Jump to content

Deploys for the week of 2013-04-08 (Updated @ 2013-04-10 08:00PDT)

Maestro Linden

By Maestro Linden,
in Second Life Server

 Share
You are about to reply to a thread that has been inactive for 4019 days.

Please take a moment to consider if this thread is worth bumping.

Recommended Posts

  • Lindens
Maestro Linden

Maestro Linden

Posted
  • Lindens
Posted

Second Life Server (main channel)

No deploys are scheduled for the main channel this week. 

 

Second Life RC BlueSteel and Second Life RC LeTigre:

BlueSteel and LeTigre are on the same server maintenance project as last week.  This project has some bugfixes, but the most notable changes are new features in LSL.  The only update for this week was a design change to improve compatibility between the new animation override system and other scripted objects that animate avatars (such as poseballs).

https://wiki.secondlife.com/wiki/Release_Notes/Second_Life_RC_BlueSteel/13#13.04.03.273434

https://wiki.secondlife.com/wiki/Release_Notes/Second_Life_RC_LeTigre/13#13.04.03.273434

Testing for this week's planned project update for BlueSteel/LeTigre didn't complete in time, so they will instead get the project that was originally only planned for Magnum.  This project includes low level changes to how the simulator handles HTTP connections.

https://wiki.secondlife.com/wiki/Release_Notes/Second_Life_RC_BlueSteel/13#13.04.05.273580

https://wiki.secondlife.com/wiki/Release_Notes/Second_Life_RC_LeTigre/13#13.04.05.273580

Scheduled Wednesday 2013-04-10 07:00-11:00 PDT

 

Second Life RC Magnum:

Magnum is on the same project as last week.  This project includes low level changes to how the simulator handles HTTP connections.  The only change this week is to fix a crash mode.

https://wiki.secondlife.com/wiki/Release_Notes/Second_Life_RC_Magnum/13#13.04.05.273580

Scheduled Wednesday 2013-04-10 07:00-11:00 PDT

 

We will be monitoring this thread as the code gets released, so feel free to note any observations you have about the server updates.  If you have a specific bug you'd like to report, please file a Jira. 

Link to comment
Share on other sites
Qie Niangao

Qie Niangao

Posted
Posted

Second Life RC BlueSteel and 
Second Life RC LeTigre
:

[...] The only update for this week was a design change to improve compatibility between the new animation override system and other scripted objects that animate avatars (such as poseballs).

I gather that this change is to address the problem of llGetAnimationOverride returning the animation name, not its UUID (as does llGetAnimationList), so furniture that doesn't contain that animation cannot llStopAnimation(llGetAnimationOverride("Sitting")) as originally designed.

Off the top of my head, I don't see a problem with this new approach, and in fact it should make most existing furniture scripts "play nicely" with the new AO script functions, without change.

It will be interesting to see what happens when the scripted object contains an animation named "sit" (although that's a perilous situation for many animation scripts anyway).

Link to comment
Share on other sites
  • Lindens
Maestro Linden

Maestro Linden

Posted
  • Author
  • Lindens
Posted

Qie Niangao wrote:

Second Life RC BlueSteel and 
Second Life RC LeTigre
:

[...] The only update for this week was a design change to improve compatibility between the new animation override system and other scripted objects that animate avatars (such as poseballs).

I gather that this change is to address the
of llGetAnimationOverride returning the animation
name
, not its UUID (as does llGetAnimationList), so furniture that doesn't contain that animation cannot llStopAnimation(llGetAnimationOverride("Sitting")) as originally designed.

Off the top of my head, I don't see a problem with this new approach, and in fact it should make most existing furniture scripts "play nicely" with the new AO script functions, without change.

It will be interesting to see what happens when the scripted object contains an animation named "sit" (although that's a perilous situation for many animation scripts anyway).

Yep, that's the idea - the change allows objects like scripted chairs and 'legacy' AO HUDs to stop animations which have been overridden.

To answer your question, if llStartAnimation(), llStopAnimation(), or llSetAnimationOverride() are called on an animation that is both in the task inventory and is one of the built-in animations (such as 'sit'), the animation in task inventory is used.

Link to comment
Share on other sites
  • Lindens
Maestro Linden

Maestro Linden

Posted
  • Author
  • Lindens
Posted

Hi Nimh, we need more details in order to investigate those slowdowns.  At the very least, it would be good to know which region that was in.  If the slowdowns aren't frequently recurring (as in you wouldn't  see itafter visiting the sim for a few minutes), we would also require some timestamps of specific incidents, in order to investigate.  Could you file a Jira with these details?

Link to comment
Share on other sites
Lucia Nightfire

Lucia Nightfire

Posted
Posted

Seems we picked up a bug in 13.04.05.273580 for Bluesteel as requesting PERMISSION_OVERRIDE_ANIMATIONS from an attachment now generates a permissions request dialog. Will wait on LeTigre regions to restart to see if they do it with the new version.


I guess I see why now. We lost the Animation Override code with this version change. And just when I started working with making an AO...

Link to comment
Share on other sites
Nimh20 Vandeverre

Nimh20 Vandeverre

Posted
Posted

More Details? Maestro--this is across the board when it comes to Mainland.  My first post with Stats was right after Help Island Pubic was restarted.  My second post stats were from Korea1 but it doesn't matter what Mainland sim. your on, it seems like this is happening on most I visit.  I mean with these conditions anywhere on the Grid no matter where you are, how do you ever expect Valve to let you guys on Steam?  I can't believe you all don't have some kind of simulator monitoring system of the about 27,000 sims showing in real time their statistics and throwing out alerts when they are running bad or down.

Link to comment
Share on other sites
Theresa Tennyson

Theresa Tennyson

Posted
Posted

Nimh20 Vandeverre wrote:

More Details? Maestro--this is across the board when it comes to Mainland.  My first post with Stats was right after Help Island Pubic was restarted.  My second post stats were from Korea1 but it doesn't matter what Mainland sim. your on, it seems like this is happening on most I visit.  I mean with these conditions anywhere on the Grid no matter where you are, how do you ever expect Valve to let you guys on Steam?  I can't believe you all don't have some kind of simulator monitoring system of the about 27,000 sims showing in real time their statistics and throwing out alerts when they are running bad or down.

Sounds like you're describing BUG-355, a problem that's been around for months. Occasionally sims with a moderate number of avatars will temporarily have their physics frame rates plummet and net time skyrocket, usually when there is a steady stream of avatars arriving. It's happened at server usergroup meetings before. When the population stabilizes the framerate goes back up. 

It only happens in certain areas - it looks like some avatars trigger the problem and others don't. When it hit the server user group meetings it was when a group of "fashionistas" showed up en-masse to discuss a problem they were having. Other sims with more, but different, avatars seem basically unaffected.

Link to comment
Share on other sites
Nimh20 Vandeverre

Nimh20 Vandeverre

Posted
Posted

Hi Maestro,


This is exactlly what I'm talking about. I believe Theresa is describing what I am seeing watching sim stats like a hawk-lol.  Hopefully there is something then can adjusted or changed to fix this bug.

 

Theresa Tennyson wrote:

Sounds like you're describing BUG-355, a problem that's been around for months. Occasionally sims with a moderate number of avatars will temporarily have their physics frame rates plummet and net time skyrocket, usually when there is a steady stream of avatars arriving. It's happened at server usergroup meetings before. When the population stabilizes the framerate goes back up. 

It only happens in certain areas - it looks like some avatars trigger the problem and others don't. When it hit the server user group meetings it was when a group of "fashionistas" showed up en-masse to discuss a problem they were having. Other sims with more, but different, avatars seem basically unaffected.

 

Link to comment
Share on other sites
Yuriko Nishi

Yuriko Nishi

Posted
Posted

hi maestro,

 

i know you usualy can´t do much about such stuff, gonna tell you anyway.

 

imo the new "interest list" stuff is a good idea, but how it´s implemented atm it is feeling more like buggy servers or bad connection.

 

a few examples:

 

1. you stand around and chat with a few friends and wonder why they´re talking to this person you can´t see, you move your cam a bit and BAM! that person is standing right in front of you for 5 minutes already, the server just didnt tell your client yet.

 

2. you arrive somewhere and only see half of the building until you either cam faaaaaaar out and then back in or slide your object detail slider down to minimum and then back up to what you want.

 

3.
 this is a horrible one that has potential to break racing in sl even more:

you drive around a corner (your camera view shifts to a part of the sim you didnt see before) and you see 10 cars parking on the track where you´re supposed to go! after about 1 sec the cars poof. this happens EVERY LAP.

connected to that, when you lock your camera on a vehicle going around, you see the car stuck in place for 2-3 seconds EVERY LAP in the same position of the sim.

another one: you face away from the track (track is in your avatars back) and you hear the engine sound looping in one place, it doesnt move anymore until you turn around, then you see the car teleporting around, not driving around, not enough position updates from the server.

 

4. you see avatars faces stuck in animations, often multiple at once. smile, cry, laugh, angry, all at once, looks like a horror movie, the server didnt tell your client the animation is finished.

 

5. you look at the minimap after a teleport, and half of the avatars sit in the sim corner at coordinates 0.0.0 until they stand up or walk around.

 

6. attachments that got turned invisible by a script are still visible, because the server didnt tell you what the script did, because it wasnt in your field of view, while the script did it...

 

need more? ^^

 

is anyone adressing these issues?

Link to comment
Share on other sites
Ayesha Askham

Ayesha Askham

Posted
Posted

Yuriko

Apart from the now well-known issue with object rendering errors (missing prims), the only other issue I have encountered, and I've seen this 3 times now, is the invisible avatar.

That one is most unsettling, since it is even more extreme than the blocking effect.  The avatar is simply not rendered until camera position is altered substantially and a new field is to be drawn.

This is related to the missing prim issue in that some vital information provided to the viewer by the sim is no longer given, or is only given at certain points.  The viewer has all the data, it just isn't told to render the avatar.

I only encountered the non-rendered avatar after the RC rolls this week, so far I have not seen it on Main Server sims.

While I do not doubt that Andrew and Simon know about this now, we do not know how or when this issue will be addressed, other than that it will be a client-side modification. :smileyfrustrated:

Link to comment
Share on other sites
Yuriko Nishi

Yuriko Nishi

Posted
Posted

the issues i described in 3 and 4 are experienced by everyone who comes to our racing events. if you dont do racing you of course cant experience such "bugs" ;)

 

anyway, i doubt any of that will ever get fixed, i spent hours with writing jiras and demonstrating other vehicle related  issues to lindens on the beta grid. they aknowledged the problems, and rolled the broken code anyway, and it stayed till now (6 month or so)

 

just wanted to mention them anyway . maybe there will be a miracle and anyone actually cares about us...

 

i think the racing community is just not on LL´s "interest list" ^^

Link to comment
Share on other sites
Qie Niangao

Qie Niangao

Posted
Posted

Yuriko Nishi wrote:

the issues i described in 3 and 4 are experienced by everyone who comes to our racing events. if you dont do racing you of course cant experience such "bugs"
;)

Then I must not understand #4. Why would missed animation updates be unique to racing?

It would seem that everyone should also be able to see #5 & #6. Given my in-world activities, I'm especially surprised not to have encountered #5: I nearly always have the mini-map up and am constantly hopping from sim to sim full of avatars, and I've never noticed them clustered at (0, 0).  Is there more to that one?

[ETA: #3 seems a big challenge, indeed. I'm finding it difficult to imagine a sensible way to fix this while retaining any of the wins from the new interestlist logic. In some meeting, Andrew mentioned the possibility of a "keyhole"-shaped area of interest consisting of a sphere around the cam location in addition to the current sweep in the cam's direction, but it seems that sphere would need to be sim-scale to handle the cars-around-the-corner problem. I'm wondering if dynamic objects (not necessarily physical) could somehow "register" as being of interest sim-wide -- but can't imagine how to keep that from being an instant target of viewer-crashing griefers.]

Link to comment
Share on other sites
Yuriko Nishi

Yuriko Nishi

Posted
Posted

did some research on the enterwebz.

 

4. might be connected to fast moving avatars while triggering animations.

speculating: driver triggers animation while in your view, drives out of your view and the animation never ends because out of your cam. seen that in the past a couple of times, now it happens all the time. this is no biggie, just looks silly.

 

5 might be connected to : Fix for “Agent appears in incorrect position to other agents after being moved by a sim teleporter” (BUG-1795). but i can´t make sure because i cant view jira anymore (why btw)

 

6. might be about BUG-1779 but again i can´t confirm, just found a description of that bug on a blog with a link to the jira which i cant view

this is a quote from that blog:

"The lack of updates being received from objects outside of the camera’s field-of-view have resulted in other issues as well, which are also fixed in the Magnum release.

For example, if you are looking at a cube which is changing colour between red and blue every few seconds, and turn your camera so that it is outside your field of view and wait for it to turn blue before camming back, the cube will briefly still appear to be red in your view before suddenly changing to blue. This is due to the fact that when the cube is outside your camera’s field of view, the updates associated with it are no longer sent to your viewer until such time as it re-enters your field of view.

With the updates deployed to Magnum, updates for objects outside your field of view are once again sent to the viewer – but at  a very low rate (around one update every 5 seconds). While this uses a little more bandwidth then sending no updates at all, it also means that objects in a state of change behave more predictably when camming away from and back to them (BUG-1779)."

 

maybe that particular ones are already fixed on rc

 

Link to comment
Share on other sites
Qie Niangao

Qie Niangao

Posted
Posted

5 might be connected to : Fix for “Agent appears in incorrect position to other agents after being moved by a sim teleporter” (
). but i can´t make sure because i cant view jira anymore (why btw)

The release notes indicate that 1795 was fixed in the most recent release to main server channel (apparently 2 April), for whatever that's worth. [EDIT: Oh, duh, I see you were actually quoting those release notes.]

(And yeah, this insanity of hiding jiras from public view has lasted much longer than I'd expected. It's increasingly a pain in the butt even for the few of us who can see them -- actually, especially for those who can -- and was an astonishingly foolish idea from the start. You'll never find a Linden who will defend it, but you'll never find one permitted to discuss it, either.)

Link to comment
Share on other sites
Levio Serenity

Levio Serenity

Posted
Posted

Despite the release notes for this version saying the interest list bug(s) were addressed, I still don't see any improvements in animated objects being rendered correctly.  I'm referring specifically to BUG-1779 (although not filed be me)  As far as I'm aware, this is affecting 100% of every animated object in SL.  I'm sure others along with me have spent an awful lot of time and energy scripting/animating things, and it's a real shame that SL looks so bad now. I am embarrassed to show our Meeroos to outside parties when the prims break apart over the place. This bug has been out there for months now, and should never have made it to the live grid, please address ASAP. -Tiger

 

Link to comment
Share on other sites
Curious Hazelnut

Curious Hazelnut

Posted
Posted

Eddie Mathieson wrote:

For the amount of money we've all invested in Second Life over the years, would be nice if it would run at least 50% the quality of all other online gaming out there!

I find Second Life to be more reliable than a recent release from EA and I am happier with the direction Linden Lab is taking Second Life than with the rumors around the next generation XBox from Microsoft.

Link to comment
Share on other sites
Andrew Linden

Andrew Linden

Posted
Posted

I logged into to Ruiz/159/80/1502 as per the instructions in BUG-1779 and finally noticed what you're Levio is talking about.  For me it was rather hard to see becuase most of the meeroos would snap together pretty fast, but I dropped my viewer bandwidth settings down to 300kbps (from 1Mbps) and could reproduce it much easier.

Each visible part of a meeroo is a child-prim.  The root is an invisible prim that typically doesn't actually move when the meeroo sits down or stands up -- each child prim moves to a new local transform to produce the effect of motion to a new stance.  The problem is that each prim is being scheduled independently of the others so they arrive out of order, and may be incorrectly sorted so that some for one meeroo will arrive much later than others for the same animal.

Hrm... I will ponder how to fix this.

Link to comment
Share on other sites
You are about to reply to a thread that has been inactive for 4019 days.

Please take a moment to consider if this thread is worth bumping.

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
 Share
Go to topic listing
×
×
  • Create New...