Why Keeping Second Life 2 Closed Source is a Good Idea, Expressed in a Single Screencap

Many SLers are upset at the announcement that Linden Lab's follow-up to Second Life will be closed-sourced, so I was planning to explain why keeping SL 2 closed source is actually a good idea -- at least to start -- when this popped up in my Google Alerts, and pretty much did my work for me:

So a forum for CopyBot users (yes, these exist) is not happy about the closed source aspect, not happy at all. Why? Well, one likely reason: An open source Second Life 2 would make it much easier to build some kind of CopyBot-type program for creating unauthorized copies of SL 2 content. (In the same way that open sourcing the Second Life viewer in 2007 also made it much easier for CopyBot-type hacks to arise and spread.)

Now, before any flames arise, let me quickly add this:

To be sure, open sourcing the SL viewer also led to a lot of great, totally legit innovation as well, and I'm not blurring that distinction at all. That said, open source is not the only way to drive innovation. The closed source Apple App Store is full of thousands of amazing apps, most created by 1-2 people, many of whom make a great living from their content, and even sell variations of their app content on other platforms. Also note that I said "at least to start". For consumer-facing products, open source is an option best worth thinking about after mass market status is achieved -- but not before. (First comes the closed iPhone, the mass market for that is established, then comes the open Android standard.) At the time, when Linden Lab opened the code, mass market already seemed to be in reach; in retrospect, however, that proved to be premature.

Please share this post:

Tweet

Comments

[DontGetHairy]

Groupthink is a psychological phenomenon that occurs within a group of people, in which the desire for harmony or conformity in the group results in an irrational or dysfunctional decision-making outcome. Group members try to minimize conflict and reach a consensus decision without critical evaluation of alternative viewpoints, by actively suppressing dissenting viewpoints, and by isolating themselves from outside influences >

How to Prevent Groupthink
http://en.wikipedia.org/wiki/Groupthink#Prevention

[Galatea]

If the viewer isn't open source, then at minimum it needs to support plugins to modify its behavior. But if it has a sufficiently powerful API for plugins, then it need not (and arguably should not) be open source, all the innovations of third-party viewers could instead integrate directly into the main viewer, and better yet, allow users to mix and match features in a way we currently can't in SL. For example, I'd love to be able to use Cool VL Viewer (my viewer of choice) and the LSL preprocessor (from Firestorm) at the same time, but having all these innovations out in separate TPVs instead of as plugins prevents that. Getting all that innovation moved into plugins for one unified viewer would do a lot to *improve* the mod-ability of the SL experience.

[Ayamo Nozaki]

I wouldn't agree, libOpenMetaverse effectively made the first copy bot before Second Life was open source.

[Melissa Yeuxdoux]

So SL2 should be closed source because security by obscurity?

[Hamlet Au]

"libOpenMetaverse effectively made the first copy bot before Second Life was open source"

Yes, and Linden Lab supported that effort and it also helped impel them to go the full open source route with the viewer:

http://nwn.blogs.com/nwn/2006/11/second_life_clo.html

[Pathfinder]

Open sourcing the viewer revealed many client-side bugs in LL's code that were then able to be fixed (some of which were classic "whoops we're trusting the viewer too much" security flaws).

Claiming that closed source *on the client side* allows greater security is a specious argument and will get you laughed out of the room by any group of real computer security experts. And claiming it will stop copybotters is just sowing the seeds of Fear-Uncertainty-Doubt.

C'mon Hamlet, you're better than this!

[Pathfinder]

Also, if LL really wants to make an innovative viewer, then someday it needs to be on this list...

http://www.curse.com/client

[Arcadia Codesmith]

I think Galatea has a point. Build for third-party/user customization through plug-ins and integrated options.

Utimately I believe that building the Metaverse will require an open standard, but declaring one at the onset is putting the cart before the horse. You've got to plant the garden before worrying about the structure of the co-op.

[cathartes aura]

Third Party Viewers didn't stop SL mass adoption. Copybot didn't stop SL mass adoption. What did stop it was the rigging of SL's economic system by LL, the 1%'r Land Barons, and some Content Creators.

You guys killed the golden goose. Your greed is what did SL in. Land is what drives every activity in SL. By rigging the land market through !%'r goodies like the Atlas Land Program, grandfathered tier, reduced tier, and free tier for certain favored 1%'rs, LL made it clear to everyone not part of the 1% that they would always operate at a loss.

No one who's got a thinking brain in his or her head is going to invest in an economy that is rigged against them so heavily. That is what killed off mass adoption of SL. You guys got too greedy... ;)

[Dr Feelgood]

Copybot is a game specific hack but universal game hacks cannot be stopped like 3D Ripper or Ninja Ripper.
They will extract contents from 99.9 of all games including SL 2.0. there is no escape.

Myself i will hold out for High Fidelity it has the most promise. also is fully open source that is a community effort.

https://highfidelity.io/company

High Fidelity i will own my own world that i have complete control over ..check
It will use crypto currency..check
It will have a marketplace..check
I will get paid for helping to power grid..check
Real build tools like voxels ..check
Anyone can create features for the game..check
High Fidelity is completely transparent in its goals and motives since day one ..check
The software is free & open source ..check

The platform empowers its residents ..check

[Shug Maitland]

The problem is that in a world "in the spirit of Second Life" effective user input is key, the earlier in the development phase the better. Linden Lab has proven itself blind to problems and features that are obvious to the residents.
Thus far open sourcing the viewer has been an extremely effective way of providing that input. Firestorm support and jira, for eample, provide LL with an input service with the wackadoodles and ill informed filtered out.
LL needs to provide SELECTIVE community access as early in development as possible (it is always easier to correct a bad idea in the suggestion stage than after implementation). The open source developers provide just what LL needs.
Of course this ASSUMES that SL2 uses a viewer at all :)

[Han Held]

I expected this to be the usual anti-open source "fear, uncertainty and doubt" -and I wasn't disappointed.

I agree with Lust's comparision of HiFi versus SL 2.0 and I agree it's a better investment of time and energy than SL 2.0 appears to be.

[Metacam Oh]

If their virtual world is closed source it's going to be crap, because once again it will be what Linden Lab thinks everyone needs, and not what everyone really needs, so yeah maybe some shit will get stolen, that's life, but closed source is a dead end, apparently they haven't learned their lesson. I'm much more interested in HiFi now.

[Ajax Manatiso]

Not making it open source protects it from copybotters just like putting a lock on your front door means burglars will never get in. It only dissuades the totally incompetent. As for a whole new virtual world - LL already did a study that said the oldbies will never leave. I would be one of them. I will stay on original SL until they shut it down, and when they shut it down, I will be done with SL and LL forever. No Hi Fidelity or anything else - I will just enjoy RL and my tech projects on my home LAN.

[Hamlet Au]

"If their virtual world is closed source it's going to be crap"

Minecraft is closed source and has a much larger userbase than Second Life and has far more user-generated content. (Minecraft machinima on YouTube runs into the hundreds of millions of views, for instance.)

[Pathfinder]

@Hamlet

Minecraft also has ~12,000 user-created plugins.

http://www.curse.com/bukkit-plugins/minecraft

Nothing inherently wrong with closed-source clients (I was only arguing against your "they're more secure" stance ;).

But if LL wants to encourage real client-side innovation, I think they should support user-created plugins like Minecraft, WoW, Skyrim, etc.

[joe]

The headline for this article should have been: "Has LL made stupid decisions about SL in the Past?"

The answer is, "Of course!"

The next headline should read: "Will LL make more stupid decisions in the future?"

[Adeon Writer]

Copybot has been remade from scratch by different sources many times. It's all about looking at packets, and learning how to automate the building system. Clientside source might help, but it's really not necessary... you just need a viewer and a proxy.

Thinking a closed source viewer would in any way stop the creation of a Copybot is like thinking Piracy protection in videogames is anything more than a waste of money.

[Metacam Oh]

What Pathfinder said, if they are going to be closed source they need user generated plugins.

[Metacam Oh]

Also, I might add, if Minecraft shut down those user mods, they would have been dead years ago, but Minecraft goes on and the mods give it infinite life.

[Pussycat Catnap]

I lean slightly in favor of a closed source client for the new SL MMO (and that is what SL is...) not for security reasons, but for reasons of the shared user experience.

That said...

If they go this route, they need to back it up with a very robust addon developer kit, like the 'LUA' language of other MMOs.

Otherwise... much of the SL experience will be lost.

I care nothing for RLV, and find it distasteful... BUT it is a vibrant part of the community...

And an LUA addon based version of RLV... sufficiently limited in scope due to the limited reach of an addon... would have very useful gaming and alt-lifestyle utility while lacking the various dangers that make RLV too extreme for me to be willing to use an RLV enabled viewer.

(ie, control of action leads to response would be unable to wrest avatar control from the actual game player - and as such D/s roleplayers would have to learn the real-life lesson that it is supposed to be about consent and not force. But at the same time, LUA-like addons could be used to enhance gamism.)

There has been too much inconsistency across the SL community for the last few years - which I believe is a big part of the motivation to try and put the lid back onto Pandora's box.

[Phadrus]

If I were to entertain the notion of being an executive at Linden Lab, I would seriously consider including OnLive’s SLGo as a key component in the overall user experience. That is, I would propose that SLGO become the primary portal by which users should access SL 2.0; going so far as to actually make it mandatory.

If pulled off successfully, it would, at a stroke, eliminate three outstanding issues that have plagued SL from the very start:

1. One client, one point of entry. Other than the relatively simplistic OnLive software, which works on most if not all OSes (and soon iOS), no additional client installations are required. Hardware and OS/video driver compatibility issues will be a thing of the past. Furthermore, everyone will be automatically upgraded to the latest version. No more waiting for the reluctant user base to get around to updating their clients. Every resident will be able to see the latest technology as it goes live.

2. Stellar performance across the board. With the sheer amount of non-optimized content, Second Life’s main Achilles heel was being graphically intensive to the point of demanding the latest hardware for the full visual experience. SLGO was created to solve that very problem and it works remarkably well. No matter if you have a tablet or a Pentium 4, if you can stream Netflix, you can view SL 2.0 at the highest settings.

3. No more copybot. With a single point of entry, rogue copybot clients and other external utilities that can be used to extract information such as textures, audio, animations and mesh data for the purpose of re-distribution will cease to be effective. There’s simply nothing to steal from a video stream other than by way of analogue.

For this to work, there would need to be substantial improvements made to the current existing SLGO service up to but not limited to the following:

- SLGO must be offered free of charge. Linden Lab will need to find a way to subsidize this – someone more resourceful than I may know a way.

- Improve latency times to provide vehicle operators with a more seamless experience.

- Higher resolutions for those with connections greater than 10 Mb should be offered while continuing to perfect their compression algorithms to ensure the highest video quality possible.

- Included with each account are cloud storage options for avatar and settings backups. Cloud storage can be accessed via account login on the web for the purpose of downloading and deleting content in addition to uploading user contributed material such as textures, video files, animations, dae files, etc.

- Assets uploaded to the SL 2.0 grid from SLGO cloud storage can then be deleted from cloud storage to conserve space.

- Screen capture resolution support of up to 4k saved to cloud storage or sent to Flickr, Facebook and email in PNG format. These are of higher quality that faithfully captures imagery beyond what is seen in the video stream.

- 16:9 video capture support with a recording length equivalent to the amount of cloud storage available. These are of higher quality that faithfully captures imagery beyond what is seen in the video stream.

- To conserve bandwidth, limit the number of simultaneous account logins from the same IP from between 3 to 5. A comprehensive system should be available to assist with populating environments with AI avatars for those who need them.

- Offer VR headset support where possible.

- Allow third parties to create extensions supported by the client for things like UI improvements, combat systems, and additional functionality via dev kits. A beta server can be provided for testing. Users should be able to choose which extensions they wish to use by enabling or disabling them in the same manner browser plugins are managed. Third Parties can also sell/license plugins on the marketplace; enabling developers to monetize their efforts if they so choose. Hard coded limitations and an active screening process can help mitigate the surge of potentially harmful extensions.

- Linden Lab should stay out of UI development in general and allow the market to be the market. The primary focus should be on maintaining and improving customer service along with the core of the platform.

- Provide greater access to advanced tools and monitoring utilities found in the full featured local client.

- Include an optional text based client that can be installed locally for basic communication, inventory and currency management.