Last week’s unexpected change to the Second Life third-party viewer policies has generated a lot of questions. As promised, I contacted Linden Lab with a list of the most commonly-asked questions and have collected the responses.
One interesting departure from normal Linden Lab communication policies is that Oz Linden’s statements on or about the new policy, verbal or otherwise, are to be interpreted as being official policy statements on behalf of Linden Lab – a privilege rarely granted; in fact, I think it is unique.
Accordingly, most of the responses from the Lab are from Oz.
Why was the policy update posted to the forum section of the community site instead of to the blog? Being a terms-of-service change, it strikes me that it should be made prominent, as it is a part of an agreement that all users must comply with individually.
We try to use the official Second Life blog for announcements when they impact all Second Life users. Because this change affects developers of third party Viewers, rather than all users of Second Life, it was announced and discussed directly with developers at the Open Development User Group, and then we posted about the changes in the forum so that the link to the post explaining the changes could be easily shared with developers.
This indirectly impacts all users, one way or another, as it is directly related to viewers that are commonly used. Why was it not announced as widely, as (for example) “Flickr Pic of the day”?
In this case, because the updates were to the guidelines specifically for developers, we tried to target our communication about what changes we made and why to them. We understand that the changes will also be of interest to the users of those Viewers, and we appreciate the opportunity to help explain the changes through exchanges like this, but our priority at the time of announcing the changes was explaining them to developers themselves.
When does the new policy come into effect?
The policy went into effect on February 24, 2012.
Which features and viewers are affected by the policy change?
There are a couple of minor changes like eliminating Viewer tags and online status, which we’ve already discussed with Viewer developers.
Will users of those viewers and the viewer developers be notified, or will the affected viewers be locked out of Second Life as soon as possible after the policy comes into effect? If there’s notification before viewers are locked out, how long will the grace period be?
We haven’t made any changes to how the policy is enforced, and at this point there are no enforcement actions planned related to these changes.
So, most third-party viewers are in violation of the policy at the time that it went into force, but Linden Lab is not planning to take any action about current viewers that are in breach of the policy?
Viewers that had features in violation of the updated policy (for example, “true online presence”) are changing them, and we don’t expect to have to take action to enforce the updates at this time.
We made the updates not in response to a specific problem right now, but to prevent the problems that could arise in the future if we don’t protect a common and safe experience for all users of Second Life. We wanted to be clearer that those sorts of features should be developed in cooperation with Linden Lab and also strengthen the protections for all users’ security and privacy.
There’s a list of “exceptions” to the policy (particularly to item 2.k). Where and when will that list of exceptions be published/announced, or will it remain verbal and anecdotal?
We don’t have plans to publish a list of exceptions to the policy. We will enforce the policy as we usually do, and if problems arise, we will deal directly with the developers involved to try to resolve those issues. Again, though, at this point there are no enforcement actions planned related to these changes.
Does this mean that text-based viewers, mobile-device viewers, and those that are customised for the disabled are forbidden when the policy comes into effect? If not, well, there doesn’t seem to be anything in the policy that would *permit* them to still be used.
Absolutely not. There is nothing in the new rules that in any way alters support for accessibility, or any other changes that are purely presentation and user interface.
The text of the policy appears to indicate otherwise. Doesn’t the text of the policy automatically take precedence over any representations about the policy? As a user, I must read and agree to this document, as I must read and agree to more than 30,000 words of other policy documents and materials.
Obviously, I expect to know exactly to what I – and others – are agreeing, and have a natural expectation for a consistent, transparent process of enforcement by Linden Lab as a part of its own obligations.
It is not always possible to write a rule or a policy in a way that is at once simple, unambiguous, and covers all future possibilities. Some judgement is required, and we believe that the developers of third party viewers are quite capable both of making good judgements and of communicating with us about what they would like to do to contribute to the Second Life experience.
If as a user you are not confident that a particular third party viewer is compliant with the policy for any reason, you should not use it. If developers are unsure whether a feature would be in violation of the policy, we encourage them to discuss it with Oz Linden.
What procedure will be in place for TPV developers who want to develop a new viewer feature? They’ll clearly need to get approval from Linden Lab before they start (since if they don’t get it later, it will be a waste of effort). Since they cannot release or change almost any feature before Linden Lab does, I suppose that means that any approval they get will have to be in writing. Hardly anyone seems to be clear on what is and isn’t permitted at this time.
The vast majority of TPV features are differences in the user interface or communications that don’t create any incompatibility with other Viewers, and as such are not affected by the new rules. If Viewer developers would like clarification on specific features, they should contact Oz Linden.
The Snowstorm proposal review process is in place for Viewer developers who have suggestions that do affect the shared experience.
Speaking of features, how does this affect the mesh parametric deformer, specifically?
The proposed mesh deformer is a good example of the kind of feature that impacts the shared experience of Second Life and is subject to the new section of the policy (2.k). It’s something that affects the shared experience, so it’s a feature that should be developed by working with Linden Lab so that it can be incorporated it into Second Life for all users. We’re very supportive of the mesh deformer proposal and are working with Qarl to incorporate it into the official Viewer as soon as it’s done. As such, we don’t expect this to be a problem.
We would prefer that testing of the new feature be done primarily on that common code base to minimize the time until we can accept it, but don’t object to TPVs incorporating test versions with it. We caution content creators and users that until work on it is complete, any objects created using test versions may not remain compatible with the final integrated feature.
Exactly how is the policy enforced?
Enforcement is described in section 8 of the policy itself. Our reason for updating the policy was not to go after a Third Party Viewer, and if problems arise with violations of the policy, we’d rather work with the developers to resolve it. Again, if developers are unsure whether a feature would be in violation of the policy, we encourage them to discuss it with Oz Linden.
The policy document itself gives me some personal qualms, as someone who has wrangled such things before from both ends.
“shared experience” – a key term in the new revision – is not defined in the agreement.
Section 8 of the TPV policy – the enforcement section – is probably my least-favourite. I mentally think of it as the “we just make some stuff up” clause, and it taints the whole document with a particular air of unpredictability.
Given two parties to the agreement, the scope, behaviour and obligations of one party (Linden Lab) within the terms of the agreement cannot be determined or predicted by the agreement.
A decade ago, I’d have laughed if anyone had shown me an agreement like that, as it opens up a substantial liability window for the Lab in my opinion (which you should not infer to be any form of legal advice).
Final thoughts
Linden Lab has made little apparent effort to “sell” the policy or policy changes, to explain why they exist, who benefits from them, and why they might be a good thing. Those are questions that I still don’t have the answers to, and that I cannot reasonably begin to speculate on.
A well-documented baseline feature-set, that content-creators could refer to with confidence, would make a lot of sense, and probably be of broad utility and benefit, but there’s nothing like that described here.
LL doesn’t need to sell it. They can – and did – ram it down our throats, and don’t care if we choke on it.
As for your objection to Section 8, all I can do us cue up Voltaire’s “USS Make Shit Up”.
@Tonya That’s the song I had stuck in my head at the time – and every time I read that policy document. I was thinking of Star Trek’s “Particle of the month” club.
“‘Cause if we find we’re in a bind
We’re totally screwed but nevermind
We’ll pull something out of our behinds,
We just make some shit up”
http://en.wikipedia.org/wiki/Standard_form_contract#Contracts_of_adhesion
So a feature that changes how users view the virtual world (mesh deformers) would violate 2.k, but a feature that changes how users view the virtual world (text-only viewers) would not violate 2.k? I’m glad they cleared that up.
I’d also like to point out that in making this policy take effect the same day they announced it, Linden Lab broke their own terms of service, which state:
and
So, congratulations to the Lab’s legal department on another bang-up job.
Thanks for this article. This cleared a lot up for me as an “end user” of a couple of viewers. I must say this is one of “those actions” by Linden Lab that are both good and bad. I can see where they are coming from with these changes. But also view them as extremely … lets say …. pushy.
I suspect there are businesses out there that could pull off this sort of explanation of a very vague policy and vaguer enforcement of “no,no don’t worry about it – you can trust us to only do what makes sense”. Unfortunately the Lab’s historical performance with their customers has not earned them the level of trust required to just relax and not worry about this sort of half-assed policy change. The capricious and scattershot enforcement of vague rules and policies that varies from one employee to the next has pretty much been a hallmark of the Lab’s SOP for quite a few years now. Poor internal communication and even worse external communication both create and compound the problems, as evidenced by the immediacy of the timing (if it is not addressing a current crisis why not give folks time to read it before announcing a policy change to TOS) and the lack of even internal consistency with their own policies.
Does the Lab even HAVE a legal department now? If they do, they might consider actually utilizing their services. Just sayin’
Okay, last blog I played devil’s advocate, now I’m going back no normal role. ; )
“We try to use the official Second Life blog for announcements when they impact all Second Life users. Because this change affects developers of third party Viewers, rather than all users of Second Life…”
Seriously? Who are they trying to kid?
“our priority at the time of announcing the changes was explaining them to developers themselves”
Ohhhh we see. So their priority was the TP devs, and not the CUSTOMERS WHO ARE PAYING THEIR BILLS and who will suddenly find their viewers not allowed to log into Second Life? (at least, that’s how it sounded.)
“Viewers that had features in violation of the updated policy (for example, “true online presence”) are changing them, and we don’t expect to have to take action to enforce the updates at this time.”
Well, that’s good to hear, considering the policy “went into effect on Feb 24, 2012″.
We made the updates not in response to a specific problem right now, but to prevent the problems that could arise in the future
Okay, I can understand and agree with that. So customers… you can now de-stress and comfortably recover from the coronaries. False alarm. Oops!
“Doesn’t the text of the policy automatically take precedence over any representations about the policy?”
Very, very good question.
“It is not always possible to write a rule or a policy in a way that is at once simple, unambiguous, and covers all future possibilities”
Or in the case of Linden Lab… it’s apparently never possible. ; )
I mentally think of it as the “we just make some stuff up” clause
Yeah, strikes a lot of people that way, it seems. I do realize in this case LL had good intentions (which would seem unusual, but in this case, I don’t think they meant to cause coronaries, although they did). It’s that ol’ LL talent of not being able to see beyond 24 hours and predict consequences.
“A decade ago, I’d have laughed if anyone had shown me an agreement like that, as it opens up a substantial liability window for the Lab in my opinion”
That was the first thought that struck me. “Is this even legally enforceable?” Of course then we get down to that old sack of “probably not, but that doesn’t mean much when we’re banned from the grid” (or in this case, our TPV won’t log in). Good luck fighting the King.
In all fairness, I’m glad LL tried to explain this better, I appreciate their “reassurances” they don’t plan to enforce the policy at this time, and that they will work with the TPV developers. But as you state Tateru, that’s not what’s written on the paper. And since at last count Linden Lab had a zero trust factor with their customers… they’ll understand if people don’t take them at their word. I do understand that this is a hole in the dam they needed to plug, and that frankly I’d have plugged a long, long time ago (well, to be honest, I’d have never made the viewer open source to start with. Call me paranoid). But I actually can understand that there are things they do not want on their grid (such as revealing user information LL doesn’t want revealed), or putting in new toys that could potentially detrimentally alter the experience of SL (LL does enough damage in that area already). ; )
Not being snide… nor even bashing LL on this one (preferred the satire route this time). Just pointing out a few obvious observations that honestly, someone at LL should have seen before they released that announcement. What they need is a Watchdog Linden to review just about everything they do and say, “Hey guys… this isn’t gonna fly very well.” I bet they wish they’d had a Watchdog on the OpenSpace fiasco and Viewer 2.
So in summary, I’m not condemning LL on this one. I understand their reasoning behind it. But like so very, very many things… this seems possibly a reasonable step that might have been more wisely implemented.
(subscribing to the comments)
I have now looked at Oz Linden’s user page on the SL Wiki, which links to his LinkedIn user profile. Based on his previous career, he’s a smart guy. But he now has the privilege of being the lead customer spokesman on this issue, and I suspect he is clamped firmly in the jaws of the Peter Principle.
His past experience of dealing with customers may well have involved corporation-to-corporation relationships involved with technical issues and people much like him.
Second Life is different. He is dealing with a highly diverse community of computer users, not exclusively with programmers and those who manage them. Not all of them are native English speakers. Many of them are on the outside of any corporate structure of obligation: “My way or the highway” doesn’t work.
It doesn’t, I think, help that he appears to be based in Massachusetts, on the other side of the continent from head office. He may not be a part of a “water-cooler network” which, even in a well-organised company, can resolve issues before anyone gets embarrassed. It takes work to maintain an equivalent community across the internet. On that aspect, Linden Labs ought to know what they’re doing, but…
(There are things you can say informally which nobody else can ever be sure of. Email and almost everything else which might be used on the internet can get logged (see Wikileaks for examples). Oz may be used to using voice-over-IP for that reason, which makes the in-world meeting with the Developers somewhat interesting.)
Quote: There are a couple of minor changes like eliminating viewer tags and online status, which we’ve already discussed with viewer developers.
MINOR?!?! Eliminating online status will break MANY scripted objects, i.e., staff boards, tip jars, marketplace delivery mail systems, rlv items, etc. All of which use online status in one form or another. LL needs to consider either an opt-in or opt-out remedy for this ‘privacy’ issue instead of a blanket change without considering the consequences. This is way too much like the Last Name issue where the response to our asking for the ability to change our last names wax to throw out Last names entirely and came up with Display Names which did nothing to solve the core issue. Its throwing out the baby with its bath water.
Jacek: At first, I didn’t understand how the deformer could be considered part of the “shared experience” either since it’s a viewer-only feature… but then someone explained it to me.
If I wear a mesh object designed to work with the deformer, and turn the deformer on in my viewer, it’ll look great — to me. But to people using a viewer without the deformer, it won’t look right at all. Therefore, the “shared experience” for deformerless users would indeed be degraded — much like how it used to be with Emerald’s “fake” extra attachment points.
“So a feature that changes how users view the virtual world (mesh deformers) would violate 2.k, but a feature that changes how users view the virtual world (text-only viewers) would not violate 2.k? I’m glad they cleared that up.”
It makes perfect sense. Clothes designed to work with the deformer will look broken on all viewers that don’t have the deformer. So although the deformer is a client-side feature, it will have an impact on server-side data, just like Emerald’s extra attachment points did.
Text-only viewers don’t have that effect, so they’re OK.
@wayfinder
Agree about the TOS changes. I suspect it would make for an interesting case if I compiled a non-compliant viewer now as I could just say I never officially saw or agreed to the new TOS.
@wolf
Regarding VoIP: note that the “open source” product Oz worked around in the past was a VoIP/SIP server.
But what caught my eye was the history of that product: developed by an open-source start-up, which (after five years) foundered and was bought by Nortel, which was in turn bought by Avaya a year later. Six months after that, Oz was at the Lab.
These kinds of experiences profoundly shape attitudes and habits.
“There is little value to be extracted out of the open source community [with which] Pingtel is associated…” –trade press reaction to the acquisition by Avaya
http://news.cnet.com/8301-13505_3-10018919-16.html
Linden Lab new rules !
If you can’t define a term, don’t expect anyone else to magically know what you are thinking either, so till you or someone can, don’t use it as the basis for discussion, let alone any kind of policy.
Don’t claim to be fixing privacy by breaking the only dependable point of presence, fix the point of presence and make it user privacy aware. We have waited better part of a decade, we don’t mind if it takes a little time to do correctly.
If you change something that impacts 50% of your user base, man up and tell them in person. Don’t just make the change and let others handle it. This is called a public relations disaster.
Hire a customer liaison with enough authority to catch things like this before they happen, perhaps even elected resident community representatives to help connect with your customers. The more you engage, the less likely you are to be surprised or mauled.
Fireice: “MINOR?!?! Eliminating online status will break MANY scripted objects, i.e., staff boards, tip jars, marketplace delivery mail systems, rlv items, etc.”
In fairness, removing display of online status from Viewers isn’t likely to affect scripted devices.
Oh wait, this is LL. “Born to bork.” Okay, now I’m worried. ;D
@Wayfinder It is being separately disabled in the LSL function call as well, unless the person being queried owns the script (or is the creator, I understand)
@Tateru: I haven’t considered all the ramifications of such, but right off hand I’d say limiting LsL function to the owner of the device is standard prevention of invasion of privacy.
In other words, if the owner sets out an “I’m Online” device it works. If someone else does so to see if the person they’re stalking is online… it doesn’t.
[...] The backpedals begin… kinda… sorta… … or you can just hire people you know and trust. Ghostbusters! Writer's Dash: Anoint Melissa on TPV [...]