DISQUS

Dusan Writer's Metaverse

Second Life

Australia

*typing and no, it wasn't deliberate.

That's a really interesting idea. I wonder how much of my frustration with SL is a matter of reaching for system functions I'm used to in every other app on my machine. I've got a disability that makes ting without a spell checker a serious challenge and I doubt my situation is unusual.

We will soon need an RSS feed for Linden Lab's ever-growing tangle of RSS feeds. Personally, I prefer podcasts to be podcasts and blogs to be blogs.

Second Life is severely under-documented and it would be great to escape the limits of oral tradition when I'm building)

On the state blog of Second Politburo (trademark pending) many contributors accused Second Politburo of speaking lawyerese. Nothing could be further from the truth. The guidelines reek of amateur bunglers playing cut-and-paste with boilerplate and some would say that's about par for the course for the company that may not be named.

Note this is a somewhat disparaging response so I am, in theory, prevented from using the name of the company that may not be named for fear of breaching their alleged intellectual property.

The measure of the silliness of these guidelines, and speaking only in terms of US domestic law, can be found by considering what instruments the guidelines purport to suspend.

1. The First Amendment to the Constitution of the United States. Apparently the press are no longer free to use the name of the company that must not be named if their text criticises said company.

2. The law of conflict of laws. Choice of law and choice of forum clauses are quite common in contracts. They derive their validity from the law of conflict of laws. The guidelines of the company that may not be named purport to suspend the principles of that law. It is unclear how you can suspend a body of law by an instrument that owes its validity to that law.

3. Many of the alleged intellectual properties of the company that may not be named have been in common use for several years. Once you let the IP cat out of the bag you do not have the right to recapture it years later.

4. Intellectual property is part of the law of property which governs claims made against the whole world. The law of contract governs claims agreed to by the parties to the contract. A court has already ruled, in Bragg v Linden Lab, that the the terms of service are a contract of adhesion and therefore draw strict scrutiny against the company that must not be named. Trying to enforce a law of property claim by getting law of contract parties to accept changes to a contract of adhesion is so ridciulous a legal tactic that it beggars belief. Indeed, it may actually have the unintended effect of abandoning an intellectual property claim against persons who do not accept the terms of the contract of adhesion.

Many companies, including now the company that must not be named, post boilerplate they know is void ab initio. It is to be hoped this is an example of an ambit claim, rather than a serious belief on the part of the company that must not be named that their internal memos somehow override the US bill of rights.

Software engineers really should stick to writing code and getting the grid up to scratch.

The article is an absurdity. Each solution they propose either does not work or has side effects that make it totally useless. I get inventory offers all the time. I do not want to set myself in busy mode and have those offers automatically rejected. Skyboxes are great. I like using mine. But if there's an avatar out there without a flight assist script I certainly haven't met them yet. You'd really wonder why LL keeps putting out spin that is not even good spin. Spin should have at least a faint chance of persuading the target audience.

'We get it' is not enough. Lindens are always getting it, What's needed is something concrete that shows LL is about to change its ways.

LL's root problem is that the code base is a small number of large undifferentiated blocks. It is poorly documented. Small changes in one segment of the code base can cascade into poor performance in a completely separate segment, because the code is cross-linked in ways that went out of use in about 1981. Tweak the UI and you find you're exploding subroutines in the asset server. LL needs a group whose task is documenting the code base in more detail. In the longer term, the degree of modularity in the code base needs to increase drastically. Waiting a little and thinking about documentation and design issues can be the mark of a great designer.

This unhappy state is a direct result of initial decisions to follow the get it running approach. It is cause for embarrassment, not the proclamation of triumph.

A monthly blog post is not a communications strategy. That's especially true when the Lindens continue to close comments after a fixed number ad so rarely respond to comments the way a normal blogger does. The way to cure the communications gap is to, um, engage with people and that means answering comments, not posting a rah rah statement and waiting for the arbitrary comment limit to kick in.

Sl is wonderful and SL is rubbish. Getting SL up and running was wonderful, but it happened a little time ago now and there's a point where 'But we were great in 2004' begins to sound a little stale. Philip could cure the problem immediately by announcing some actual improvements to the user experience.

One thing that springs to mind is the limit on groups. 25 was great when SL first rolled, 255 inventory items was equally great when SL first rolled. I don't know abut you but my inventory is a tad larger than 255 items these days, and I constantly juggle myself in and out of groups trying to keep track of the stuff that interests me. I'd be really surprised if raising the number of groups you can join would have any adverse consequences and it would give everyone a much better sense that SL is paying attention to the soft world, people's interactions, as well as the hard world, coding improvements.

In the longer term a way must be found to allow real crowds at least on some sims. When Philip addressed a group of companies recently he was asked a lot of questions about return on investment a company can expect from SL. The Royal Liverpool Philharmonic played to less than 90 avatars, although they seem to think they were communicating with 9 million. Until ROI is better, the level of outworld engagement is not going to improve. And that means we need crowds.

I think Philip's missing the point. Linden Lab needs to hire someone who can write manuals and they badly need some expertise on interface design. All his answers are engineering answers, and they just don't address problems with the user experience.

The way to measure lousy user experience is not the numbers who stay, but the numbers who sign up and never return after one or two sessions. Don't get me wrong, I adore SL. I just hate the interface and I hate even more the lousy documentation available to new users. I'm sure Torley Linden is a really nice guy, and his videos are great to watch, but they're not good instructional materials, and they cannot replace a couple of well-written how to pages.

Linden Lab also badly needs some professional advice on their legal proposals. Country-specific avatar rules are going to be a nightmare because criminal laws run by location, not the origin of the individual. In Gutnick's case the High Court held that a defamation happens where the material is downloaded from the Internet, not where it is uploaded. It's a civil case, but it's hard to see why an Australian court would not reason by analogy that a crime occurs where the criminal matter is downloaded. Gutnick's case has been either followed, or cited with approval, outside Australia, although it's also been heavily criticised in the US.

Linden Lab could find themselves in significant difficulties if a court outside the US held the Gutnick principle sounds in criminal law and that downloading matter in that court's jurisdiction is therefore criminal conduct. Applying different rules to avatars according to their citizenship does not solve the problem at all. Let's say a citizen of Country X opens a child pornography shop in world. A criminal court in Country Y will not have the slightest interest in the nationality of the uploader, no matter what the laws of Country X may be. A criminal court is likely, in extreme cases, to find an offence and authorise extradition.

Gary McKinnon was extradited from Britain to the US for breaking US law, even though he was not a US citizen and his alleged criminal conduct happened entirely outside the US.

Philip's a great engineer. That doesn't make him a legal authority. Linden Lab need to broaden the places they're looking for advice, both in world and outside. They also, just quietly, need a better feedback mechanism than Lindens wandering around in world.