Would removing voice entirely from the grid actually be an overall upgrade to SL now?

Stop! You're taking away their scapegoat!!

btw voice also leads to child labor.

I voted for removal, but not because I think Voice imposes an unbearable processing load on the system. I think it'd be an "overall upgrade" in the sense of IMPROVING communication.

I've spent the last two evenings at Public Orientation Island, handing out Lucrezia's Newbie Kit. The audio chaos is incredible. People playing loud and annoying music with their gestures, or on the voice channel, voicers muttering to other voicers about how lame someone else is (as if the other person can't hear them), impossible to tell who's speaking... it's quite awful.

Please, LL, PLEASE bring peace and quiet back to SL! Offer Voice as an option for things like corporate meetings, or education seminars, or machinima producers if you must, but take it off the general grid. It's cacophony!

And how exactly does removing voice stop the griefers from using annoying music with their gestures? This is something thats been happening for years now its not new.

As for voice its simple people - it's called mute. Just open up the active speakers find the loudest person (which isn't hard) and mute them.

As for the load, the real load on the main grid is using distance to put how loud it should be. And that really isn't going to be that much overhead. Voice Servers aren't a huge factor. If you say it is then by that standard Audio and video streams would have an effect on sim performance because they're tied into SL yet using different systems.

Ahem: Beta testing software in parts of the Main grid? Sounds to me like it would cause problems. Quite frankly it would not matter if it was stable for YEARS or if the project caused an issue right off the bat: The issue is this project. Full Stop on that one.

You do not test out new code on a server system in the manner they are doing it ... This WILL cause problems. said problems may not manifest at once, but they are still there.

As for the number of users .... That is academic. The point of this is that they altered something in the server code somewhere and it is now causing issues.

since Voice has been a part of the server code since the first version of the Voice First Look client - it isn't Voice that is the culprit.

Sigh.

The amount of users is not, in fact, academic. Actually, it's quite possibly at the heart of the issue - Voice has never been adequately tested "en masse", if you will, until it got deployed in the main client. I don't think it's a stretch to point out that exactly at the same time that happened, the grid became immensely unstable, while the protocol switch - that "something in the server code somewhere" that they altered - happened some time back, without causing instability issues.

Now, if you are trying to make the case that the ability to run two fairly different clients on the same grid is inherently going to caused stability issues, I would agree with you. But to insist that voice has nothing to do with it - circumstantial evidence to the contrary - without an official statement one way or the other from LL is I think being a bit presumptive.

You are correct sir! lol

I hear is also causes obesity!

Don't forget that you can get a virus with it as well!

Don't forget that it will be the decline of Western Civilization as we know it!

the moment voice was placed into the Server code and the First Look client released, there should have been a rather large flood of people downloading the Client to test it. The fault there is on the user base.

currently there is NO concrete evidence that Voice is the source of the issues.

Simple Logic and Common Sense however says that mucking about with the system to test software on the main grid is going to cause severe issues - and no, I'm not talking about the client software either. I'm talking about the server software.

again, if voice were the problem, this would have occurred a while back. It didn't. End of story on that one.

what happened? LL more than likely started testing new code recently.

Perhaps there would have been such a flood had many users actually been interested in voice in the first place. Saying that the user base is at fault for not bothering to test a feature the apparently overwhelming majority never wanted is a little unreasonable. We as users are not being paid to mess around with features we don't care about and don't want, despite our lack of interest in them.



No, but given that the current issues and voice going live happened at exactly the same time, there's at least very strong circumstantial evidence.



... whereas, by contrast, there's not even circumstantial evidence of this, let alone concrete. It's just pure speculation.

Maybe the real culprit is cosmic rays!

And the times that we saw pretty much the exact same kinds of grid problems over the last year, before voice, were caused by.. ?

The forum citizens have spoken. Wow, I expected it to be a bit more even.

A temporal glitch that gave a us a pre-lagged echo of Voice one year before the fact.

It's that evil, that Voice.

Bullshiat. It's NEVER the job of the clients to make sure the service provider adequately bug tests something. The fact that we do it at all is a bonus.

Besides, as Alex says... perhaps there was no flood of users because a "flood" of users didn't want voice, or at least didn't care enough about it to subject themselves to buggy beta software to test it. Or were perfectly willing to test it, but were put off by LL repeatedly ignoring specific complaints about the client, such as I was.




I'm pretty sure I said it was circumstantial evidence. But it's convincing circumstantial evidence, that makes logical sense.



I'm not disputing this, and never have.



Bloody hell man. Have you EVER been present for a feature rollout before? The transition from beta to live is almost ALWAYS problematic because of the influx of new users. Hang around an MMORPG if you don't believe me - name ONE MMORPG that has had a launch go off without a hitch, despite the fact that the game was working fine the day before in beta on exactly the same code. Furtheremore, this is not the first time LL has done exactly this - had a feature that worked fine in beta and then collapsed when they transitioned it to the live code. They have a track record here.



And this is PURE speculation, whereas at least my theory has some pretty compelling circumstantial evidence on it's side.

That's funny. However, I agree with the person you commented on...

There have been many instances of glitches, inventory problems, login errors, lag, etc...and some of the times, people have claimed SL is borked...

This was all before voice.

So, why don't we just get over the hump...let them smooth out the rough edges.

Many things.

The problems we are experiencing aren't "voice problems". It's just LL's usual overload and lack of proper testing causing havoc. Voice may have triggered it this time, but there have been plenty of other things in the past to do so as well.

They will smooth things out, or they won't, entirely without our input or lack thereof. It's pretty well established that they don't devote any measurable time to monitoring what goes on here, so in the meantime people vent. It makes them feel better and doesn't hurt anything.

I voted no for semantic reasons.

"upgrade" and "downgrade" in my book are not synonymous with 'improvement'.

Would SL be improved, overall, if SL was downgraded to remove Voice? Too vague to say.

Would ripping voice out make the world more stable? Doubtful.

Would the grid run better? There would likely be no change.

Would the client software run better? Sure.

Would lag between customer and the grid decrease? Sure, with less competition for bandwidth on the customer's connection.

Would people who can't type well enjoy SL more without voice? Probably not.

So many factors to measure and weight in for an 'overall improvement' score.

With or without would make no difference to me currently, so, again, 'no'.

I suggested in another thread that voice in and of itself is not the cause of every problem from a coding standpoint, and I still believe this is true. I also don't believe that the problem becomes server side until what I outline below happens to enough of the clients.

The fact that many people are claiming that use of Skype and others are not degrading thier performance to the same degree as the built in voice seems to support (at least to me) the fact that the rendered sound being used by the inline service is the culprit. From what I have been able to understand from looking at the whitepaper from Diamond (the company that provides the services), there are two separate components that are especially critical in the process - one is the 3D rendering/spatial positioning engine, and the other is the stream management engine. I believe that one or both of these services is being handled on the client side embedded in the voice executable that runs separately.

If this is the case, then the client is now being asked to not only render massive amounts of 3D graphical data, but is also having to render 3D positional sound as well. The resulting bottleneck is similar to the one that causes the makers of FPS games to recommend turning off EAX and hardware 3D sound if performance is lagging badly.

When these clients use up their resources they begin sending malformed packets into the network, causing collisions and increasing packet resends and network traffic. As the network gets dirtier, it also gets less efficient, and slower. The effect on any given computer depends on how close to pegged its CPU remains and whether or not it can keep its cache and buffers clear enough to send packets without interruption.

I could be wrong about all of this, and it doesn't totally explain why people not actively using voice still experience these problems, but I do believe that voice is just sending so much data to already overworked clients that they can't communicate on the network anymore.

Incorrect.

From Dictionary.com (direct copy and paste):

up·grade /n. ˈʌpˌgreɪd; adj., adv. ˈʌpˈgreɪd; v. ʌpˈgreɪd, ˈʌpˌgreɪd/ Pronunciation Key - Show Spelled Pronunciation[n. uhp-greyd; adj., adv. uhp-greyd; v. uhp-greyd, uhp-greyd] Pronunciation Key - Show IPA Pronunciation noun, adjective, adverb, verb, -grad·ed, -grad·ing.
–noun
1. an incline going up in the direction of movement.
2. an increase or improvement: an upgrade in the year's profit forecast.
3. a new version, improved model, etc.: The company is offering an upgrade of its sports sedan.
4. an increase or improvement in one's service, accommodations, privileges, or the like: If the ship isn't full we'll receive an upgrade to a deluxe stateroom.
5. something, as a piece of equipment, that serves to improve or enhance: a full range of upgrades available for your computer.
–adjective
6. uphill; of, pertaining to, on, or along an upgrade.
–adverb
7. up a slope.
–verb (used with object)
8. to promote to a higher grade or rank: He has been upgraded to senior vice president.
9. to improve or enhance the quality or value of: to upgrade property by landscaping it.
–verb (used without object)
10. to improve the quality, value, effectiveness, or performance of something: Buy this basic computer and upgrade whenever you're ready.

Note especially definitions 2, 4, 9 and 10.

Good opportunity to correct a misconception.

1. There is no excuse for not testing a feature that has been implemented into the main grid code and is now able to be tested there by the users. None whatsoever. The Lindens did their testing in an Alpha grid, some residents did testing on the Beta grid.

Guess what happens when it hits First Look? That's right - You're supposed to do as suggested and use the test client as often as possible. Not wanting the feature is irrelevant.

2. Your 'evidence' is not convincing whatsoever. all you have done is string together a few Blog posts.

3. I'm well aware of Linden Lab's track record. The alterations to the message system came about before Voice - the project allows for far greater issues to occur over time.

4. do you REALLY expect to see a Blog post by a Linden telling us when they are testing new software using this project of theirs? They did it so they wouldn't have to tell us!

5. Until someone comes up with concrete evidence that Voice is to blame for this - voice is not the problem. The alterations to the server to allow their little testing project to function are. Full Stop.

6. Until the entire user base has been polled concerning the Voice feature, no one can state if the majority of residents wanted Voice. This forum, other forums, and various Blogs do NOT show the opinions of the entire user base. Full Stop.

[ ] Yes.
[ ] No.
[X] No idea.
[X]

Did I miss pie?

I saved you a slice..



..but hush, don't tell anybody!

Bzzzt, wrong. A paying customer is not "supposed to do as suggested" and used buggy software to spare them the expense of doing real QA testing. Sorry, that wouldn't fly anywhere except in the MMO world, and not even in most of those.



I haven't done anything relating to blog posts at all, you must be thinking of someone else.



But what CAUSED these greater issues? Could it be, oh... a few thousand new voice users suddenly taxing the systems that were never tested with that stress load?



Wow, that's right up there with "the proof of God is that you can't prove he doesn't exist".



No, not "full stop", no matter how much you like your little phrase. The simple fact is you have not one iota of evidence more than I do that your hypothesis is correct. I at least have circumstantial evidence, you don't even have that. But regardless of who is right or wrong, the truth of the matter isn't altered by how much "concrete evidence" we can find.



Seriously? You couldn't possibly sound more pretentious right now.

This forum consists of speculation. Yours and mine. And all your silly little "full stops" in the world don't prevent me from posting my speculations, so you might as well just stop trying. The whole user base has never been polled, and never will be - but that doesn't mean I can't make guesses, using what evidence we have.