Layers prototype

[hoako]

Hi!

In the patch tracker is a patch for two script.
http://sourceforge.net/tracker/index.php?func=detail&aid=1953238&group_id=11113&atid=311113
This scripts add Layer functionality easily to the document. Also Key Bindings
could be set for set assigning layers. (hiding and deleting not yet implemented,
but will be)

I haven't committed. I'll wait until i get the feedback.

Cheers!
Joaquín


-------------------------------------------------------------------------
This SF.net email is sponsored by the 2008 JavaOne(SM) Conference
Don't miss this year's exciting event. There's still time to save $100.
Use priority code J8TL2D2.
http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

[hoako]

Joaquín Duo wrote:
> Hi!
>
> In the patch tracker is a patch for two script.
> http://sourceforge.net/tracker/index.php?func=detail&aid=1953238&group_id=11113&atid=311113
> This scripts add Layer functionality easily to the document. Also Key Bindings
> could be set for set assigning layers. (hiding and deleting not yet implemented,
> but will be)
>
> I haven't committed. I'll wait until i get the feedback.
Seems it has problems while loading a Document Containing a Layer Manager. Need
to check this more.

Cheers!
Joaquín


-------------------------------------------------------------------------
This SF.net email is sponsored by the 2008 JavaOne(SM) Conference
Don't miss this year's exciting event. There's still time to save $100.
Use priority code J8TL2D2.
http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

[hoako]

Joaquín Duo wrote:
> Hi!
>
> In the patch tracker is a patch for two script.
> http://sourceforge.net/tracker/index.php?func=detail&aid=1953238&group_id=11113&atid=311113
> This scripts add Layer functionality easily to the document. Also Key Bindings
> could be set for set assigning layers. (hiding and deleting not yet implemented,
> but will be)
Tim while making this prototype if found two bugs:
User properties:
When user properties are created via script, when the file is reloaded, they are
reordered missing it's original order. (this doesn't happen on manually created
user properties)

NullOutputScript:
When opening a file it evaluates its script before its user properties are created.
I guess is has to do with /scripting/null_output_script.cpp
line 51: connect_script_changed_signal(sigc::mem_fun(*this,
&null_output_script::on_execute));

Check here to test them: (opening the K-3D files is enough)
http://sourceforge.net/tracker/index.php?func=detail&aid=1955922&group_id=11113&atid=111113

Can you point me how/where should i fix these bugs?

Also i did used the int32_t and inode* user properties successfully, thanks
again for that fix!

Cheers!
Joaquín


-------------------------------------------------------------------------
This SF.net email is sponsored by the 2008 JavaOne(SM) Conference
Don't miss this year's exciting event. There's still time to save $100.
Use priority code J8TL2D2.
http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

[bART Janssens-2]

On 5/2/08, Joaquín Duo <hoakoduo@...> wrote:
> When user properties are created via script, when the file is reloaded, they
> are
> reordered missing it's original order. (this doesn't happen on manually

I think there is some reordering going on in
persistent_container.h/.cpp. Why is the order important?

> NullOutputScript:
> When opening a file it evaluates its script before its user properties are
> created.
> I guess is has to do with /scripting/null_output_script.cpp
> line 51: connect_script_changed_signal(sigc::mem_fun(*this,
> &null_output_script::on_execute));

Yes, the script is executed whenever it is changed. It is also
executed when one of the properties is changed, but not when a
property is added or removed, though maybe it should.

Cheers,

--
Bart

-------------------------------------------------------------------------
This SF.net email is sponsored by the 2008 JavaOne(SM) Conference
Don't miss this year's exciting event. There's still time to save $100.
Use priority code J8TL2D2.
http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

[bART Janssens-2]

On 4/28/08, Joaquín Duo <hoakoduo@...> wrote:
http://sourceforge.net/tracker/index.php?func=detail&aid=1953238&group_id=11113&atid=311113
> This scripts add Layer functionality easily to the document. Also Key
> Bindings
> could be set for set assigning layers. (hiding and deleting not yet
> implemented,
> but will be)
>
> I haven't committed. I'll wait until i get the feedback.

Very cool, amazing what the python scripting can do, especially in your hands :)
I like the way it works, very intuitive.

Cheers,

--
Bart

-------------------------------------------------------------------------
This SF.net email is sponsored by the 2008 JavaOne(SM) Conference
Don't miss this year's exciting event. There's still time to save $100.
Use priority code J8TL2D2.
http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

[hoako]

Bart Janssens wrote:

> On 4/28/08, Joaquín Duo <hoakoduo@...> wrote:
> http://sourceforge.net/tracker/index.php?func=detail&aid=1953238&group_id=11113&atid=311113
>> This scripts add Layer functionality easily to the document. Also Key
>> Bindings
>> could be set for set assigning layers. (hiding and deleting not yet
>> implemented,
>> but will be)
>>
>> I haven't committed. I'll wait until i get the feedback.
>
> Very cool, amazing what the python scripting can do, especially in your hands :)

Thanks a lot!
I was thinking about creating a panel and connecting them to this plugin. Later
create the whole functionality on C++.
I believe many features could be prototyped on python and later passed to C++.

> I like the way it works, very intuitive.
Great!

Anyway, I think the idea of adding a layer int32_t property for the
Meshinstances (and some other nodes, like cameras, grids, ligths) could be used
in future C++ implementations. Also this way an object could be part of several
layer having multiple properties.

Cheers!
Joaquín



-------------------------------------------------------------------------
This SF.net email is sponsored by the 2008 JavaOne(SM) Conference
Don't miss this year's exciting event. There's still time to save $100.
Use priority code J8TL2D2.
http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

[hoako]

Hi again!

Bart Janssens wrote:
> On 5/2/08, Joaquín Duo <hoakoduo@...> wrote:
>> When user properties are created via script, when the file is reloaded, they
>> are
>> reordered missing it's original order. (this doesn't happen on manually
>
> I think there is some reordering going on in
> persistent_container.h/.cpp. Why is the order important?
Well in this particular case every layer has a name. The ideal is having the
name next to the en/disable property. Save a document with the layer manager on
it. When you open it all the layers names went down.

I believe respecting the order is very important for the coherence of GUI in
general.

>
>> NullOutputScript:
>> When opening a file it evaluates its script before its user properties are
>> created.
>> I guess is has to do with /scripting/null_output_script.cpp
>> line 51: connect_script_changed_signal(sigc::mem_fun(*this,
>> &null_output_script::on_execute));
>
> Yes, the script is executed whenever it is changed. It is also
> executed when one of the properties is changed, but not when a
> property is added or removed, though maybe it should.

I guess there is a misunderstanding on the problem. This cause troubles when
opening a file with already created properties from a script on the script node.
Like now (when opening):
1- Creates the script.
2- Evaluates the script.
3- Add the user properties. (without checking that they already are if the
script created them)
You may test this also loading a file with a layer manager. This only happens
with NullOutputScript.

I guess it should evaluate the script after adding the properties.

Cheers!
Joaquín


-------------------------------------------------------------------------
This SF.net email is sponsored by the 2008 JavaOne(SM) Conference
Don't miss this year's exciting event. There's still time to save $100.
Use priority code J8TL2D2.
http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

[bART Janssens-2]

On 5/4/08, Joaquín Duo <hoakoduo@...> wrote:
> Anyway, I think the idea of adding a layer int32_t property for the
> Meshinstances (and some other nodes, like cameras, grids, ligths) could be
> used
> in future C++ implementations. Also this way an object could be part of
> several
> layer having multiple properties.

It's a good idea, but I think it should be formalized in an interface
that provides access to the collection of layer properties. That way,
it's easy to query if
- an object supports layers
- what its layer properties are

Cheers,

--
Bart

-------------------------------------------------------------------------
This SF.net email is sponsored by the 2008 JavaOne(SM) Conference
Don't miss this year's exciting event. There's still time to save $100.
Use priority code J8TL2D2.
http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

[Timothy M. Shead]

Bart Janssens wrote:

> On 5/4/08, Joaquín Duo <hoakoduo@...> wrote:
>  
>> Anyway, I think the idea of adding a layer int32_t property for the
>> Meshinstances (and some other nodes, like cameras, grids, ligths) could be
>> used
>> in future C++ implementations. Also this way an object could be part of
>> several
>> layer having multiple properties.
>>    
>
> It's a good idea, but I think it should be formalized in an interface
> that provides access to the collection of layer properties. That way,
> it's easy to query if
> - an object supports layers
> - what its layer properties are
>  

Layers are really just one type of grouping mechanism, one that happens
to have some related semantics drawn from 2D drafting.  Adding
group-related data to nodes won't scale for the same reason visibility
flags don't scale.  Suggest you take a look at

http://www.k-3d.org/wiki/Node_Grouping_Design

Tim


-------------------------------------------------------------------------
This SF.net email is sponsored by the 2008 JavaOne(SM) Conference
Don't miss this year's exciting event. There's still time to save $100.
Use priority code J8TL2D2.
http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

[hoako]

Timothy M. Shead wrote:

> Bart Janssens wrote:
>> On 5/4/08, Joaquín Duo <hoakoduo@...> wrote:
>>  
>>> Anyway, I think the idea of adding a layer int32_t property for the
>>> Meshinstances (and some other nodes, like cameras, grids, ligths) could be
>>> used
>>> in future C++ implementations. Also this way an object could be part of
>>> several
>>> layer having multiple properties.
>>>    
>> It's a good idea, but I think it should be formalized in an interface
>> that provides access to the collection of layer properties. That way,
>> it's easy to query if
>> - an object supports layers
>> - what its layer properties are
>>  
> Layers are really just one type of grouping mechanism, one that happens
> to have some related semantics drawn from 2D drafting.  Adding
> group-related data to nodes won't scale for the same reason visibility
> flags don't scale.  Suggest you take a look at
>
> http://www.k-3d.org/wiki/Node_Grouping_Design

Good point. Thanks for the tip!
I guess visibility groups, groups and collections could all be merged together,
isn't it?
Also this may help on the problem hiding materials objects, or you prefer
another mechanism?

Cheers!
Joaquín


-------------------------------------------------------------------------
This SF.net email is sponsored by the 2008 JavaOne(SM) Conference
Don't miss this year's exciting event. There's still time to save $100.
Use priority code J8TL2D2.
http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

[Timothy M. Shead]

Joaquín Duo wrote:

> Timothy M. Shead wrote:
>  
>> Layers are really just one type of grouping mechanism, one that happens
>> to have some related semantics drawn from 2D drafting.  Adding
>> group-related data to nodes won't scale for the same reason visibility
>> flags don't scale.  Suggest you take a look at
>>
>> http://www.k-3d.org/wiki/Node_Grouping_Design
>>    
> Good point. Thanks for the tip!
> I guess visibility groups, groups and collections could all be merged together,
> isn't it?
> Also this may help on the problem hiding materials objects, or you prefer
> another mechanism?
>  

These are orthogonal - reducing clutter in the Node List Panel is
something that can be done today.  I believe Joe has requested it from
awhile back.

Cheers,
Tim


-------------------------------------------------------------------------
This SF.net email is sponsored by the 2008 JavaOne(SM) Conference
Don't miss this year's exciting event. There's still time to save $100.
Use priority code J8TL2D2.
http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development