Blendigo 3.8.25.0

[Headroom]

I believe the Indigo SDK would allow rendering in the Blender viewport. The developers would be able to tell you more about that.



Sent from my iPad using Tapatalk

[Oscar J]

This sounds amazing!!!

[zurkdahool]

I believe the Indigo SDK would allow rendering in the Blender viewport. The developers would be able to tell you more about that.

Well.... with a little help from my friends

Btw. where do i find the Rotation/Scale parameter for textures in blendigo?

edit 1: example what i mean with a clay render button mentioned before:

[dakiru]

Btw. where do i find the Rotation/Scale parameter for textures in blendigo?

In UV/Image editor, if I understand correctly.

[lycium]

So much exciting stuff to code!

Great work!

Since 2/3 Glare members are in Berlin (and Nick thinks about visiting this year too!) we should meet up maybe

[zurkdahool]

So much exciting stuff to code!

some words about my progress on blendigo...

blendigo is well coded and overall it is done very clever in some way... but i found some strangeness inside the code (e.g. class handling, duplicated code and so on...) which makes it hard to me to extend blendigo in a professional way without 'dirty hacks' or in a way which makes sense to me at the moment. ... or i don't get it

For example: i tried to implement the uid tag into the exported scene file and change the 'internal_medium_name' tag to 'internal_medium_uid' tag with at least effort. to be successful a uid must be used everywhere (for every xml-node). because if you are saving the scene from indigo and have no unique uid it will be generated... i don't have to say whats happened if it's not unique and you try to reload the scene Also the right coupling between medium and material uid must be done on uid base.... Of course it is possible but the way i have to implement this feature into the existing code is not what i like. I am also thinking about features like <model2> ....


at the moment i stopped coding and thinking about what to do next. some ideas which are whiz through my head:

a) refactoring blendigo
personally my intension is to contribute blendigo with extensions and additional features but i am not interested in refactoring blendigo at the moment

b) re-write essential parts (e.g. igmesh / xml export) in C++ as a python extension
other exporters goes this way and it works great. maybe in the future it is also possible to integrate also a live preview with this way. I really like this idea to refresh my rusty c/c++ knowledge

c) integrate 'blendigo' directly in blender
integrate directly in blender (like octane renderer does with the special blender version). i don't like it because to much work....

d) ask someone who can explain me some internals about the blendigo code

e) leave blendigo as it is
no way...


i will contribute to blendigo in the future but at the moment i am not sure how i can do that...

[Oscar J]

It would be a crying shame if you didn't get the assistance you need with this. I think you might be exactly the guy blendigo needs!

[zurkdahool]

hmmmm... maybe,maybe not. who knows .... i think i should free my mind from the current blendigo code and think about a new approach.... will free my mind at the weekend with paragliding and do some 'architecture summit' after the weekend.... maybe i will come with new ideas

[Oscar J]

Nothing like some good ol' paragliding to clear your mind.

[zurkdahool]

Nothing like some good ol' paragliding to clear your mind.

Exactly! like my club mate: http://xc.dhv.de/xc/modules/leonardo/in ... tID=680265 he is now so relaxed after >150km *hohohooho*

[OnoSendai]

blendigo is well coded and overall it is done very clever in some way... but i found some strangeness inside the code (e.g. class handling, duplicated code and so on...) which makes it hard to me to extend blendigo in a professional way without 'dirty hacks' or in a way which makes sense to me at the moment. ... or i don't get it

I have also found the blendigo code to be a bit confusing, trying to work out the control flow etc.. So it's not just you

[zurkdahool]

Hi,

after thinking about what to do next i have an idea what i am doing now. First step is a kind of a blendigo re-write. Because there are big changes in the code base i can not guarantee that the code will be compatible to the current blendigo code. We will see (maybe it becomes a fork). i don't know at the moment...

Next step for me is to move the exporter code to a new python module (written in c++) which makes it hopefully easy to connect it later to the indigo sdk (which means we can use the blender viewport). this is the result to keep blendigo available for all platforms (e.g. as far as i know there is no linux sdk available at the moment.) thanks to Ono for the sdk support!

at the moment i am doing some planning and 'software architecture' stuff . don't expect too much in the next weeks because there is a lot of work and i am doing this in my free time. i will post some news about the progress from time to time here in the forum. beside of that i will contribute some code to the original blendigo code...

[snorky]

...

[afecelis]

+1 for rendering within blender's viewport!
Then, and only then, shall Indigo be a trueRT engine. Not to mention how much time it will save us by looking at the changes right away and being able to adjust them without having to wait for the scene to get exported to the main Indigo application.

what if... we ran some sort of crowdfunding to support zurkdahool 's work? I'd be willing to donate, dunno about my other fellow blenderheads

[Headroom]

Blendigo's error messages have ever been a test for my patience.
I've removed every object from the scene except the camera and the hemi light (HDR).
Rendering had worked just a few operations before I edited one object by adding a few plane meshes to act as light sources.

'Mesh' object has no attribute 'indigo_camera'

What the heck does that mean ?

Edit:

OK, now that heart rate is down again I can report that identified the defiant mesh and simply deleted it. I still have NO idea how to relate the error message to that mesh. Moving the mesh to another, not rendered layer also did not make a difference. The error still appeared. The mesh had to be deleted.