Nabble - k3d-development

Re: Interesting way to use CGAL.

2009-12-01T21:44:19Z

On Wed, Dec 2, 2009 at 2:20 PM, Timothy M. Shead <tshead@...> wrote:

> Daniel Scott Matthews wrote:
>>
>> http://openscad.org/
>>
>> Embedding an OpenSCAD like system within a K-3D object could be one
>> method for reducing the complexity of working with Boolean geometry,
>> the output would be a mesh, potentially with multiple material's
>> applied to different parts. The text property that contains the
>> OpenSCAD code would have access to linked user properties to allow for
>> it to be parametric.
>
> Not entirely clear from the web page, but it looks to me as if OpenSCAD just
> renders the geometry using OpenGL, it doesn't know how to extract the
> surfaces ... correct me if I'm wrong here.
>

It is not interactive, it is a compiler for geometry.

It is the description language that I find interesting, the approach
to working with booleans.
http://en.wikibooks.org/wiki/OpenSCAD_User_Manual/The_OpenSCAD_Language

It does geometry I/O with DXF and STL
http://en.wikibooks.org/wiki/OpenSCAD_User_Manual/STL_Import_and_Export

Loose the GUI and you have module, with a user editable text property,
that emits geometry.

------------------------------------------------------------------------------
Join us December 9, 2009 for the Red Hat Virtual Experience,
a free event focused on virtualization and cloud computing.
Attend in-depth sessions from your desk. Your couch. Anywhere.
http://p.sf.net/sfu/redhat-sfdev2dev
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

Re: Interesting way to use CGAL.

2009-12-01T20:20:16Z

Daniel Scott Matthews wrote:
> http://openscad.org/
>
> Embedding an OpenSCAD like system within a K-3D object could be one
> method for reducing the complexity of working with Boolean geometry,
> the output would be a mesh, potentially with multiple material's
> applied to different parts. The text property that contains the
> OpenSCAD code would have access to linked user properties to allow for
> it to be parametric.

Not entirely clear from the web page, but it looks to me as if OpenSCAD
just renders the geometry using OpenGL, it doesn't know how to extract
the surfaces ... correct me if I'm wrong here.

Cheers,
Tim

[tshead.vcf]

begin:vcard fn:Timothy Shead n:Shead;Timothy org:www.k-3d.org email;internet:tshead@... title:Founder x-mozilla-html:FALSE version:2.1 end:vcard

------------------------------------------------------------------------------
Join us December 9, 2009 for the Red Hat Virtual Experience,
a free event focused on virtualization and cloud computing.
Attend in-depth sessions from your desk. Your couch. Anywhere.
http://p.sf.net/sfu/redhat-sfdev2dev
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

Interesting way to use CGAL.

2009-12-01T13:28:19Z

http://openscad.org/

Embedding an OpenSCAD like system within a K-3D object could be one
method for reducing the complexity of working with Boolean geometry,
the output would be a mesh, potentially with multiple material's
applied to different parts. The text property that contains the
OpenSCAD code would have access to linked user properties to allow for
it to be parametric.

------------------------------------------------------------------------------
Join us December 9, 2009 for the Red Hat Virtual Experience,
a free event focused on virtualization and cloud computing.
Attend in-depth sessions from your desk. Your couch. Anywhere.
http://p.sf.net/sfu/redhat-sfdev2dev
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

Re: Creating NURBS curve caps

2009-11-17T20:13:15Z

Bart Janssens wrote:

> Both the trim and Coons patch method are in my repository now, though
> I haven't merged with sf.net in a long time, so there may be conflicts
> if anyone is brave enough to try pulling in the changes.

Builds cleanly here ...

Cheers,
Tim

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

Re: Creating NURBS curve caps

2009-11-17T14:45:20Z

On Sun, Nov 8, 2009 at 12:20 PM, Carsten Haubold <CarstenHaubold@...> wrote:
>> Unfortunately, both methods fail when the curve is convex, so I tried
>> using a Coons patch. This does not work in the general concave case
>> either, however, so I propose we follow Ayam's method completely:
>> - If the curve to cap is planar, create a NURBS plane over the
>> bounding box and trim that using the curve

Both the trim and Coons patch method are in my repository now, though
I haven't merged with sf.net in a long time, so there may be conflicts
if anyone is brave enough to try pulling in the changes. I'll merge
with sf.net once all NURBS plugins are converted to the new structure.
In the mean time, I'll see if I can make some screencasts and put them
on my blog.

> That was exaclty the reason why I chose the other approach, I didn't want to
> have two different methods doing the same thing just for different end
> types. But true, my solution fails at convex shapes.
> Is is possible to attach or detach a NurbsCreateCaps node from within lets
> say a NurbsExtrudeCurve plugin when (un-)checking the "create caps" option?
> Then that would be nicest way to handle that.

I've removed the flag, and added a "NurbsCapPatch" action script
instead, that uses a (new) NurbsExtractCurve plugin to extract end
curves and then applies NurbsCreatePatch.

It would be nice if in 0.9, we could add a way to bind a UI to such a
script. In this particular case, we really need control over:
- What method to use (Coons or trimming)
- Should the caps be created in the U or the V direction?
- Should both ends be capped?

Cheers,

--
Bart

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

Re: .k3d file type required for materials management.

2009-11-08T19:06:15Z

On Mon, Nov 9, 2009 at 12:49 PM, Timothy M. Shead <tshead@...> wrote:

> Daniel Scott Matthews wrote:
>>
>> On Mon, Nov 9, 2009 at 12:29 PM, Timothy M. Shead <tshead@...> wrote:
>>>>
>>>> It would save a lot of work if I could select a MultiMaterial and save
>>>> it and all it's requirements, connected nodes and bitmaps to a .k3d
>>>> file with the extension .k3dm so that it could only ever me merged,
>>>> not loaded as a scene. I'd also like to be able to have .k3dm.gz files
>>>> that have all the required bitmaps included. The .k3dm files should be
>>>> able to hold more than one MultiMaterial and there should be a File
>>>> menu item to save all materals to a .k3dm or k3dm.gz file.
>>>
>>> Setting aside details of file extensions ...
>>>
>>> In-general, all k3dml files are equal - some just contain more stuff
>>> than others. Similarly, there's no distinction between "opening" a
>>> document, and "importing" a document - both operations execute the same
>>> code, the only difference is that the former happens on a new, empty
>>> document.
>>>
>>> So the foundation is already there, the only thing you need is the
>>> ability to save a subset of a document, instead of the whole thing.
>>>
>>
>> That is what I thought, it is the dependency management that is
>> tricky, and maintaining the property linking.
>
> Yep, I don't think there's any alternative to going through the process of
> developing some use-cases (your materials are perfect here) where we select
> the appropriate subset of a document to be saved.
>

Well there is the smart graph walking method, or the user guided
method, or a combination of both where the user audits a graph
representation of the automatically selected nodes before they are
saved to a file.

The user based method requires an interactive graph view of some type.

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

Re: .k3d file type required for materials management.

2009-11-08T18:49:41Z

Daniel Scott Matthews wrote:

> On Mon, Nov 9, 2009 at 12:29 PM, Timothy M. Shead <tshead@...> wrote:
>>> It would save a lot of work if I could select a MultiMaterial and save
>>> it and all it's requirements, connected nodes and bitmaps to a .k3d
>>> file with the extension .k3dm so that it could only ever me merged,
>>> not loaded as a scene. I'd also like to be able to have .k3dm.gz files
>>> that have all the required bitmaps included. The .k3dm files should be
>>> able to hold more than one MultiMaterial and there should be a File
>>> menu item to save all materals to a .k3dm or k3dm.gz file.
>> Setting aside details of file extensions ...
>>
>> In-general, all k3dml files are equal - some just contain more stuff
>> than others. Similarly, there's no distinction between "opening" a
>> document, and "importing" a document - both operations execute the same
>> code, the only difference is that the former happens on a new, empty
>> document.
>>
>> So the foundation is already there, the only thing you need is the
>> ability to save a subset of a document, instead of the whole thing.
>>
>
> That is what I thought, it is the dependency management that is
> tricky, and maintaining the property linking.

Yep, I don't think there's any alternative to going through the process
of developing some use-cases (your materials are perfect here) where we
select the appropriate subset of a document to be saved.

Cheers,
Tim

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

Re: .k3d file type required for materials management.

2009-11-08T18:39:05Z

On Mon, Nov 9, 2009 at 12:29 PM, Timothy M. Shead <tshead@...> wrote:

>> It would save a lot of work if I could select a MultiMaterial and save
>> it and all it's requirements, connected nodes and bitmaps to a .k3d
>> file with the extension .k3dm so that it could only ever me merged,
>> not loaded as a scene. I'd also like to be able to have .k3dm.gz files
>> that have all the required bitmaps included. The .k3dm files should be
>> able to hold more than one MultiMaterial and there should be a File
>> menu item to save all materals to a .k3dm or k3dm.gz file.
>
> Setting aside details of file extensions ...
>
> In-general, all k3dml files are equal - some just contain more stuff
> than others. Similarly, there's no distinction between "opening" a
> document, and "importing" a document - both operations execute the same
> code, the only difference is that the former happens on a new, empty
> document.
>
> So the foundation is already there, the only thing you need is the
> ability to save a subset of a document, instead of the whole thing.
>

That is what I thought, it is the dependency management that is
tricky, and maintaining the property linking.

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

Creating NURBS curve caps

2009-11-08T18:37:01Z

>> Is is possible to attach or detach a NurbsCreateCaps node from within
>> lets say a NurbsExtrudeCurve plugin when (un-)checking the "create
>> caps" option?

> I think that logic would be at the GUI level. Since the
> NurbsExtrudeCurve shouldn't care about it.
> So probably those plugins should inherit a "iCapsNurb" interface that
> the GUI should detect and react.
> Probably another GUI plugin should take care of creating and deleting
> the NurbsCreateCaps plugin.
> Well, I am just guessing. Tim will say correctly.

Yep, Joaquin is correct ... pipeline manipulation needs to happen at a
UI level. As always, an operation ("Create Caps") that can be usefully
used in multiple places belongs in a separate plugin.

Cheers,
Tim

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

.k3d file type required for materials management.

2009-11-08T18:29:22Z

> It would save a lot of work if I could select a MultiMaterial and save
> it and all it's requirements, connected nodes and bitmaps to a .k3d
> file with the extension .k3dm so that it could only ever me merged,
> not loaded as a scene. I'd also like to be able to have .k3dm.gz files
> that have all the required bitmaps included. The .k3dm files should be
> able to hold more than one MultiMaterial and there should be a File
> menu item to save all materals to a .k3dm or k3dm.gz file.

Setting aside details of file extensions ...

In-general, all k3dml files are equal - some just contain more stuff
than others. Similarly, there's no distinction between "opening" a
document, and "importing" a document - both operations execute the same
code, the only difference is that the former happens on a new, empty
document.

So the foundation is already there, the only thing you need is the
ability to save a subset of a document, instead of the whole thing.

Cheers,
Tim

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

Planet K-3D

2009-11-08T14:59:41Z

Folks:

Just a "heads up" that we've started a new site to help keep track of
what's happening in the K-3D world, "Planet K-3D", located at

http://planet.k-3d.org

... it includes feeds from the blogs, forums, wiki, "official" source
repository, and will grow to include other K-3D-related sources as they
appear.

Enjoy!
Tim

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

Re: Creating NURBS curve caps

2009-11-08T10:29:18Z

Carsten Haubold wrote:

> 2009/11/8 Bart Janssens <bart.janssens@...
> <mailto:bart.janssens@...>>
>
> Moving this discussion about NURBS curve caps to the list, so everyone
> can benefit.
>
> In the original GSoC project, Carsten implemented NurbsCreateCap using
> a ruled surface between two halves of the curve loop. When updating
> ExtrudeCurve, I tried a simpler approach, using the same method as in
> NurbsDisk.
>
> Unfortunately, both methods fail when the curve is convex, so I tried
> using a Coons patch. This does not work in the general concave case
> either, however, so I propose we follow Ayam's method completely:
> - If the curve to cap is planar, create a NURBS plane over the
> bounding box and trim that using the curve
>
> That was exaclty the reason why I chose the other approach, I didn't
> want to have two different methods doing the same thing just for
> different end types. But true, my solution fails at convex shapes.
>
> - Use Coons patches if the curve is non-planar
>
> I had and have no idea of Coons paches so I can't help you there.
>
>
> This would be the behavior of the "create caps" option we have on i.e.
> ExtrudeCurve, and NurbsCreateCaps would have a flag to enable Coons
> even on planar curves.
>
> Alternatively, we could drop the "create caps" flag everywhere and
> require the use of the NurbsCreateCaps plugin whenever caps are
> needed.
>
> Is is possible to attach or detach a NurbsCreateCaps node from within
> lets say a NurbsExtrudeCurve plugin when (un-)checking the "create
> caps" option?
> Then that would be nicest way to handle that.

I think that logic would be at the GUI level. Since the
NurbsExtrudeCurve shouldn't care about it.
So probably those plugins should inherit a "iCapsNurb" interface that
the GUI should detect and react.
Probably another GUI plugin should take care of creating and deleting
the NurbsCreateCaps plugin.
Well, I am just guessing. Tim will say correctly.

Cheers!
Joaquín

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

Re: Creating NURBS curve caps

2009-11-08T03:20:57Z

2009/11/8 Bart Janssens <bart.janssens@...>

Moving this discussion about NURBS curve caps to the list, so everyone
can benefit.

In the original GSoC project, Carsten implemented NurbsCreateCap using
a ruled surface between two halves of the curve loop. When updating
ExtrudeCurve, I tried a simpler approach, using the same method as in
NurbsDisk.

Unfortunately, both methods fail when the curve is convex, so I tried
using a Coons patch. This does not work in the general concave case
either, however, so I propose we follow Ayam's method completely:
- If the curve to cap is planar, create a NURBS plane over the
bounding box and trim that using the curve

That was exaclty the reason why I chose the other approach, I didn't want to have two different methods doing the same thing just for different end types. But true, my solution fails at convex shapes.

- Use Coons patches if the curve is non-planar

I had and have no idea of Coons paches so I can't help you there.

This would be the behavior of the "create caps" option we have on i.e.
ExtrudeCurve, and NurbsCreateCaps would have a flag to enable Coons
even on planar curves.

Alternatively, we could drop the "create caps" flag everywhere and
require the use of the NurbsCreateCaps plugin whenever caps are
needed.

Is is possible to attach or detach a NurbsCreateCaps node from within lets say a NurbsExtrudeCurve plugin when (un-)checking the "create caps" option?
Then that would be nicest way to handle that.

Regards,
Carsten

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

Creating NURBS curve caps

2009-11-08T01:19:55Z

Moving this discussion about NURBS curve caps to the list, so everyone
can benefit.

In the original GSoC project, Carsten implemented NurbsCreateCap using
a ruled surface between two halves of the curve loop. When updating
ExtrudeCurve, I tried a simpler approach, using the same method as in
NurbsDisk.

Unfortunately, both methods fail when the curve is convex, so I tried
using a Coons patch. This does not work in the general concave case
either, however, so I propose we follow Ayam's method completely:
- If the curve to cap is planar, create a NURBS plane over the
bounding box and trim that using the curve
- Use Coons patches if the curve is non-planar

This would be the behavior of the "create caps" option we have on i.e.
ExtrudeCurve, and NurbsCreateCaps would have a flag to enable Coons
even on planar curves.

Alternatively, we could drop the "create caps" flag everywhere and
require the use of the NurbsCreateCaps plugin whenever caps are
needed.

Cheers,

Bart

On Tue, Nov 3, 2009 at 9:17 PM, Bart Janssens <bart.janssens@...> wrote:

> On Tue, Oct 27, 2009 at 11:39 PM, Daniel Scott Matthews
> <dsmatthews@...> wrote:
>> What about some sort of triangulation of the space enclosed by the
>> curve set. Like the sort of routines used in FEM pre-processor
>> utilities, http://www-users.informatik.rwth-aachen.de/~roberts/software.html
>>
>> Or a set of quads that then have a trim curve?
>>
>
> I'm currently looking at Coons (or Gordon) patches, that seems to be
> the way Ayam does it:
> http://ayam.sourceforge.net/docs/ayam-4.html#ss4.27
>

.k3d file type required for materials management.

2009-11-07T20:47:28Z

It would save a lot of work if I could select a MultiMaterial and save
it and all it's requirements, connected nodes and bitmaps to a .k3d
file with the extension .k3dm so that it could only ever me merged,
not loaded as a scene. I'd also like to be able to have .k3dm.gz files
that have all the required bitmaps included. The .k3dm files should be
able to hold more than one MultiMaterial and there should be a File
menu item to save all materals to a .k3dm or k3dm.gz file.

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

Re: How do I run K-3D with limits on both Memory and Processor usage.

2009-11-01T19:14:24Z

On Mon, Nov 2, 2009 at 9:28 AM, Timothy M. Shead <tshead@...> wrote:

> Daniel Scott Matthews wrote:
>>
>> I have a quad core system with 4GB RAM running a server kernel yet
>> there are occasional bugs with K-3D that lock up the system completely
>> and I suspect it is due to a memory leak that is grabbing 100% of the
>> resources. I also have 32GB of swap in a fast RAID configuration but
>> that just lets K-3D gobble up even more memory. :-)
>>
>> How can I run the application so that it is limited to 2GB of RAM (no
>> swap at all) and only 2 cores, so that I can switch to the debugger
>> and halt the application in a way that will let be know what part of
>> the code was so greedy?
>
> As long as K3D_BUILD_INOTIFY_MODULE, K3D_BUILD_NGUI_SPACE_NAVIGATOR_MODULE,
> and K3D_ENABLE_PARALLEL are "OFF", there shouldn't be any multi-threaded
> code running at all.
>
> Take a look at the documentation for "ulimit" to see how you can limit the
> time, memory, files, and threads that a process can use.
>

That did the trick, it looks like the issue is triggered by the
3DSMeshReader but is in the SDK, and this is on a file that did load
OK in the past! I even hand edited the .k3d file so that the .3ds file
was in the same directory to reduce complications, no luck.

xterm &

ulimit -m 2000000 -v 1000000
make run-gdb/fast
run

...
DEBUG: Triggering event for /home/daniel/projects/k3d_artwork/COW_L.3DS
INFO: Loading /home/daniel/projects/k3d_artwork/COW_L.3DS using 3DSMeshReader

Program received signal SIGSEGV, Segmentation fault.
0x00002aaaab49e944 in k3d::selection::operator<< (Stream=Cannot access
memory at address 0x7fffe9264ff8
) at /home/daniel/projects/k3d/k3dsdk/selection.cpp:42
42 {
(gdb) bt
#0 0x00002aaaab49e944 in k3d::selection::operator<< (Stream=Cannot
access memory at address 0x7fffe9264ff8
) at /home/daniel/projects/k3d/k3dsdk/selection.cpp:42
#1 0x00002aaaab49ead0 in k3d::selection::operator<< (Stream=...,
RHS=@0x7fffffffa6cc) at
/home/daniel/projects/k3d/k3dsdk/selection.cpp:88
#2 0x00002aaaab49ead0 in k3d::selection::operator<< (Stream=...,
RHS=@0x7fffffffa6cc) at
/home/daniel/projects/k3d/k3dsdk/selection.cpp:88
#3 0x00002aaaab49ead0 in k3d::selection::operator<< (Stream=...,
RHS=@0x7fffffffa6cc) at
/home/daniel/projects/k3d/k3dsdk/selection.cpp:88
#4 0x00002aaaab49ead0 in k3d::selection::operator<< (Stream=...,
RHS=@0x7fffffffa6cc) at
/home/daniel/projects/k3d/k3dsdk/selection.cpp:88
#5 0x00002aaaab49ead0 in k3d::selection::operator<< (Stream=...,
RHS=@0x7fffffffa6cc) at
/home/daniel/projects/k3d/k3dsdk/selection.cpp:88

...
ad infinitum

------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

Re: How do I run K-3D with limits on both Memory and Processor usage.

2009-11-01T15:28:56Z

Daniel Scott Matthews wrote:

> I have a quad core system with 4GB RAM running a server kernel yet
> there are occasional bugs with K-3D that lock up the system completely
> and I suspect it is due to a memory leak that is grabbing 100% of the
> resources. I also have 32GB of swap in a fast RAID configuration but
> that just lets K-3D gobble up even more memory. :-)
>
> How can I run the application so that it is limited to 2GB of RAM (no
> swap at all) and only 2 cores, so that I can switch to the debugger
> and halt the application in a way that will let be know what part of
> the code was so greedy?

As long as K3D_BUILD_INOTIFY_MODULE,
K3D_BUILD_NGUI_SPACE_NAVIGATOR_MODULE, and K3D_ENABLE_PARALLEL are
"OFF", there shouldn't be any multi-threaded code running at all.

Take a look at the documentation for "ulimit" to see how you can limit
the time, memory, files, and threads that a process can use.

Cheers,
Tim

[tshead.vcf]

begin:vcard fn:Timothy Shead n:Shead;Timothy org:www.k-3d.org email;internet:tshead@... title:Founder x-mozilla-html:FALSE version:2.1 end:vcard

------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

How do I run K-3D with limits on both Memory and Processor usage.

2009-11-01T14:35:51Z

I have a quad core system with 4GB RAM running a server kernel yet
there are occasional bugs with K-3D that lock up the system completely
and I suspect it is due to a memory leak that is grabbing 100% of the
resources. I also have 32GB of swap in a fast RAID configuration but
that just lets K-3D gobble up even more memory. :-)

How can I run the application so that it is limited to 2GB of RAM (no
swap at all) and only 2 cores, so that I can switch to the debugger
and halt the application in a way that will let be know what part of
the code was so greedy?

------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

Re: Just upgraded to Ubuntu 9.10 Karmic Koala and all my GL problems are fixed.

2009-11-01T14:00:50Z

Daniel Scott Matthews wrote:
> On Mon, Nov 2, 2009 at 4:49 AM, Timothy M. Shead <tshead@...> wrote:
>> Timothy M. Shead wrote:
>>> Daniel Scott Matthews wrote:
>>>> It broke the build of the ODE module so I have turned that off that
>>>> part of of my build config until I can determine why.
>> Looks like the upgrade to Ubuntu 9.10 removed the ode-dev package on my
>> system. After re-installing it, the ODE module builds fine.
>>
> That fixed it, is this a bug for the Ubuntu people?

Not sure whether this is expected behavior or not.

As a general observation, upgrading your OS calls for an automatic
re-run of CMake no matter what ;)

Cheers,
Tim

------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

Re: Just upgraded to Ubuntu 9.10 Karmic Koala and all my GL problems are fixed.

2009-11-01T13:14:10Z

On Mon, Nov 2, 2009 at 4:49 AM, Timothy M. Shead <tshead@...> wrote:

> Timothy M. Shead wrote:
>>
>> Daniel Scott Matthews wrote:
>>>
>>> It broke the build of the ODE module so I have turned that off that
>>> part of of my build config until I can determine why.
>
> Looks like the upgrade to Ubuntu 9.10 removed the ode-dev package on my
> system. After re-installing it, the ODE module builds fine.
>

That fixed it, is this a bug for the Ubuntu people?

------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

Re: Just upgraded to Ubuntu 9.10 Karmic Koala and all my GL problems are fixed.

2009-11-01T10:49:56Z

Timothy M. Shead wrote:
> Daniel Scott Matthews wrote:
>> It broke the build of the ODE module so I have turned that off that
>> part of of my build config until I can determine why.

Looks like the upgrade to Ubuntu 9.10 removed the ode-dev package on my
system. After re-installing it, the ODE module builds fine.

Cheers,
Tim

[tshead.vcf]

begin:vcard fn:Timothy Shead n:Shead;Timothy org:www.k-3d.org email;internet:tshead@... title:Founder x-mozilla-html:FALSE version:2.1 end:vcard

Re: Just upgraded to Ubuntu 9.10 Karmic Koala and all my GL problems are fixed.

2009-10-31T18:42:58Z

Daniel Scott Matthews wrote:
> It broke the build of the ODE module so I have turned that off that
> part of of my build config until I can determine why.

Starting my upgrade now ;)

Cheers,
Tim

[tshead.vcf]

begin:vcard fn:Timothy Shead n:Shead;Timothy org:www.k-3d.org email;internet:tshead@... title:Founder x-mozilla-html:FALSE version:2.1 end:vcard

Just upgraded to Ubuntu 9.10 Karmic Koala and all my GL problems are fixed.

2009-10-31T14:41:54Z

It broke the build of the ODE module so I have turned that off that
part of of my build config until I can determine why.

------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

OSS Usability Improvement

2009-10-28T07:43:18Z

Hi

Please find below a link to a survey related to my PhD research work to evaluate OSS usability improvement from Contributor's point of view.
It shall not take more than 5 minutes of your precious time.
Your identity is neither required nor recorded. The participation is highly valued and appreciated.

http://www.kwiksurveys.com/online-survey.php?surveyID=OLHOO_22480cb3

Thank you and Best Regards

Arif

Re: Render-able Annotation types - 2, 3, 4, 5...n point relations

2009-10-27T18:08:56Z

Daniel Scott Matthews wrote:
> On Tue, Oct 27, 2009 at 3:59 PM, Timothy M. Shead <tshead@...> wrote:
>> Taking a simple example, if you mean "make dimension lines a primitive type,
>> and provide painters for them", that's what I've had in mind.
>
> Yes, but split the data calculation from the representation because
> the data is handy for other tasks such as parametrics or animation.

Sure, with this model, you have a modifier that takes geometry as input,
and produces geometry + dimension lines as output. As always, you can
have as many different modifier types as you like.

Cheers,
Tim

[tshead.vcf]

begin:vcard fn:Timothy Shead n:Shead;Timothy org:www.k-3d.org email;internet:tshead@... title:Founder x-mozilla-html:FALSE version:2.1 end:vcard

Re: Render-able Annotation types - 2, 3, 4, 5...n point relations

2009-10-26T23:13:01Z

On Tue, Oct 27, 2009 at 3:59 PM, Timothy M. Shead <tshead@...> wrote:

> Daniel Scott Matthews wrote:
>>
>> I'd like to be able to use a node type that took point selections
>> across one or more other nodes (mesh etc.) and output a set of
>> floating point values that described the relationships between the
>> points. These relationships would at a minimum include distance,
>> angle, normal and center. This data could then be used by special
>> annotation nodes that used established CAD standards to represent the
>> relationships numerically and diagrammatically. These annotations (and
>> ideally all annotations) would be render-able.
>
> Taking a simple example, if you mean "make dimension lines a primitive type,
> and provide painters for them", that's what I've had in mind.

Yes, but split the data calculation from the representation because
the data is handy for other tasks such as parametrics or animation.

The code needs to be smart enough to handle sets of 1-n selected
points across multiple nodes.
1 point could have XYZ absolute position and polar coordinates
2 points can have distance, center, vector, as well as the 1 point
data for each point.
3 points has center, plus the data for three sets of two points.

Then there are the calculation of angles for the combinations of all
set of three points in the selected points and bounding rectangles and
boxes.

The idea being that there is a set of routines for calculating useful
geometric data from sets of points and that data can be used by a set
of nodes that present the data in appropriate ways.

>> I have notices there is some code in the API to do some of this, is it
>> an existing WIP?
>
> Not sure what you're referring to here ...

I thought I saw some relevant routines for calculating some of
required data. It was probably just part of something else like the
physics stuff.

------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

Re: Render-able Annotation types - 2, 3, 4, 5...n point relations

2009-10-26T22:59:19Z

Daniel Scott Matthews wrote:
> I'd like to be able to use a node type that took point selections
> across one or more other nodes (mesh etc.) and output a set of
> floating point values that described the relationships between the
> points. These relationships would at a minimum include distance,
> angle, normal and center. This data could then be used by special
> annotation nodes that used established CAD standards to represent the
> relationships numerically and diagrammatically. These annotations (and
> ideally all annotations) would be render-able.

Taking a simple example, if you mean "make dimension lines a primitive
type, and provide painters for them", that's what I've had in mind.

> I have notices there is some code in the API to do some of this, is it
> an existing WIP?

Not sure what you're referring to here ...

Cheers,
Tim

[tshead.vcf]

begin:vcard fn:Timothy Shead n:Shead;Timothy org:www.k-3d.org email;internet:tshead@... title:Founder x-mozilla-html:FALSE version:2.1 end:vcard

Render-able Annotation types - 2, 3, 4, 5...n point relations

2009-10-26T17:35:56Z

I'd like to be able to use a node type that took point selections
across one or more other nodes (mesh etc.) and output a set of
floating point values that described the relationships between the
points. These relationships would at a minimum include distance,
angle, normal and center. This data could then be used by special
annotation nodes that used established CAD standards to represent the
relationships numerically and diagrammatically. These annotations (and
ideally all annotations) would be render-able.

I have notices there is some code in the API to do some of this, is it
an existing WIP?

------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

Re: Is it possible to nest RenderManCSGOperator nodes?

2009-10-24T19:07:10Z

Daniel Scott Matthews wrote:

> See attached .k3d file.
>
> I expected to see an intersection of a sphere with the intersection of
> two other spheres.
>
> The node graph (see attached .svg file) looks like I'm using the
> correct connection logic but if I render it with Aqsis the result
> shown only the first intersection.
>
> Is this a bug, or an error in my method?

You probably need to turn-off visibility for your first
RenderManCSGOperator, it is rendering A & B, which is obscuring the
desired A & B & C.

Cheers,
Tim

[tshead.vcf]

begin:vcard fn:Timothy Shead n:Shead;Timothy org:www.k-3d.org email;internet:tshead@... title:Founder x-mozilla-html:FALSE version:2.1 end:vcard

Re: NGUI idea, feasibility question.

2009-10-24T18:51:29Z

Daniel Scott Matthews wrote:

> How feasible is it to have a toolbar that acts as a buoyant stack,
> where user-actions are listed according to the frequency of use.
>
> If I have used actions A, B, C, D, E, F, and G, 4, 3, 2, 2, 1, 1, 1
> times and I then use action H twice and I once my toolbar would look
> like this A, B, H, C, D, I, E, F, G i.e. the toolbar configuration
> takes into account how recently as well as how frequently a
> user-action is performed.
>
> using the toolbar would require a single mouse click or two key
> strokes, one to select the toolbar and then a number to select the
> actions, with a ten action stack (0-9), else two numeric keys (00-99)
> which is probably overkill.
>
> The selection number would be displayed with the icon of the
> user-action to ensure that the user knew which keys were appropriate
> for the toolbars current state, this is necessary because the it's
> dynamic nature.
>
> Is this idea useful and do-able?

Maybe and definitely ...

Cheers,
Tim

[tshead.vcf]

begin:vcard fn:Timothy Shead n:Shead;Timothy org:www.k-3d.org email;internet:tshead@... title:Founder x-mozilla-html:FALSE version:2.1 end:vcard

Is it possible to nest RenderManCSGOperator nodes?

2009-10-24T18:08:01Z

See attached .k3d file.

I expected to see an intersection of a sphere with the intersection of
two other spheres.

The node graph (see attached .svg file) looks like I'm using the
correct connection logic but if I render it with Aqsis the result
shown only the first intersection.

Is this a bug, or an error in my method?

------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

RenderManCSGOperator_nesting_test_1.k3d.tar.gz (6K) Download Attachment
RenderManCSGOperator_nesting_test_1.k3d_Node_Graph.svg.tar.gz (2K) Download Attachment

NGUI idea, feasibility question.

2009-10-24T16:47:39Z

How feasible is it to have a toolbar that acts as a buoyant stack,
where user-actions are listed according to the frequency of use.

If I have used actions A, B, C, D, E, F, and G, 4, 3, 2, 2, 1, 1, 1
times and I then use action H twice and I once my toolbar would look
like this A, B, H, C, D, I, E, F, G i.e. the toolbar configuration
takes into account how recently as well as how frequently a
user-action is performed.

using the toolbar would require a single mouse click or two key
strokes, one to select the toolbar and then a number to select the
actions, with a ten action stack (0-9), else two numeric keys (00-99)
which is probably overkill.

The selection number would be displayed with the icon of the
user-action to ensure that the user knew which keys were appropriate
for the toolbars current state, this is necessary because the it's
dynamic nature.

Is this idea useful and do-able?

------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

Re: Build broken

2009-10-23T23:56:37Z

On Sat, Oct 24, 2009 at 4:04 PM, Daniel Scott Matthews
<dsmatthews@...> wrote:

> On Sat, Oct 24, 2009 at 3:52 PM, Daniel Scott Matthews
> <dsmatthews@...> wrote:
>> I am getting this error even after a full hg pull/update; ccmake ... ;
>> make clean; make
>>
>> I am about to dete my local copy and pull doen a fresh copy of the
>> repo to see if that helps.
>>
>>
>
> No Luck, the main repository is currently in a state that does not build.
>

Working again, now with funky new selection rectangles, thanks Tim!

------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

Re: Build broken

2009-10-23T23:04:28Z

On Sat, Oct 24, 2009 at 3:52 PM, Daniel Scott Matthews
<dsmatthews@...> wrote:
> I am getting this error even after a full hg pull/update; ccmake ... ;
> make clean; make
>
> I am about to dete my local copy and pull doen a fresh copy of the
> repo to see if that helps.
>
>

No Luck, the main repository is currently in a state that does not build.

------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development

Build broken

2009-10-23T22:52:32Z

I am getting this error even after a full hg pull/update; ccmake ... ;
make clean; make

I am about to dete my local copy and pull doen a fresh copy of the
repo to see if that helps.

[ 46%] Building CXX object
modules/highlighting/CMakeFiles/k3d-highlighting.dir/rubber_band.cpp.o
/home/daniel/projects/k3d/modules/highlighting/rubber_band.cpp: In
member function ‘void module::highlighting::rubber_band::draw(const
k3d::gl::render_state&)’:
/home/daniel/projects/k3d/modules/highlighting/rubber_band.cpp:126:
error: ‘const class k3d::rectangle’ has no member named ‘left’
/home/daniel/projects/k3d/modules/highlighting/rubber_band.cpp:126:
error: ‘const class k3d::rectangle’ has no member named ‘top’
/home/daniel/projects/k3d/modules/highlighting/rubber_band.cpp:127:
error: ‘const class k3d::rectangle’ has no member named ‘right’
/home/daniel/projects/k3d/modules/highlighting/rubber_band.cpp:127:
error: ‘const class k3d::rectangle’ has no member named ‘top’
/home/daniel/projects/k3d/modules/highlighting/rubber_band.cpp:128:
error: ‘const class k3d::rectangle’ has no member named ‘right’
/home/daniel/projects/k3d/modules/highlighting/rubber_band.cpp:128:
error: ‘const class k3d::rectangle’ has no member named ‘bottom’
/home/daniel/projects/k3d/modules/highlighting/rubber_band.cpp:129:
error: ‘const class k3d::rectangle’ has no member named ‘left’
/home/daniel/projects/k3d/modules/highlighting/rubber_band.cpp:129:
error: ‘const class k3d::rectangle’ has no member named ‘bottom’
/home/daniel/projects/k3d/modules/highlighting/rubber_band.cpp:134:
error: ‘const class k3d::rectangle’ has no member named ‘left’
/home/daniel/projects/k3d/modules/highlighting/rubber_band.cpp:134:
error: ‘const class k3d::rectangle’ has no member named ‘top’
/home/daniel/projects/k3d/modules/highlighting/rubber_band.cpp:135:
error: ‘const class k3d::rectangle’ has no member named ‘right’
/home/daniel/projects/k3d/modules/highlighting/rubber_band.cpp:135:
error: ‘const class k3d::rectangle’ has no member named ‘top’
/home/daniel/projects/k3d/modules/highlighting/rubber_band.cpp:136:
error: ‘const class k3d::rectangle’ has no member named ‘right’
/home/daniel/projects/k3d/modules/highlighting/rubber_band.cpp:136:
error: ‘const class k3d::rectangle’ has no member named ‘bottom’
/home/daniel/projects/k3d/modules/highlighting/rubber_band.cpp:137:
error: ‘const class k3d::rectangle’ has no member named ‘left’
/home/daniel/projects/k3d/modules/highlighting/rubber_band.cpp:137:
error: ‘const class k3d::rectangle’ has no member named ‘bottom’
make[2]: *** [modules/highlighting/CMakeFiles/k3d-highlighting.dir/rubber_band.cpp.o]
Error 1
make[1]: *** [modules/highlighting/CMakeFiles/k3d-highlighting.dir/all] Error 2
make: *** [all] Error 2

------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
K3d-development mailing list
K3d-development@...
https://lists.sourceforge.net/lists/listinfo/k3d-development