VRML breaks collaboration?

Hello, list!

Among the various meanderings I've had within Croquet lately, I've been working with VRML.  I wonder if anyone can comment on the following?

By simply adding a VRML model to any of the demo worlds (e.g., KAT, SimpleDemo, MPEG Demo, Sailing, etc.) and then remotely connecting to said demo world, the world comes to a halt.  In particular, it turns red and a progress bar opens up stating,

"Checkpoint failure - searching image references..."

Naturally, the progress bar doesn't ever make it to the end.  Is there a fix for this?

There are good reasons why an environment such as Croquet should offer strong support for VRML and, ultimately, X3D.  Is it simply the case that Croquet's level of ability in this area is alpha quality?  If so, are there any plans in motion to address this?

Thanks,

-RC

Sounds like a bug. [Somehow a piece of the on-island model (e.g., a  
vertex list or some such) is being shared and held onto by something  
off-island (some part of the importer). That's a no-no, and results  
in a failure when a snapshot of the island is made for the new  
participant.]

You can report it to http://opencroquet.org/index.php/Bug_Reporting 
with the steps to reproduce, including a model that reproduces the  
problem.

I'm not sure how to answer the "alpha quality" question. I haven't  
run across non-alpha quality VRML data transfer in any system, so I  
tend to use other formats. (ASE, 3DS, OBJ, DAE,...)  But that's just  
my (lack of) experience.  My own personal view (which others will  
quite fairly disagree with), is that at this stage in the evolution  
of a technology, I concentrate only on making the stuff that I need  
work. I'll generally use a different way to get something done (e.g.,  
by transforming serialization formats in programs that are designed  
for digital content creation and exchange), rather than getting hung  
up on something that isn't mission critical to my immediate project's  
success.  If there are folks listening who find VRML to be a critical  
part of their workflow, then they'll be quite interested to get a  
reproducible test case.

-Howard

On Oct 25, 2007, at 10:43 AM, RC wrote:

It's an open source *community* project, with the usual expectations.  Besides reporting the bug, it's certainly understood that those who observe the bug are welcome to develop a fix and submit it, or others who want to contribute and learn are equally welcome to work on such bugs.   The no-no Howard cites is a helpful clue for that process - it's part of the limits of the "island" technology at the moment, though it may not be well-documented.

Howard Stearns wrote:

Sounds like a bug. [Somehow a piece of the on-island model (e.g., a vertex list or some such) is being shared and held onto by something off-island (some part of the importer). That's a no-no, and results in a failure when a snapshot of the island is made for the new participant.]

You can report it to http://opencroquet.org/index.php/Bug_Reporting with the steps to reproduce, including a model that reproduces the problem.

I'm not sure how to answer the "alpha quality" question. I haven't run across non-alpha quality VRML data transfer in any system, so I tend to use other formats. (ASE, 3DS, OBJ, DAE,...)  But that's just my (lack of) experience.  My own personal view (which others will quite fairly disagree with), is that at this stage in the evolution of a technology, I concentrate only on making the stuff that I need work. I'll generally use a different way to get something done (e.g., by transforming serialization formats in programs that are designed for digital content creation and exchange), rather than getting hung up on something that isn't mission critical to my immediate project's success.  If there are folks listening who find VRML to be a critical part of their workflow, then they'll be quite interested to get a reproducible test case.

-Howard

On Oct 25, 2007, at 10:43 AM, RC wrote:

Hello, list!

Among the various meanderings I've had within Croquet lately, I've been
working with VRML.  I wonder if anyone can comment on the following?

By simply adding a VRML model to any of the demo worlds (e.g., KAT,
SimpleDemo, MPEG Demo, Sailing, etc.) and then remotely connecting to said
demo world, the world comes to a halt.  In particular, it turns red and a
progress bar opens up stating,

"Checkpoint failure - searching image references..."

Naturally, the progress bar doesn't ever make it to the end.  Is there a fix
for this?

There are good reasons why an environment such as Croquet should offer
strong support for VRML and, ultimately, X3D.  Is it simply the case that
Croquet's level of ability in this area is alpha quality?  If so, are there
any plans in motion to address this?

Thanks,

-RC
-- 
View this message in context: http://www.nabble.com/VRML-breaks-collaboration--tf4691613.html#a13409188
Sent from the Croquet - Dev mailing list archive at Nabble.com.