Hi,

The collada is not in the away3d.loaders.parsers.collada?what is the instead?

You can use preFab to import DAE file ,then export it as obj file

it’s a working-around. I am just wondering does away3D support collada model for now? as I checked the old version of API reference, it did support.

COLLADA is not supported in Away3D 4.0 right now. We are planning to support it in the future, but we are not rushing there because frankly, it’s not a format you should be using. There are better alternatives for most of the frequently-used features that COLLADA supports, so we encourage you to use one of those instead, e.g. 3DS for mesh data, MD5 for skeletal animation, and in due time AWD2 for anything.

COLLADA is a very complex format which isn’t implemented the same in all applications. Different generators support different “flavors” of COLLADA, meaning that on top of supporting a very complex base format, a parser implementer (e.g. us) has to take all these peculiarities into account. This makes the process of writing a parser for COLLADA file, that “just works”, very difficult.

Add to this our strong opinion that COLLADA is a redundant format that people shouldn’t be using in the first place for real-time applications, and you can see why we choose to focus on other areas of the engine right now. Not to say we don’t want COLLADA support. We do, and we will add it when we find the time (or the right contributor to implement it.)

What are you trying to import? Just simple mesh/geometry or scene data, or animation as well?

I agree,and I use 3DS for mesh importing now in my project.
I am trying to import mesh and animations. it seems I will import animation using MD5, is there a tutorial about this?
Thanks

Richard Olsson - 25 July 2011 07:50 AM

COLLADA is not supported in Away3D 4.0 right now. We are planning to support it in the future, but we are not rushing there because frankly, it’s not a format you should be using. (...)

Hi.

I respectfully disagree: I believe that developers should be the only ones to know and decide what formats they should or shouldn’t be using.

Collada has it’s place and is anything than redundant. It is an interchange format, not intended for retail run-time use (but rather for production run-times; it is a format explicitly created for real-time content after all, contrary to your opinion). People using it as their retail format shouldn’t probably be in a position to make such a call anyways, at least not in a professional context. That means, it’s very good that you try to educate, but you might as well elaborate about when and why it is not recommended (instead of just listing the issues of implementing a parser, which does not really inform users imo.).

For example, we use Collada extensively in our content pipeline and tools (written in AS3) at work and it made a lot of things a lot easier! We don’t want to miss it! It can be used to create complex scenes (with embedded meta-data and placeholders, if necessary) using different tools operating on the same data, like 3DS Max and Photoshop. As mentioned above, it is the perfect format to support in our own tools, which use the same run-time as the final products for visualization.
Once a pipeline is set, it’s easy for the artist to use it, which would not be possible without Collada, as it is now.

Yet I want to let you know, that I am very much aware of the issues most of the exporter implementations create for parser authors due to slight variations in the document structure; I have some (mostly bad) experience with that myself.

But after all, I would really like to see support for Collada return to Away3D 4!
I can imagine you could support an explicit (and well documented!) set of features and structures that pipeline creators need to use, so we can decide what plug-ins and settings are the best for our needs (e.g. use the Collada refinery to achieve compatibility). It might make it easier for you to support Collada that way. Although I have to admit, that I haven’t thought this through beyond the general idea.

Best regards.

@TheSillyOne:
I’m sorry if I come out harsh against COLLADA. I believe I have time and time again explained in different threads on these forums and on the old mailing lists why COLLADA is not a great format to be used in those areas where the Flash Player (and Away3D) is generally used, i.e. web applications and games. Because of this, and the effort required to implement it, we focus on other things.

You have to understand that we are independent developers, with no budget, creating this product for you for free, because we enjoy doing so. We listen carefully to our community, and appreciate all comments, but in the end we have to decide what we want to spend our time on, because our time is finite. Usually it comes down to weighing the factors of effort, benefit and “fun” or “interest”. Quick bug fixes require minimal effort but provide great benefit. A new animation system might require a very large effort, but the benefit might be so large that it justifies the effort.

Personally I think that a new file format (AWD2) and a flexible loading framework are things that are fun to produce and benefit the community a great deal, which is why I have focused most of the couple of thousand hours that I’ve put into Away3D 4.0 into creating these particular features.

Using the very same formula, David for example thinks that shaders are a great combination of fun, benefit for the community and reasonable effort, which is why he has invested a lot of his time into an absolutely awesome materials system. Fabrice thinks the same of tools and mesh generation utilities.

What I’m trying to say is that if we are to prioritize our tasks (of which there are many) we will focus on improving things like physics integration, game tooling, unified animation systems and a lot of other things, before we get around to creating a COLLADA parser. Simply because it is our opinion that a COLLADA parser requires effort that is disproportionate to the amount of benefit and “fun” that it provides. Why do we think so? Well, because it’s not an ideal format, and it can only be used in tooling and development (and is not suitable for online deployment) which is not where most of our users use Away3D. By this I’m not saying that COLLADA is a completely useless format. Obviously it has a lot of uses. Most of them are just not in the fields where Away3D is used.

If you personally feel that a COLLADA parser is something that benefits you and the community a lot, and you think it’s fun and a sensible amount of effort to implement one, then you’ll be happy to know that all parsers in Away3D are plug-ins, and you can easily create your own. If it’s good, and you would like to contribute it, then we would gladly incorporate it. Because as I said, we don’t have anything principally against having a COLLADA parser in Away4; we just do not have the time to create both that and all the other things that are on our list.

On the other hand, if you don’t think it’s fun or a reasonable amount of effort, but it’s crucial for your company that there is a COLLADA parser in Away4, then you are also very welcome to hire someone (e.g. one of us in the Away3D team) to implement it for you. That is the only way for us to free up enough time to get it done in a fixed timeframe.

I hope that you’ll appreciate the situation that we are in, and I’m sorry if it sounded like I have some sort of prejudiced opinion against COLLADA. I have read the 300 page specification and I have formed (IMO) a rather educated opinion about it. That opinion is that while COLLADA has it’s uses, I’m not convinced it’s more important for Away3D than finishing the AWD2 format specification and tools, improving the animation’s framework, fixing a bunch of bugs and building some rather exciting tools that you will soon be hearing about. So unless you can convince me I’m wrong, that’s what I will be spending my nights and weekends on over the next couple of weeks and months.

Please contact me offline if you want to contribute a COLLADA parser or hire one of the team members to do it!

Cheers
/R

Thanks a lot for the response!

Well, I am perfectly aware of all the very valid points you made! My point was just this: “Collada has its uses and there is nothing else quite like it for those uses yet.”, and really nothing more, no pushing you to put back Collada support into Away3D 4 at all. Your work is very much appreciated, do not worry!

I understand why it’s more important to concentrate on other things to implement (yet the “fun” factor is irrelevant to me personally, as much as it helps to get things done) and you are absolutely right! These things *are* more important than a Collada parser! I was just saying that I’d like to see it back eventually.

On the other hand, I’m looking forward to the new AWD2 format, which would indeed be a lot more important and if done right, could replace Collada for us, at least in the latter parts of our pipeline! What would really make or break it would be exporters for the most common content creation tools (we use 3DS Max 201x here), especially for animation.

One more thing to wrap this up: You could have a link to a post that you can paste or a sticky thread regarding Collada support in Away3D and what its intended use is so (especially new) users can be informed more easily.

And yes: you guys rock! Keep up the good work!

Best regards.

Silly, here’s an idea. Why not write a Collada parser for Away3D 4.0 and share it with the community? You seem to be very familiar with the format and as you mentioned, you already have AS3 apps that read it?

Just an idea

I agree with both of you. Collada is a very useful exchange format, not really intended for runtime applications. However, in some applications, for example a browser based 3D editor of any kind, or one that allows users to upload 3D content, or for database generated 3D, it is sadly missed in Away3D 4.0.

An XML based language is always going to be useful on the web, imho.

theMightyAtom - 23 September 2011 02:09 PM

Silly, here’s an idea. Why not write a Collada parser for Away3D 4.0 and share it with the community? You seem to be very familiar with the format and as you mentioned, you already have AS3 apps that read it?

Just an idea

Ha, yeah, I wish I had the time, but I have to write a complex framework, set up a pipeline and integrate some existing technology for my company, so there’s no time for that. I have attempted to write a Collada parser in C++ in the past, so I know what I am up against. Fortunately, there’s Assimp.

Best regards.

@theSillyOne:

can you tell me if latest Assimp version can convert collada to md5 ?

in the Assimp-viewer for windows i can import my collada model and it shows up correct.

right now i am installing visual studio to check out the complete (sdk) version of Assimp.

on their website they say it works as a converter too.

ive got not much knowedge about visualstudio and c++, so if it will not convert collada to md5, you would save me propably lots of time by telling me.

thanx

80prozent - 23 September 2011 04:30 PM

@theSillyOne:

can you tell me if latest Assimp version can convert collada to md5 ?

Unfortunately, that’s something I haven’t tried, cannot try now and thus, just don’t know.

There is a documentation and there are some good tutorials for Assimp, so I hope you’ll get your answer as fast and painlessly (that is actually a real word) as possible.

Best regards.

thanx for answer

i had a closer look at assimp and decided that for my needs its far to much new stuff to learn right now….maybe another time when i need more of its features than one conversion. for now i keep focus on as3 and hope i can figure another way to convert between those formats.

I decided a while back to try to learn the full Collada specification. WOW is it complex! It became very clear to me very quick how huge a task writing a parser for it would be. The complexity of it is also the reason different parsers often behave differently. A huge mess

Richard’s AWD format is going to be a much better replacement. Check out the WIP spec on it here

http://code.google.com/p/awd/downloads/list

It’s what Collada should have been from the beginning

Still focussing on MD5 and 3DS support does seem to favour the use one creation application over the other.

Being a softimage user and animator myself 3ds and especially md5 is not ideal as of yet. It also seems as if autodesk is favouring going around rather then through the middle. Meaning, using crosswalk, dotxsi, fbx and collada to move from soft to maya and max and vice versa. This might also go for maya and the nonautodesk houdini and blender?

But then indeed neither is collada it would seem. Though it is open where 3ds, sadly is not.

How about fbx support? (or even dotxsi, sorry couldn’t resist;))

All in all it seems there are no sturdy pipelines in sight as of yet. Unity is starting to look promising in this respect.

Thanks for the amazing work on away3d.

Proscenium might fill in the gap here, it has a Collada parser. Although I haven’t tested it, yet.

We’re currently using MD5 and Wavefront Object for prototyping.

Best regards.