How do we build a more coherent community around Cobalt?

On Wed, 2008-03-05 at 13:14 -0800, Edwin Castro wrote:

> Anyway, I hope to build something like that. I'm in learn mode right
> now though and waiting for Croquet to mature a little more... perhaps
> it's time to find others who might find such a project fun...
Sounds like a great idea! I'd like to see that one happen as well, just
for fun. Ric

--
================================================
My father, Victor Moore (Vic) used to say:
"There are two Great Sins in the world...
..the Sin of Ignorance, and the Sin of Stupidity.
Only the former may be overcome." R.I.P. Dad.
Linux user# 44256 Sign up at: http://counter.li.org/
http://www.sourceforge.net/projects/oar
http://www.wayward4now.net  <---down4now too
================================================

Thanks to everyone for their input on this. It looks like, for the short term, we will be putting together a portal site that will contain the following elements:

- A community-editable wiki, probably based on Mediawiki

- Discussion forums, probably based on JForum

- Information on how to connect and participate with the IRC channel

- A Jabber chat server (details TBD)

Clearly this is all very preliminary and will need to be fleshed out more as we start putting this infrastructure together.

Long term, the goal is pretty clear that people want to use Cobalt as the community collaboration tool.

Hi Matt,

I'm really happy to see that this action with Cobalt is really happening as we had previous conversation about something like it.

The technical details about the PreCobalt communication looks like got to the conclusion and I'm looking forward to see the portal and have all those channels for communicate.

As I see there are some of you who actually put Cobalt together and others who keen to contribute but have less or minimal knowledge about squeak, programming etc.

It's one of my goal to learn from you guys programming, developing Cobalt to power our developer and user network.

I think it is a great opportunity doing the Cobalt development in a training process that nurtures the new users and developers.

I think forming WorkGoups, figuring out who can lead and teach others, who wants to learn etc. are the first steps.

I feel by doing this will build a coherent community around Cobalt.


Thanks for your activity!

Attila



    

 



  


 

Thanks to everyone for their input on this. It looks like, for the short term, we will be putting together a portal site that will contain the following elements:


- A community-editable wiki, probably based on Mediawiki

- Discussion forums, probably based on JForum

- Information on how to connect and participate with the IRC channel

- A Jabber chat server (details TBD)

Clearly this is all very preliminary and will need to be fleshed out more as we start putting this infrastructure together.

Long term, the goal is pretty clear that people want to use Cobalt as the community collaboration tool.

---

Climb to the top of the charts! Play the word scramble challenge with star power. Play now!

Janet,

I am completely agree with Janet.

This is an unofficial Croquet support effort. http://croqueteers.com/

It is in the early stages. Everyone is wellcome.

________________
Jose L. Salmeron
http://www.upo.es/eps/salmeron

Hi Janet,

>
>> I had a lot of
>> trouble finding examples worlds beyond "I am a rabbit being driven
>> over a field"
>
>
I've was frustrated by this too, (& I hope I don't sound negative
either)  being excited about the technology of OpenCroquet, but always
seeing the same back-of-a-rabbit's-head examples (the third-person
perspective is annoying)*. There's very little information out there,
and if it wasn't for Smalltalk's largely self-documenting code, it would
even more difficult to understand how to build anything with Croquet.
Nevertheless, there are people out there building things with
OpenCroquet (I'm one of them), and when they have built things worth
sharing, the pace of development will quicken. What we will need though
is a way to connect our worlds together, preferably via OpenCroquet
itself, and across future versions of OpenCroquet (without too many
MessageNotUnderstoods!).

Another thing. Many many programmers will come to OC without any
knowledge of Smalltalk, let alone Squeak. These people are unaware of
the history & motivations being Morphic and Tweak, and they won't know
the cultural side of programming in Smalltalk or best practices. Most
people will come from a Java or C++ background and will need to leave
much baggage at the door. It would be great if there was a forum as
Janet suggests, mailing list or chat server, or Croquet World, where
newbies can ask stupid questions without fear of being laughed at (I'm
not saying that happens here). I've been thinking about making a world
which somehow visualizes the classes of the Squeak image. If this was
itself a Croquet World, then it would have dynamic views of the  Squeak
image it's running in, something like a live class browser world that
shows all the relationships between classes, so that newbies can wander
around the classes (a simple network of labeled boxes would do to begin
with), point to them and ask questions of each other: "what does  this
TMessageRouter do ?" etc, and people could leave annotations and
comments (and postcards to examples). There could be CroquetClassWorld,
SqueakBasicsWorld, SmalltalkForNewbiesWorld.


* This brings-up a related point which I think DAS blogged about once:
while people familiar with OpenCroquet know that the primitive graphics
are beside the point, people who don't know about OC see the
cardboard-box rabbit examples and think that's all there is to it. Of
course, the architects of OC didn't have time for art because they were
busy making the technology, but unfortunately the fickle masses reward
shallow good-looking things over deep plain things.

Hi Guys,

On Wed, Mar 5, 2008 at 3:33 PM, Matthew Schmidt
<[hidden email]> wrote:
I agree with Darius' comment that there is a danger to creating
copious documentation that quickly becomes out of date, but I do think
the optimum is a bit more than we currently have.  With that in mind,
let me answer the question posed above.

What I want is:

 - A place for new folks to enter the community.  Eg, a couple of
stable web portals to serve as a documentation anchor for people
coming into the community.  I think this is largely done, and with
time as things stabilize more web documentation will just happen.

 - A stable out-of-world location for people to get questions answered
and to develop a sense of community.  At some point it would largely
become volunteers helping with installs, once the world is stable
enough to support in-world community.  IRC  seems to be a reasonable
option here.

 - A stable in-world location for people to develop a sense of community.
   * That place would be expected to break from time to time in the
early stages.
   * It would likely lag behind other development spaces in code
deployment, favoring
      stability over features.
   * people could verify their install went well
   * people could read in-world documentation, and see in-world
examples of portals,
     documentation, browsers, avatars and so on
   * people could use the data sets as a reference implementation.

My interest is largely in using this educationally, but  social
interaction is a core part of human nature and this will be used
socially.  I welcome the social aspect of this, and think at some
point in the future we have to ponder how to capitalize on that.

Janet

Hi Edwin,

On Wed, Mar 5, 2008 at 4:14 PM, Edwin Castro
<[hidden email]> wrote:
>
> When I first encountered croquet I envisioned a space where I could meet my
> role-playing buddies and play table-top role-playing games in the virtual
> world.

  I had similiar thoughts.  I have a vision for what I want, but I
cannot realize it on my own, so I join communities heading in the
right direction.  My initial goal is to simply have open protocols
to allow interchange amongst people and spaces.  The rest can be
layered on top of a sufficiently thought-out protocol.

  I envision a single application that is both client and server, when
you install it you select a virtual house on a virtual plot of land,
if you have access to a large server you can place your house in a
virtual city on that server.  In your front yard you have something
that looks like a stargate, which serves as both inbound and outbound
portal to other servers.  If the protocol is robust, the same
interface can take you to a system doing massive multiplayer gaming
like Eve, Warcraft, SecondLife, etc as well as let you have a few
friends connect to your server and play virtual monopoly on your
virtual dining room table.  Why should every game have it's own
engine?  If the engine is well thought out, and you are willing to
throw enough CPU at it, you should be able to have a generic world
server implement most any 3D based environment, be it a MMO game, a
scene-graph walk through, a learning environment, something like Alice
(3D environment for teaching programming), SecondLife or a virtual
board game in a virtual kitchen.

The system could have a backend database for crypto-keys, PKI (public
keys), Kerberos tickets, local authz databases and so on.  The backend
database could have a set of world common objects crypto-signed by the
default world.  All other objects are signed by their respective
authors and you can accept them or not as your world policy dictates.
Every world builder that wanted to author unique objects would either
need to hand them to the world server for signing, or get a unique
world ID assigned.  128 bit world IDs are not too burdensome in a
sparse space where folks either accept all, accept none, or accept
according to policies.  For example one could use the PGP trust
paradigm and accept world IDs you have personally swapped signed keys
with, that well known entities have signed keys for, or that friends
of friends have swapped signed keys with.  By requiring crypto-signing
and persistent identity, you acquire faith that the rabbit in front of
you claiming to be Julian is really Julian.  If your kids are playing
in this space, and the space has reasonable policy, then it becomes
challenging for predators to enter the space.  If you play shoot-em-up
games, your policy would be possibly more permissive, or possibly
friends only, depending on what you are looking for.  The problem of
grief-ing in the game space is also solved by persistent identity.  If
someone wants to build a MMO where they are god and everyone dies at
their whim, they are free to do so, but nobody will really want to
join it.  Persistent IDs allow stable virtual communities to form,
with faith that the person in front of you is who they say they are.
 They also allow folks who require anonymity to have it, but still be
a persistent person.  Virtual alcoholic anonymous meetings, virtual
rape survivor meetings and so on, all have a need for persistent ID
not traceable to a real person.

The world could work with existing 3D meshes, perhaps openGL or SDL
(Simple DirectMedia Layer, an open DirectX implementation).  All
meshes are signed by someone and that someone is persistent in this
space.

Anyhow, those layers can all be placed on top of an existing protocol.
 Having something working first will allow people to learn and get
experience, and from that experience more advanced systems will
emerge.  For now, it's best to get something running, and allow the
more advanced stuff to follow later.

There is much to do,

Janet

On Wed, Mar 5, 2008 at 4:28 PM, Peter Moore <[hidden email]> wrote:
>
> It doesn't seem like my Jabber idea has caught fire. Does it seem too
> inelegant? Too much work?
>

I like Jabber and IRC, and would use whatever folks prefer.

Janet

Peter,

On Wed, Mar 5, 2008 at 2:21 PM, Peter Moore <[hidden email]> wrote:
>
>   I propose that we use Jabber (XMPP) as our main communication channel.

OK, works for me.

> The advantage to using Jabber vs. IRC is that there is already some support for
> Jabber in Cobalt. It's currently possible for someone running Adium or
> Pidgin (eventually gmail's chat) to communicate with people inside of any
> space.

That is a powerful incentive to go this route.

> So rather than having a single persistent space we can create ad-hoc
> meeting spaces using Jabber to send Croquet "postcards" as invitations.
>

Ad hoc is useful for lots of things, but I still favor a stable
location.  When I was first
starting I wanted to connect to some place, and say hello to a human who would
say hello back.

Janet

I second the suggestions about the collaborative portals (community
wikis) and in-world meeting places (notably, the "hello networked
world" default place).

To me, what could bring a lot of community development (since we are
talking of an early stage browser) would be support for standard web
XMLRPC integration, so that the tons of public web APIs out there
could be used in-world (i'm thinking of freebase-like services,
mapping, translation and media APIs) for visualization purposes.

Also, support for external cross-platform projects (ex: GStreamer
framework for media processing or conferencing, webkit for true
standard-compliant web rendering) like it's been done with ODE and
data repositories (ex: 3D database) could both lower the huge
maintenance and dev work by the croquet/cobalt core team and unleash
the community effort.

Scripting support has been mentioned as well, and i guess that a
Python-like language would lower the entry barrier for development.

In short, i'm more or less inclined to think that given the state of
the project, the initial community could well be made of developers at
first. Not to say that there is no Croquet developers community, but
that a huge part of willing developers are kind of frightened by
learning a new language and having to write a lot from scratch
again...

What do you think ?

FLo

well I suppose I should chime in.

Every good community needs a T shirt and hat :) 

a placeholder for the idea:

http://i273.photobucket.com/albums/jj228/waufrepi/CobaltT.jpg

then perhaps donate the profits to the OLPC foundation. 

           wfpi

> I have just one concern regarding a possible large push in this direction.
> With the code subject to many changes, debugging, and experimentation to
> make it usable, it will likely break compatibility with any and all
> tutorials, documentation, code comments, art, avatars, buildings, meshes,
> textures, etc.
>  
You need documentation of what's there now, to attract enough people to
the project. With enough people attracted to the project there will be
more people to support documenting efforts.
More people also means more development, and more sharing of code. We
can't keep waiting for companies developing commercial croquet products
to share some of their code.

For what it's worth, I'll try to keep croqueteers.com updated as time
allows, and some other people have already been contributing. I'm also
planning to put up a forum with different sections for newbies and pro's.

croqueteers.com will move to a beefier server in the future, and when
that happens it could in theory support running croquet worlds etc,...
If any interesting ideas come out of this discussion I'm willing to put
that server at disposal of the community to implement them.
I'm the server admin for that server, but the owner is a friend of mine
who is actually paying for it. He was willing to let me use it and the
future
server to help develop a croquet community.

waufrepi III wrote: I like the T-shirt idea :) Things like that help develop a sense of
community.

I have just created a brand new Croquet forum on croqueteers.com
direct link: http://croqueteers.com/forum/

Please, suggest ideas, use it, collaborate, be a community!

(Not that I think these lists are bad, but as mentioned before in the
community thread here on this list, good forums and a wiki would be
essential tools in building a community. I hope I'm not upsetting any
one by pushing forward with all this so aggressively, I just really want
to get a good community going with tools that are more accessible to the
common people. Remember any of you can suggest/change any thing for both
the wiki and the forums. I want to do this as open as  humanly possible.)

Maybe this should be posted on the croquet-user list as well, since
users can request help there, but I'll leave that to someone else, I
don't want to spam the lists with my stuff.

The forum sections and boards are just a draft really, what I could come
up with as I did the initial install. If any of you feel like there's
something there which shouldn't be there, or if there's something not
there which should be there, let me know and I'll make the modifications.

-John Sennesael

John Sennesael wrote:
> (Not that I think these lists are bad, but as mentioned before in the
> community thread here on this list, good forums and a wiki would be
> essential tools in building a community. I hope I'm not upsetting any
> one by pushing forward with all this so aggressively, I just really want
> to get a good community going with tools that are more accessible to the
> common people. Remember any of you can suggest/change any thing for both
> the wiki and the forums. I want to do this as open as  humanly possible.)

Go for it. The only way things get done is when people do them ;-) You
might want to consider a bridge between the two though. For one thing,
it would make it easy for people like myself who are intrinsically
email-centric to help answering the questions in the forum. For another,
it would mean that the forum looks a little bit more lively than it does
right now ;-)

> Maybe this should be posted on the croquet-user list as well, since
> users can request help there, but I'll leave that to someone else, I
> don't want to spam the lists with my stuff.

Why not? I think it's perfectly appropriate.

Cheers,
   - Andreas

> Go for it. The only way things get done is when people do them ;-) You
> might want to consider a bridge between the two though. For one thing,
> it would make it easy for people like myself who are intrinsically
> email-centric to help answering the questions in the forum. For
> another, it would mean that the forum looks a little bit more lively
> than it does right now ;-)
It already has an option to get email notifications when someone replies
to one of your threads, however it currently can not let you reply to
these messages.
I'll have to see if I can create or find a forum mod which will do just
that. I agree that a bridge would be helpful. It doesn't look lively
right now because it was just created, about an hour ago ;)
It will probably take some time.
>
>> Maybe this should be posted on the croquet-user list as well, since
>> users can request help there, but I'll leave that to someone else, I
>> don't want to spam the lists with my stuff.
>
> Why not? I think it's perfectly appropriate.
I'll wait a day or two, and see if other people reply or suggest things
here first.

Ps: Thanks for the response, it is encouraging :)

-John Sennesael