Re: [Instances|Group|Model] Thoughts

[Vince Fortin <vfortin(at)gmail.com>]

Guy Guy Guy

All I can say is Patrick does a great job correcting incongruities!

On 11/30/05, Wayne Williams <w1343(at)comcast.net> wrote:

I think it'd be great to see you posting some of your musings on XSI. I
don't think anyone on this list has much trouble understanding your English
either....I don't anyways. With the community as small as it is still,
having great minds all grouped together discussing their thoughts on the
inner workings of XSI is really beneficial to the userbase as a whole imo.
There are those of us who aren't as technically savvy as the bloggers and we
plebes stand to learn a great deal from their contributions.....so hopefully
we'll see you posting a little something as your schedule/desire permits :D

----- Original Message -----
From: "Guy Rabiller" < guy(at)alamaison.fr>
To: <XSI(at)Softimage.COM>
Sent: Wednesday, November 30, 2005 12:49 PM
Subject: Re: [Instances|Group|Model] Thoughts


>
> Hey Francois,
>
> that's funny because I never read the emails in threaded view, and
> following your message I've switched to threaded view and realized that
> it was deep under the started thread "montreal's snow". Sorry about
> that, I should have started a new thread.
>
> Regarding XSI-Blog, I dont feel having the soul of an article writer,
> especialy with my english level, and if I remember correctly you have to
> post often and regularly with interresting topics, unfortunately most of
> my posts here are whinnings and I'm affraid that it would not be
> appropriate there.
>
> That said I really like this blog, I read it with great interrest,
> especialy Luc-Eric's posts, trying to decipher his state of mind, a
> great mystery to me :-)
>
> I don't know, may be I should post there anyway, who knows, we might
> find a better communication channel by deciphering each others ;-)
>
> --
> guy rabiller | 3d technical director (at) LaMaison
>
>
>
> Francois Lord wrote:
>
>> Guy, have you thought about blogging on xsi-blog? This is exactly the
>> kind of material that people want to have easily accessible. I'm saying
>> this because your post ended up in the middle of a thread that was in
>> the middle of another thread (which was not too serious in the first
>> place), so it might not have gotten the attention it deserved.
>> (shameless plug for Patrick's site)
>>
>>> Btw, regarding Instances, Models and their Instance Groups, when you
>>> take a closer look at what's going on when you deal with Instances:
>>>
>>>
>>> 1) By instancing a Model, an 'InstanceGroup' is created under the Model.
>>> This is a special kind of a Group, because it has only one Parameter (
>>> as for the "Associated Models" Group of a Light ), but despite this fact
>>> it is a class of siGroupID, and of Type "#Group", and nothing, afaik,
>>> distinguishes it from a regular Group, even from the API point of view.
>>>
>>> -> Suggestion: Add to InstanceGroup some Property, or subclass its
>>> Type ?
>>>
>>>
>>> 2) The only thing which is done by instanciating a Model wich cannot be
>>> reproducing manualy by the user is the creation of this special
>>> InstanceGroup. From here, you can manualy create an empty Model ( Model
>>> -> Create -> Model -> New Model ), select it, then right click to the
>>> InstanceGroup of one instanciated Model and -> Add To Group. Now our
>>> empty Model is an Instance and it as been automagicaly recognized as
>>> such, and it works. But why then, can't we drag&drop Models into
>>> InstanceGroups ? We have to right click on an InstanceGroup -> Add To
>>> Group.
>>>
>>> -> Suggestion: Allow drag&drop of Models onto InstanceGroups ?
>>>
>>>
>>> 3) A side effect of the 2) observation is that you can add to an
>>> InstanceGroup a non-empty Model. Lets say you have a Model wich contain
>>> a Cube, and you add it to the InstanceGroup of another Model. It works,
>>> you now can see both the Instanciated object and the original Cube at
>>> the same time. Cool to add some instanciated objects to an existing
>>> Model. But for this workflow to be efficient, we would need to be able
>>> to create this famous InstanceGroup on a Model without actualy
>>> instanciate the Model itself by creating other Models.
>>>
>>> -> Suggestion: Add a Command "Add InstanceGroup" to the contextual menu
>>> of a Model ?
>>>
>>>
>>> 4) An Instance can be part of several InstanceGroups ! You can add a
>>> empty ( or not ) Model to the InstanceGroup of several instanciated
>>> Models. Unfortunately, instance are not combined, you allways see the
>>> instance of the first Model you added to its InstanceGroup.
>>>
>>> -> Suggestion: Combine Model Instances taken from several
>>> InstanceGroups ?
>>>
>>>
>>> 5) Idealy, more than combining Model Instances taken from several
>>> InstanceGroups, there could be a way to select wich Instance source is
>>> active for an Instance added to several InstanceGroups. And an
>>> animatable way. So we could switch on the fly wich Instance Source is
>>> seen from an Instance, at what time.
>>>
>>> -> Suggestion: Not sure how to present this feature, but could worth the
>>> brainstorming on it.
>>>
>>>
>>> 6) Now, take our non empty Model added to an InstanceGroup of another
>>> Model. If you want to Instanciate this non empty Model it doesnt work,
>>> well, it works but just like if you added an Instance of the original
>>> instanciated Model. If we could really Instanciate instanciated model (
>>> including their own objects ) then we could compose cascaded instances.
>>> And this could open an all new world of Instance usablity.
>>>
>>> -> Suggestion: Add cascade ( or iterative ) instanciation to non-empty
>>> Model added to InstanceGroup(s). Another way to present it is to be able
>>> to have an InstanceGroup on an Instance.
>>>
>>>
>>> Just some thoughts.
>>> --
>>> guy rabiller | 3d technical director (at) LaMaison
>>>
>>>
>>>
>>
> ---
> Unsubscribe? Mail Majordomo(at)Softimage.COM with the following text in body:
> unsubscribe xsi

---
Unsubscribe? Mail Majordomo(at)Softimage.COM with the following text in body:
unsubscribe xsi