(Update: If importing 3D models in Second Life is important to you, please vote for this issue on Second Life’s Issue Tracker: http://jira.secondlife.com/browse/SVC-2634. You’ll need to login with your SL name and password first)
I’ve been using Second Life for a number of years now to run experiments, through our ‘Studio Wikitecture’ group, to see if a Metaverse, such as SL might one day act as the platform for collective intelligence in architecture and urban planning. I’m also an active member of the ‘RL Architects in SL’ group. As you can imagine, most of the members that compose these two groups are architects. What I have found, participating in these groups over the last year or so is that new members are always excited about using Second Life as either a collaborative tool or as a platform to exhibit their work. Unfortunately, however, their interest soon wanes when they find out there’s no easy way to import in models from third party programs. I realize there’s a number of grass-roots initiatives out there that have developed rough and crude ways to import from the following:
Although, I give my utmost respect to the programmers that have developed these projects, to say they are crude, is to pay them a compliment. What invariably happens is that after sharing these links with the many people that ask, they come back (if they come back at all) even more confused and frustrated having labored through the elaborate and evolved process of copying and pasting pieces of code back and forth between one program and the other. The process is about as easy as painting with rice grains.
Having had a number of conversations over the year with people about this, I can say with confidence, that SL’s lack of portability is the number one hurdle for our demographic and the main reason why many people never come back. Although speculation, I would imagine this is a major hurdle for other groups as well.
What I don’t really understand is why this issue is not pushed more by the SL community at large. I have noticed there were a number of issues posted on SL’s Issue Tracker that call for portability of a number of various file types, such as .OBJ, .3DM, .3DS, .DXF, & .DAE.
What I don’t really understand is why, firstly, the overall SL community is not voting on this en masse and secondly, why have some of these grass-roots initiatives outlined above, just withered on the vine? Some of these projects are over two years old, with no sign of life or continued evolution since they were first posted.
Having been in the middle of this conversation for awhile, it seems the standard responses usually involve two factors: technology and/or SL’s economy—Technology, from the aspect that it’s currently still too difficult to do and economically, from the aspect that the sudden influx of new models would dilute the value of existing in-world creations, resulting in a negative impact on SL’s economy. Although I’m sure there are more reasons, these seem to be at the forefront of the discussion.
What confuses me, from my perspective anyways, is that these reasons still don’t seem plausible to me and I’m found wondering if I’m missing a valuable part of the equation.
First, although I have a limited background in programming, it appears from the existence of these grass roots projects, that portability is indeed obtainable. Having dabbled a little with each project above, I realize the process is laborious—cutting and pasting code from one program to the other.
I also realize that most of these conversion programs drastically simplify the form when imported into SL, such as textures being stripped off, and meshes and certain objects such as cylinders and sphere’s being simplified down to plain ‘box’ prims in SL. Although these are indeed hurdles, the technology is currently there to do this on a very limited basis.
What I don’t understand, is why these projects haven’t evolved into a more user-friendly format after the years they have been in place. This is just a lack of user-interface design verses a lack of back-end programming. Even though they are crude and might only import texture striped SL boxes in some cases, i know that I, as well has a horde of others, would still jump at the chance to have access to a tool like this and in most cases would actually pay good money for such a thing.
On to the economic end of the argument. I guess I can see on the surface, why a number of residences might be fearful of a world where seamless importing and exporting becomes the norm. I would imagine their argument springs from the fear that they either think their creations would be diluted by all these new models or that their creations could easily be exported and in turn imported, further diluting their value. I think these are legitimate concerns, but are there not ways of regulating this? Could you not add an additional layer of modifications rights that would give the owner of the in-world object the choice on whether they allowed subsequent owners to export out their creations?
From the importing perspective, yes, I’m sure such functionality would initially effect the market. The problem is however, that in the long run, the longer SL holds out on this functionality, the more negatively it will impact the economy, that is, when they finally do implement it. I’m not an economist, but I would imagine this is the same phenomenon you see when age old tariffs are removed and the economy in which they were trying to protect in the first place, finds itself unprepared to compete in the more innovative and efficient economy that was growing up all around them the whole time.
The assumption here, of course, is that they will do this eventually. The reason being, I believe, is that they will be forced to from market pressures from other virtual worlds offering portability as standard service and perhaps. In the end, perhaps this is exactly the reason why SL hasn’t offered this service yet. There’s no real viable competitor yet.
Although part of the reason for writing this post is to rant a little, my main objective is to start a conversation around why you don’t think portability has become common place for the end-user by now. I just wanted to share my confusion with everyone and see if there are others that are either just as confused, or perhaps have a more nuanced answer to this problem.
And by all means, if this too is a thorn in your side, please place your vote on Second Life’s Issue Tracker. Pick your poison. .OBJ, .3DM, .3DS, .DXF, & .DAE. Voting is easy, just login in first.
And if you’re a developer, just know that I’ll be the first to put my order in and I’m sure I wouldn’t be alone. The market seems to be begging for it.


















15 comments
Comments feed for this article
January 15, 2008 at 1:38 am
csven
From the programming perspective, you also have to recognize that there doesn’t appear to be sufficient incentive to develop such tools; especially not when another virtual world (e.g. Novoking) can come along and allow exactly the sort of input/output you’re discussing (mesh data).
Rather than meshes, however, I’d prefer to see more parametric tools inside SL as well as a means to move them in and out. There are ways to do this, but it really has to start with Linden Lab imo. And they don’t seem particularly concerned with adding additional modeling functionality (and data exchange); at least not until they’ve addressed some important core issues with the platform. So until stability and concurrency improve, and “web on a prim” arrives, I don’t expect much (I won’t get into what *might* happen should they open source the back end).
In the meantime, IBM’s involvement with the OpenSim effort is worth watching. For reference: http://blog.rebang.com/?p=1411 . It might not actually result in much change in the modeling tools, since they’re probably looking at 2D pattern interchange, but who knows. At least the mindset is there.
January 15, 2008 at 3:12 am
Importing and Exporting in Second Life: Oh, How I Pine « The ARCH
[...] Theory Shaw, import tools, ryan schultz By Ryan Schultz (Theory Shaw in SL), cross-posted from Studio Wikitecture [...]
January 15, 2008 at 4:18 pm
Mace
I just wanted to point out two things. The first being that although I would love if SL supported the import of OBJ or 3DS meshes, that is a massive can of worms. The 3D community so so massive and there is such a glut of free content out there that the grid would be bombarded with all the half assed 3D meshes created by hobbiest 3D artists like myself. Not to mention that you would see a wave of content from DAZ, Renderosity, RDNA and a number of other being resold in SL. Don’t think it would happen? Right, go “art” shopping sometime. Copyright infringment is already rampant on the grid do we really need to invite more of it? Second, there is the SL Prims import engine for building in 3DS Max then importing that into SL. It works pretty well and if I feel like building offline I just fire up 3DS Max.
Honestly, LL needs to consider file types but more so they need to add three things to their toolset. The ability to extrude faces, the ability to chamfer any face and a simple UV Unwrap tool. Personally I would like to see them add a simple scripting wizard for the creation of simple gadgets.
January 15, 2008 at 9:53 pm
The Grid Live » Second Life News for January 15, 2008
[...] Studio Wikitecture Importing and Exporting in Second Life: Oh, How I Pine. Quote from the site - I’ve been using Second Life for a number of years now to run experiments, [...]
January 15, 2008 at 11:10 pm
theoryshaw
Hi Mace,
i guess i would say providing better tools to a larger section of the population has always resulted in better content, period. Of course, such assess does result in more drivel as well, but we are all experts, in this sense, of filtering this out. That’s why collaborative filtering works.
seen a good youtube video lately? Imagine if the only format that youtube accepted was 8mm video, there would be less of both the good stuff and the bad stuff.
January 15, 2008 at 11:55 pm
Mace
Ok, somehow I think we are talking about two different animals. You are are still ignoring the simple fact that if people could import the various mesh types then you would have creaters of 3D content being ripped off because lazy SL merchants would immediately start re selling the available content. The amount of money tied up in copyright infringment right off the bat would be enough to cause LL not to allow this. Keeping in mind there has always been a rumor of OBJ import because the base SL figure is a Poser2 mesh/OBJ. It has been roughly four years and LL is still making money hand over fist despite not having the ability to import other mesh types. Not to mention that SL is already not terribly user friendly. Most people who bring home an e-machine from the local computer store can probably get on the grid albeit with lag which is going to greatly limit their SL experience, probably frustrate them and send them off to play World of Warcraft. I wouldn’t even want to think about someone trying to import a hair object made in Maya or even a simple cloth object. I understand your frustrations but I think you are trying to pound a square peg into a round hole.
Don’t get me wrong I am not against the idea but there are valid arguments against it and if you can show me a virtual world which is allowing common types of meshs to be imported, I will eat crow without salt. But if you want to increase the quality of content 100x then the answer is a better tool box not file portability. Granted that is my opinion based on what would be easier to accomplish. Basically I think we are more likely to the SL tool kit expand before they allow the import of other file types.
That being said, go try SL prims I forget the URL but you can google it and it is very user friendly.
January 16, 2008 at 1:59 am
csven
While importing meshes isn’t at the top of my wish list, I don’t have as big a problem with it as Mace seems to have.
First off, just like textures, a cost could be associated with uploading a mesh file. The $10L is so trivial it’s essentially non-existent. People routinely ask for and receive that amount from other avatars when asked “You got $10L so I can upload a photo?” It’s not a barrier.
Now, if the cost weren’t trivial, things change. First you decrease the amount of uploads. Second, the value in uploading has to offset the cost. Maybe someone just *really* wants their model uploaded so they can show it off because at, say US$20/upload it’s no longer trivial.
Now, that puts it in the range of a “necessary recoverable”, and that probably means sales. Except the sales process might not be the same as in-world models. Linden Lab *could* tag uploaded meshes so that they don’t Copy (after all, they once claimed they would incorporate some kind of system to help resolve IP disputes).
But No Copy affects the legit folks as well. Now the system has shifted… perhaps too far. So how to bring it back a bit for the legitimate types who want to sell their 3D? Well, it’s Linden Lab’s servers, so it seems they could either vet the uploads themselves or hire someone else to “certify” not just authenticity but RL information. That way, when someone uploads a ripped model, there’s a database on-hand for comparison and accountability purposes. Thus, someone wanting to upload a model pays their $20 and then has to be vetted in order to sell it; and if they’re ripping someone off they lose the money and the account as well as being potentially sued.
That’s top of my head, but I think it’s apparent something *could* be accommodated. Cory O. mentioned during one interview I read that a hybrid model environment would eventually be needed. And there is *already* a mix so I think it’s the control issues that are among the bigger ones with which they need to deal.
On the other hand, for a few years now I’ve been wanting to see such a system apply to *all* uploads. That would mean the client would also have a built-in set of graphics tools (e.g. Illustrator/Photoshop replacements). They already have some kind of inworld animation system tool (though when it’ll be rolled out, who knows). And if they went to procedural textures they could solve a lot of storage issues on that count. In addition, they could also both improve the parametric modeling tools as well as incorporate some mesh tools.
But all of this is a ways off, I’m afraid. Hope I’m wrong, but not holding my breath.
January 16, 2008 at 2:02 am
csven
“Don’t get me wrong I am not against the idea but there are valid arguments against it and if you can show me a virtual world which is allowing common types of meshs to be imported, I will eat crow without salt.”
“A new feature that truly differentiates Novoking from SL and Hipihi is that it allows users to upload their own content created in Photoshop, 3DMax and Maya. So besides aiming for just the ‘newbies’ the developers at Novoking have left plenty of room for the more experienced content creators too.” - http://mastersofmedia.hum.uva.nl/2007/10/24/virtual-worlds-in-modern-china
January 16, 2008 at 4:14 pm
Mace
I will check out novoking as soon as I get home.
I hope to gods that LL does not include a set of imaging tools in the toolkit. Most 3D modeling apps really dont take much of your system resources until you start rendering. By comparison Adobe tools take a great deal of your system resources and adding that to the SL toolkit would be like trying to run a race standing in 3ft of molasses.
Not sure what you mean by rolling out the inworld animation tool, Sven, I know someone now who does poses and animations in world I just dont recall the name of the app. As for charging $20 an upload well preposturous is the first word that comes to mind. I am talking about adding three tools to the modeling ability of the tool set. They already have the basis of those tools in the tool kit. Sven, your solution would cost more in time and money then it is worth. Right now the SL tool kit isn’t too far off from similar and less user friendly development kits like UnrealEdit and hiring someone or even a team of someones just to certify uploads? Dude! Do the math. Whats the final cost if you pay a five person team for this job? Hourly, benefits, 401k? Seriously, why? When an exisiting team of code monkeys can just work out how to add three measely tools to SL. I think you guys are on some heavy drugs.
January 16, 2008 at 11:06 pm
csven
“Not sure what you mean by rolling out the inworld animation tool, Sven, I know someone now who does poses and animations in world I just dont recall the name of the app.”
To my knowledge the inworld tool developed internally by LL has not been released, so that person is almost certainly uploading anims. If you’re not familiar with the LL tool, I’d suggest doing a search. btw, it’s said to have the potential to kill off the animation industry inside SL since it makes creating individual custom anims so easy. Might want to alert your acquaintance.
-
“As for charging $20 an upload well preposturous is the first word that comes to mind.”
Fair enough. And “predisposed to in-the-box thinking” is what comes to mind when I read your comments. To each their own.
-
“I am talking about adding three tools to the modeling ability of the tool set. They already have the basis of those tools in the tool kit. Sven, your solution would cost more in time and money then it is worth.”
Based on what analysis and evidence do you base your factual claim? Or is it just an opinion?
As to cost? Cost to whom? Linden Lab? Worth? Worth to whom? You?
I suppose it didn’t occur to you that someone *outside* LL might add functionality all on their own. For free. For example, Georgia Tech’s recent modifications to the client will, afaik, be made available to everyone. That is, iirc, mentioned in their video.
And of course, if/when the back-end goes open source, we should all *expect* additional applications and implementations. Isn’t that what an open API is for? Isn’t that the big news around Facebook’s 3rd party devs (esp the Scrabulous guys today)? Isn’t that why 3DS Max’s plug-in capabilities were so popular back in the early 90’s? Extending and/or modifying the core application? Maybe SL shouldn’t have *any* tools in the core client and people should be able to add plug-in modules as needed. That would simplify the oft-cited confusing UI issue, no?
Excuse me if I’m not limiting myself to three little things.
-
“hiring someone or even a team of someones just to certify uploads? Dude! Do the math. Whats the final cost if you pay a five person team for this job? Hourly, benefits, 401k? Seriously, why?”
Ask Makena/There. They’ve been doing it for years.
I made the point of calling out Linden Lab’s effective ownership of Second Life. So long as SL is *not* the internet - is a self-contained virtual world - an option might be feasible because they do, in fact, control the space. Thus, if the upload cost for a mesh was… say… US$1000, they’d probably not have to hire *anyone*. But hey, at least the option is there, and that’s a good thing afaic.
In addition consider that many of the problems with SL seem to be asset-related. Certainly seems like it’s already in need of some rethinking (and maybe they’re already decided; again, Cory O. in an interview already said a hybrid model system was inevitable).
In the end, perhaps LL will run two grids: one that’s basically what you have now and another that they restrict in the way I’m suggesting they could. Furthermore, perhaps 3rd parties will eventually either completely do away with uploading OR allow any & all uploading. We’ll see.
The internet allows anything and everything to be uploaded so if Rosedale’s goal is for his tech to *become* the new 3D internet, then limits *will* go away in the core application. Thus your initial problem comes back to haunt you anyway, doesn’t it?
-
“I think you guys are on some heavy drugs.”
And I think you’ll enjoy the crow… without salt.
While you’re munching away, please give a little thought to the possibilities instead of casually dismissing any discussion of anything that seems on the surface to be “preposturous”.
January 23, 2008 at 4:22 pm
Elvina
back to the original question:
“my main objective is to start a conversation around why you don’t think portability has become common place for the end-user by now.”
I’m not an expert and not that tech but consider these questions:
why do you think there used to be only very limited prim tools in SL from the very beginning until the very recent addition of sculpties? Why do you think the existing sculpties are limited to 32×32 vertices mesh as their highest possible resolution? why do you think there wasnt the ability to import any mesh from the very beginning? why do you think there are just a certain amount of prims possible in a sim? i guess the answer would be: processing power. The more prims there are, the more vertices/faces/polygons they have the bigger is the rendering weight.
You all know how unstable and full of all kind of problems SL is always used to be. So i’m afraid, until they find ways to make it more stable and run smoothly with the tools we have right now, we wont see any changes any time soon: neigther 3 extra “simple” add-on tools (which btw could result in possibility to have endless amount of vertices/faces per object!) nor arbitrary meshes support. It would just mean creating even more problems for themselves while they cannot even cope with the existing ones!
I’m just trying to be realistic here, I myself have just learn how to achieve a “decent” quality of sculpties from my detailed max models and I’m endlessly thankful to those who make those “grass roots projects” possible and even downloadable free of charge! And I will gladly vote for any extra modeling tool, be it simple chamfer or a mesh importer. I just don’t believe it will be there any time soon.
January 23, 2008 at 4:47 pm
Elvina
Quote from http://wiki.secondlife.com/wiki/Sculpted_Prims:_FAQ :
“Q. Sculpted prims are limited to 32×32 vertices at their highest resolution. If I’ve uploaded a 256×256 sculpt map, how can I get the sculpted prim to show that detail?
* You cannot. Currently in Second Life, we use the “prim economy” to prevent builders from “stealing” rendering resources from each other. Each region is allowed 15,000 prims, each of which may have 1024 vertices. These resources are divided proportionally amongst the region’s landowners. This balance works well with the current state of rendering hardware. There are no plans to increase these numbers in the near future. “
January 23, 2008 at 4:47 pm
Elvina
Quote from http://wiki.secondlife.com/wiki/Sculpted_Prims:_FAQ :
“Q. Sculpted prims are limited to 32×32 vertices at their highest resolution. If I’ve uploaded a 256×256 sculpt map, how can I get the sculpted prim to show that detail?
* You cannot. Currently in Second Life, we use the “prim economy” to prevent builders from “stealing” rendering resources from each other. Each region is allowed 15,000 prims, each of which may have 1024 vertices. These resources are divided proportionally amongst the region’s landowners. This balance works well with the current state of rendering hardware. There are no plans to increase these numbers in the near future. “
January 25, 2008 at 11:51 pm
csven
From what I understand from researching these issues a few years ago, Elvina, afaik you’re mistaken on several counts.
“why do you think there used to be only very limited prim tools in SL from the very beginning until the very recent addition of sculpties?’
Simple. Linden Lab is not a 3D development company. That particular expertise did not until relatively recently, afaik, exist within the company. They outsourced the 3D toolset work.
“why do you think there wasnt the ability to import any mesh from the very beginning?”
Simple. Prim data is lightweight. Streaming/compression is Rosedale’s strength. From what I understand they wanted 3D which didn’t incur significant bandwidth. CAD-style parametric/procedural models arguably do this more efficiently than mesh data (note that “Spore” will be similar in this regard).
“why do you think there are just a certain amount of prims possible in a sim?”
According to things I read some time back, prims were limited not because of their own resource requirements, but because more prims potentially meant more textures, and *those* were the biggest cause of concern.
The official forums have been pretty badly gutted, but you might find some preserved references for how LL made some decisions and, if true, it wasn’t always what one might expect.
Consequently, “processing power” is not *the* answer. It’s a part of a bigger answer that includes streaming the 3Ddata, streaming the texture data, aso.
“The more prims there are, the more vertices/faces/polygons they have the bigger is the rendering weight.”
But rendering weight is a client issue. It’s local. And it can be dialed down locally. It’s already done this way. But Linden Lab can’t really control the textures attached to those prims. They get streamed to everyone regardless. And that gets to the issue I mention above. It’s also related to the “prim economy” in that it’s not purely a tech issue but also a social issue.
btw, the consultant who coded the modeling tools offered to add more capabilities during early development (and even relatively recently, according to his blog). Linden Lab declined, apparently. For what reason I don’t know. However, “limited” prim tools != limitations on vertices. An extruded square is still a six-sided cube in the end, and an extruded pentagon would only be seven. A twisted, tapering toroid is a bigger hit than an extruded I-beam.
Of interest: someone recently pointed out on their blog that the 3D engine was subdividing the flat sides of a cube into not two triangles but eighteen! I’ve not verified what that’s all about, but it suggests to me that the LOD system may be doing some odd things. LL is probably aware of this but it’s probably not high on their priority list since what’s streamed won’t change.
May 2, 2008 at 5:59 pm
Stef
I have read the preceeding article with interest. While I am not an archetect I too would like to import more freely. However you might look at AC3D’s second life tools. They might have what you are looking for as they seem to have a means of importing such things