Austin wrote:

>I thought I would test the current 0.6.7 post-fixes and tool r/11301
>from the GIT web interface zip download.
>... with Hypergrid, UGRM (R=AI) setup with 14 .xml specified region files
>a) My avatar is stood with its ankles about 0.1m to 0.2m underneath
>surfaces (whether the object surfaces are 0.5m or 2m deep its still
>the same depth of penetration. If I go on solid terrain, the avatar
>starts failing, drops through the terrain surface, etc. Or if you
>move it a bit, it has ankles below ground with bent knees. If I go
>back on to an object, the avatar straightens up, but still has
>anjles below ground.

I notice that this occurs on every startup with 0.6.7 postfixes.  But
then if you leave things alone after login for about 50 seconds, I
get a Opensim console message

[OGS1 USER SERVICES] Updated User appearance in
http://virtual.aiai.ed.ac.uk:8002

and then the avatar jumps up a bit to be in the right position. Could
it be that the avatar size information is benign passed through to
Opensim.exe  a lot later than in 0.6.6

However, as soon as I move the avatar in the client after that mesa
sage, it again drops down to be ankle deep in the object.  Cod it be
that the late update of avatar position is not being passed through
to the viewer client

_______________________________________________
Opensim-users mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/opensim-users
Reply | Threaded
Open this post in threaded view
|

Re: Test of 0.6.7 post-fixes

Ai Austin
In reply to this post by Ai Austin
At 10:49 30/10/2009, Ai Austin wrote:
>d) I have had one issue when the regions went down while a Viewer
>logged in but could not repeat that, but will do more testing now.



I have been testing for the last 2 days or so,. and about 1 time in
every 3 attempts to start up and log in, I get a message via the User
Server that the "


[LOGIN] Region is not available for login. System.Net.WebException:
the operation has time out.

I have to reboot the UGRM and opensim.exe to get things going
again.  Anyone else observed something like that with the 0.6.7
postfixes (11301) version?



_______________________________________________
Opensim-users mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/opensim-users
Reply | Threaded
Open this post in threaded view
|

Re: Test of 0.6.7 post-fixes

Ai Austin
In reply to this post by Ai Austin
More on testing 0.6.7 postfixes (r11301). I am using a UGRM setup
with R=AI and a number of regions all on same server.  MySQL is DB in use.

I observe in the message server has a regular message every 5 minutes saying

[SERVER] Registered with http://virtual.aiai.ed.ac.uk:8002

and that continues 3 or 4 times, then I get a red message

Unable to connect to server
http://virtual.aiai.ed.ac.uk:8002.  Server not running?

but after a while I see the successful register messages again. That
has been happening in a soak test overnight all night.

At times I can connect both locally, and over a connection from
outside the local network. , at others I cannot connect and see  a
user server message
[LOGIN] Region is not available for login. System.Net.WebException:
the operation has time out.

I also see this morning that when a user client logs in or logs out I
get a red message in the opensim,.exe console saying:

      Failed to notify message server of agent leaving for <UUID?

and the messaging server gives a red error saying:

      Unable to connect to server http://virtual.aiai.ed.ac.uk:8002.  Server n

At the same time there seem to be a number of data base connection
lost and reconnecting errors in the Grid and message servers.

Anyone else having these sorts of problems with 0.6.7?  My 0.6.6
(r9903) system on which this is based was very stable and did not
show these sorts of problems.

_______________________________________________
Opensim-users mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/opensim-users