Saturday, February 27, 2010

The Third-Party Viewer Policy: The shoddiest policy from the Lindens

As I noted in the previous post, coinciding with the release of Viewer 2.0 was the release of a third-party viewer policy created by Linden Lab. At the time I had read it, I did not think much that it was a bad policy nor think it would negatively impact third-party viewer developers.

I was wrong.

In fact, there's so much wording in the policy that not only negatively affected developers, but also violates terms of the GPv2.

For example, the GPL explicitly disclaims liability of a developer, yet the policy puts it on after the fact. Also, the GPL states that you may not impose further restrictions, conditions or terms of distribution beyond what the GPL specifies. The policy does exactly that.

But my biggest beef with the policy is Section 2(b), which flat out bans export of full permission content without the legal consideration when the creator of a piece of content grants exportability to the content's users. This effectively kills content licensed under Creative Commons, open source, public domain and other content whose license allows exportability.

There are other issues as well. In fact, as it stands no third-party viewer can be in compliance with the policy and be able to distribute. But thankfully, as Tateru Nino's Massively article points out, the Lab has brought the policy back to legal for a rewrite(this has been confirmed in the opensource-dev mailing list). Here's to hoping they get it right this time!

BTW, the only folks who are seriously lauding the policy as it is now, are those who are opposed to third-party viewers in general, so that should be a clue as to what's wrong with it if still in doubt.

Thursday, February 25, 2010

My Impressions on SL Viewer 2.0: the good, the bad, and the ugly

The other day I had gotten the word that Viewer 2.0 had finally been made available to the general SL community. I've had the chance to give it a test run and while I generally approve it as a major step forward, there are a few annoyances that I found. So here's my quick rundown of the good, the bad, and the ugly of Viewer 2.0:


The Good

I'll start with the new features introduced besides the new UI. The most welcome new feature is Shared Media. Ho..ly... crap! A truly functional web browser on a prim. It's per face, per prim, can have several media elements, and puts parcel media to shame. Flash content works(though I've yet to test on something HTML5-friendly like Theora). I've used it to do a tweet completely within a prim that has Shared Media enabled on a face. So we can expect some pretty crazy media mashups from machinima makers and artists soon.

There are two new clothing pieces introduced in Viewer 2.0: tattoos and alphas. Tattoos are pretty self-explanatory. Alphas are basically the replacement for invisi-prims on avatars to hide body parts, used for non-human avatars like tinies and dragons.

Now I'll get right down to the new "web browser" UI design. While residents who have become very used to the old viewer's design will no doubt find the redesign frustrating at first, once you start to find where all the menus have moved to, you'll find that virtually no feature sets have been actually removed. Landmarks are still in your inventory(as some like Prok feared they would get replaced with a "bookmark"-like feature), the build tools are practically unchanged(with the exception of adding Shared Media settings to the Texture tab), and chatting/IM has gotten a bit easier to keep track of: profile pics are shown before each line of chat and when an avatar says multiple lines of chat in an IM before anyone else, those lines are grouped under that avatar's name automatically instead of having the name repeated every single line.

There's more stuff about the viewer that I like, but I'll defer that to this blog post from Second Tense, as it is covered in pretty good detail.

The Bad and The Ugly

I have to lump the two categories as one because quite frankly, I don't see any serious UI or feature issues that would put me off using Viewer 2.0 when it becomes the mainline official viewer. That being said, there is one major thing that needs to be addressed: it's seems pretty laggy that it's consistently sucking up over 100% CPU usage. Granted, it will get worked on, but it has to be said now because it can be enough of a turn off for some folks.

And while I said that virtually no feature sets have been removed, there does seem to be one thing missing, though I'm not going to crumple up in despair about it nor ignore it either(Well, it's still there although not as clean as in 1.23 according to Prok).

Also, it appears they are changing their open source release strategy by attempting to label Snowglobe 2.0 as the source code for Viewer 2.0 and onwards. That may particularly irk third-party viewer devs a bit, especially in light of the new third-party policies and registry launched alongside Viewer 2.0. I'm irked simply because there are no Linux binaries of Snowglobe 2.0 available yet(this has happened before, and I'm not gonna slog through trying to compile it). Also, Linden Lab needs to be reminded that when you release any GPLed viewer to the general public, you must provide the exact sources used to make it, not the sources to a distinct derivative, as Snowglobe is(unless it is the exact sources with just the name and logo different, then it's okay).

Verdict

Viewer 2.0 is a big step forward, though there will be shock and adjustment pains. If you can get through the initial shock, then you'll find there's a lot of stuff in 2.0 you'll like.

Tuesday, February 23, 2010

More on the CDS controversy

Thanks to a tip received in the previous post, I've found out that there potentially may be more to CDS' database than simply an avatar UUID/name and viewer info. In a very long thread over at SLUniverse, Skills Hak makes the claim that "pc info" gets stored(IP and MAC addresses*), as another method of detection once the copybot viewer makers figure out ways to get around CDS. If this is the case, then that's completely against the TOS, if not flat out illegal. The only way to conclusively find out for sure, is to have the whole CDS system(client, bots and database) submitted to Linden Lab for a full audit, as Skills' word is proving to be not enough.

Also, CDS users apparently have the ability to set up their system not to ban on detection, but observe and record into the CDS database. Although seemingly a much more benign option, this is still a sort of networked vigilantism, as while the detected avatar is not banned from that particular parcel or sim, he/she is wherever a CDS is set to automatically ban, and the avatar's info is still propagated throughout the CDS network.

If I find out more or new developments arise, I'll be sure to post it here.

*Update: It appears Skills put up a FAQ where she says that "The specifics of the data cannot be disclosed. However, no MAC or IP addresses are stored, nor any other sensitive data. The most personal data is the avatar key and which viewer was used." However, I'm not convinced unless she submits CDS for auditing, as I mentioned above.

Monday, February 22, 2010

How CDS is engendering vigilantism among paranoid content creators

It didn't take long for it to start hitting the fan. Within days of releasing a scripted product that claims to 100% detect and ban users of copybot viewers, we're starting to see how this item can engender vigilantism in content creators who want a quick, zero tolerance "solution" over the lengthy legal procedures recommended by Linden Lab and copyright law itself.

What brings the point home are the comments by the sim owner whose CDS initially banned the blogger's avatar, and the very enlightened comment by a Mr. Peter Stindberg:
sabinagully said...

Hi everyone! I am Sabina, the owner of Magika. I wanted to clear this whole thing up. I hope many people get to read this post.

Arora Zanzibar have indeed been banned from Magika for the use of Neillife. Skills have taken the time to double check and it is true. Arora is just trying desperatly to clear her name, therefor this blogpost and all the fuzz that comes with it, fairly annoying if you ask me. It has come to her attention several times by other people on here to contact ME about it. She has not contacted me.
Arora might not be a frequent user of neillife, but she has been caught at some point using it. Anyone using Neillife, Cryolife, Fucklife etc is not ever welcome at Magika. There is no use blaming Skills for this. It is MY choise and my choise alone to use this system and keep her banned. Anyone who disagrees. well buhu lol.
Peter Stindberg said...

It gets even more bizarre when you see some strong advocates of shutting down 3rd party viewers in general now praising this system.

What this whole debate shows, however, is that content creators feel the pain, and feel the need to turn to take the law in their own hands out of frustration that LL does not do anything. THIS is the REAL issue here.

I do not trust a fully automated system. I work in software development and have quite a critical distance to fully automated systems. The question is not IF they screw up, but WHEN they screw up. But I can't blame a single content creator to buy and use such a system. However the deployers of such a system take on a HUGE responsibility. LL gets blamed a lot for the intransparent shoot-first-ask-questions-maybe-later process of Abuse Reports. The developers and deployers of this system seem to run into the same trap. An automated system without a clear, transparent appeal process is a loose weapon. This whole incident has made this very clear.

I am a marketing and PR specialist, and from my professional view this was a close call to a complete PR disaster. The developers of the system have a great chance to give the community an effective tool. But it needs transparency of procedures and clear communications.

Maybe this incident was a warning shot.
The plain truth here is, is that not only did Arora have to appeal to Skills, she may very well have to appeal to every parcel/estate owner who is currently using CDS, because it is designed to operate like BanLink, a previous product that engendered vigilantism among land owners. Also, it seems plausible that at no time did she actually engage in the act of copybotting. For the purposes of comparison, Linden Lab won't ban users for using copybot viewers, but they will ban a user if found to engage in copybotting or any other illegal activity. CDS, and its' growing network of users, may in effect interfere with the Lab's ability to properly mitigate copybot claims, particularly when automated AR functionality is implemented into CDS.

If CDS instead detected copybotting activity rather then copybot viewers, then there wouldn't be much controversy about it right now.

I should note, btw, that Arora was successful in clearing her name from the CDS database, after taking her case to Skills(Her name may still be in many banlists affected by CDS, though).


I should also say that I am in no way defending copybotters here, I am simply making the case that CDS poses the danger of networked vigilantism and escalation in SL. Think of a forming "JLU of content creators minus the spandex", and you get the picture.