more WiscWorlds

Okay, I finally got it to work!  Your app looks like it has a lot of
potential.  I know from reading some of your blog entries that your
team is aimed at USING the Croquet environment within your educational
setting, rather than just developing the SDK itself.

Here are a few observations:

1.  All of the spaces are pretty empty.  Is there something else here
that's not obvious?
2.  I created a few objects by various means in the menus, and then
quit Croquet and went back in.  All of these objects were gone.  Is
that by design?  How could I make them stick around?  Does this deal
with the capabilities of the "everyone" logon ID?
3.  When a new text window is created and has focus so that the text
can be edited, (on Windows) just pressing any of the modifier keys
(ctrl, shift, alt) by itself without pressing any other key causes
things to happen.  Text gets selected, the halo comes up on the text
object, etc.  This seems to be a bug.
4.  Using some of the menu items causes the Croquet window to lose
focus because, for example, the TSpace browser comes up in the Squeak
world outside of Croquet.
5.  You should include the simple fix that Mark McCahill demonstrated
in one of his movies to take the MessageTally out of the TPainter
"keep" action.  This also causes the Croquet window to lose focus,
although I'm sure that it would be nice for someone to tune up or
plugin-ize the inflation subdivision code.  I saw a tuning device
being used by a bagpipe band last weekend - would that help?
6.  I got the webcam support to work - well, kind of.  The webcam view
in the window over my avatar's head (looking in a mirror) was there
but not correct.  It looked like there was a size difference between
what the camera supplied and what the display expected such that the
camera image was at least partially replicated about 6 times
horizontally across the display.  The vertical was also off but
different.
7.  You are using changesets and an update stream process to maintain
this app.  Does that mean that you are not in a hurry to provide any
of the components, like the RFB/VNC stuff, to the straight SDK-using
community?  What is your view, in 25 words or less, of the update
stream vs. Monticello package repository as it relates to the SDK
community?

There are several more questions that I have, but this is a good starting set.

Thanks for any insight you can provide, and keep up the good work!

On Jul 9, 2006, at 8:04 AM, David Faught wrote:

Both 1 and 2 are symptoms of the continuity server being down.

If there's nothing connected to the router to supply the definition  
of the current world state when you join, then the software on your  
machine will build the world itself. [Sure, it should tell you. But  
then again, it shouldn't be the case that there is no continuity  
server. Besides, it DOES tell you if you're running with Transcript  
open.]

The hardcoded scripts to build the worlds create pretty empty spaces  
(just a terrain and a portal to the commons space) -- except for the  
the initial commons world and the demo world, which is the intro  
space for the full "BFD" demo, including the portals to Mars, 3Body,  
Underwater, etc.

Note: having no cserver is an "undefined" situation that WiscWorlds  
just doesn't handle gracefully. For example, if you connected to the  
router for a world, built your own, and then added something, and  
then someone else joins that router, sees no cserver, and builds  
their own island, too, they won't have the object you created. When  
you try to do something with that object: BOOM.

I've reset the machines. However, at the moment, the cserver is a box  
that can't handle two of the demo WiscWorlds that we use for stuff on  
campus. So they're just empty right now.
   - One is a virtual art studio and gallery called Critiquet. It  
normally includes some empty studios in a hallway outside the  
"courtyard", and one studio that has influences, snapshots of works-
in-progress, and a finished 3d artwork.
   - The other is a "Crime Scene Investigation" exercise called  
Forensics. It has a car, house with furnishings, woods, lake, two  
dead bodies, and some clues.
I'll try to reset again tomorrow morning with the "right" stuff.

> 3.  When a new text window is created and has focus so that the text
> can be edited, (on Windows) just pressing any of the modifier keys
> (ctrl, shift, alt) by itself without pressing any other key causes
> things to happen.  Text gets selected, the halo comes up on the text
> object, etc.  This seems to be a bug.

yup. Wanna fix it?

> 4.  Using some of the menu items causes the Croquet window to lose
> focus because, for example, the TSpace browser comes up in the Squeak
> world outside of Croquet.

Yup. Don't do that.

> 5.  You should include the simple fix that Mark McCahill demonstrated
> in one of his movies to take the MessageTally out of the TPainter
> "keep" action.  This also causes the Croquet window to lose focus,
> although I'm sure that it would be nice for someone to tune up or
> plugin-ize the inflation subdivision code.  I saw a tuning device
> being used by a bagpipe band last weekend - would that help?

heh. Yeah, the next version should include all of Mark's goodies.

> 6.  I got the webcam support to work - well, kind of.  The webcam view
> in the window over my avatar's head (looking in a mirror) was there
> but not correct.  It looked like there was a size difference between
> what the camera supplied and what the display expected such that the
> camera image was at least partially replicated about 6 times
> horizontally across the display.  The vertical was also off but
> different.

That's what it looked like for me, too, prior to 088-
CameraSpecificFormSize.
If you've got all the updates and still have this problem, then  
there's a bug.  See if you can fix it! (Hey, this is the dev list!)


> 7.  You are using changesets and an update stream process to maintain
> this app.  Does that mean that you are not in a hurry to provide any
> of the components, like the RFB/VNC stuff, to the straight SDK-using
> community?

Mark McCahill has volunteered to do that, but see below. He doesn't  
want more stuff added to the repository, but he has all the code.

Then again, so do you!  Also, the raw VNC stuff, including the server-
side, is at http://atsosxdev.doit.wisc.edu/croquet/Xrfb/

>   What is your view, in 25 words or less, of the update
> stream vs. Monticello package repository as it relates to the SDK
> community?

No organization (including Mark's U.Minnesota) is being paid to  
organize code for other folks to use -- and that's a huge amount of  
work. So it always falls off the bottom of the ToDo list. I think  
this should and will be an issue for the Croquet Consortium once it  
gets activated.


>
> There are several more questions that I have, but this is a good  
> starting set.
>
> Thanks for any insight you can provide, and keep up the good work!
>

On 7/9/06, Howard Stearns <[hidden email]> wrote:
> No organization (including Mark's U.Minnesota) is being paid to
> organize code for other folks to use -- and that's a huge amount of
> work. So it always falls off the bottom of the ToDo list. I think
> this should and will be an issue for the Croquet Consortium once it
> gets activated.

Sadly I am not getting paid at all to play with Croquet.  However, in
my opinion Howard's remark is not entirely true.

It may be that someone has enough enlightened self-interest to start
(a) packaging their own code or (b) produce a generic packaging system
(similar to the 'save to file' option in a Squeakland eToy.

I say enlightened self-interest because

(a) packaged code may lead to more testers and more bug fixes and
contributions for your own code.

(b) more automated packaging may make it easier for you to test
others' contributed worlds / models etc. - so you can incorporate them
in your own research/teaching etc.

All that said, I agree with Howard that support for packaging would be
a great use of consortium funds (or of Department of Education /
taxpayer money.)

David

On 7/9/06, Howard Stearns <[hidden email]> wrote:
> On Jul 9, 2006, at 8:04 AM, David Faught wrote:
> > 7.  You are using changesets and an update stream process to maintain
> > this app.  Does that mean that you are not in a hurry to provide any
> > of the components, like the RFB/VNC stuff, to the straight SDK-using
> > community?
>
> Mark McCahill has volunteered to do that, but see below. He doesn't
> want more stuff added to the repository, but he has all the code.

I'm confused by this statement.  Mark has graciously made a repository
available for public contributions, why would he not want more stuff
added to the development team (Contributions?) repository?

> No organization (including Mark's U.Minnesota) is being paid to
> organize code for other folks to use -- and that's a huge amount of
> work. So it always falls off the bottom of the ToDo list. I think
> this should and will be an issue for the Croquet Consortium once it
> gets activated.

At least some ARE being paid for Croquet-related work.  That's a
start!  Any news lately of the Croquet Consortium kindling?