[Bf-committers] Blender 2.8 and OpenGL version

[Mike Erwin]

My impression is that mesh data will be the same, Cycles materials will be
the same, (new) real-time materials will be similar to Cycles materials to
the extent that similarity makes sense. Should not have to relearn the
tools or APIs when switching between real-time and offline rendering
approaches.

Blender Internal materials will be de-emphasized or dropped.

I don't know enough about particles to comment there.

I'm writing up some Viewport 2.8 ideas / docs / designs and will share
these very soon. Initially it will be "what merwin thinks", but should give
us something to talk about and eventually arrive at "what Blender folks
think".

Mike Erwin
musician, naturalist, pixel pusher, hacker extraordinaire

On Mon, Apr 18, 2016 at 11:54 PM, Owen Hogarth II <gurenchan at gmail.com>
wrote:

> HI guys
>
> I read this and the notes from a few of the more recent coder meetings and
> I had a few questions. I noticed that the new 3.2 renderer will break
> backwards compatibility with previous versions of the particle system.
>
> I have been doing some work with the node system in blender for a project,
> just basically accessing the nodes via the python script to export the mesh
> data, materials, etc....
>
> With the talk of a totally new node system as well, will that mean a
> rewrite of the node's api, would that also break backwards compatibility as
> well if I was doing a lot of work with the current api?
>
> Best
> Owen
>
> On Mon, Apr 18, 2016 at 3:55 AM, Blendergit <blendergit at gmail.com> wrote:
>
> > Yes, the ID is not related to OpenGL...and I like your idea. The object
> ID
> > could be:
> >
> > Object ID = Blender file GUID + Sequential Number
> >
> > This would be great...and of course a python function like
> > getObjectbyGUID(int).
> >
> > The change to 2.8 and all recode is the moment to think in this task. The
> > point is to get an immutable id.
> >
> > Antonio
> >
> > From: Piotr Arlukowicz
> > Sent: Sunday, April 17, 2016 6:20 PM
> > To: bf-blender developers
> > Subject: Re: [Bf-committers] Blender 2.8 and OpenGL version
> >
> > It is not necessarily connected to openGL, but having an unique random ID
> > stored in each Blend file upon creation/first-save will be also
> beneficial
> > and very welcomed. It could be even a date since epoch - anything - which
> > could help distinguish files copied and changed from files created from
> > scratch.
> >
> > pio
> >
> > ​Piotr
> > ​ Arlukowicz, BFCT​
> >
> > ​*YT: /user/piotao?feature=guide*
> > *FB:* */polskikursblendera* *TW:*
> > */piotao*
> > *Blender Network:* *https://www.blendernetwork.org/piotr-arlukowicz
> > <https://www.blendernetwork.org/piotr-arlukowicz>*
> > *Polski Kurs Blendera:* http://polskikursblendera.pl
> >
> >
> >
> >
> > 2016-04-17 12:10 GMT-04:00 Blendergit <blendergit at gmail.com>:
> >
> > > Hi Mike,
> > >
> > > Thanks for the answers
> > >
> > > If we remove things like a simple draw_line and need VBO and shaders,
> any
> > > simple add-on gets complex. Maybe a library with primitives (line,
> > polygon,
> > > set color,...) could help a lot and create a standard API.
> > >
> > > And about matrix, are we going to use GLM?
> > >
> > > PS: Something really useful for Python (not related to OpenGL) is to
> have
> > > a IMMUTABLE object identifier.
> > >
> > > Add an identifier would be as simple as add 1 to any arbitrary number.
> > > It's not necessary keep a sequence...the problem is if you link
> > files...you
> > > can get conflicts, so the best idea could be use something like GUIDs.
> > >
> > > In database design, the unique identifier must keep equal all the time,
> > > and in Blender you must get objects by name, and this name can change,
> > so a
> > > lot of functions would be possible if we add a simple integer immutable
> > > value for each object.
> > >
> > > Maybe, now you have already this, but I have never seen this
> identifier.
> > >
> > > Antonio
> > >
> > > From: Mike Erwin
> > > Sent: Sunday, April 17, 2016 5:56 PM
> > > To: bf-blender developers
> > > Subject: Re: [Bf-committers] Blender 2.8 and OpenGL version
> > >
> > > Hi Antonio,
> > > Yes to 1) and 2) for sure. 3) I have not thought of deeply, but it
> seems
> > to
> > > follow from 1 and 2. The underlying context will be GL 3.2 core.
> > >
> > > Because we'll be targeting OpenGL 3.2 core profile, lots of older but
> > > convenient functions / techniques will not be available. Matrix stack,
> > > batch rendering, etc. So we'll be replacing those as needed, coming up
> > with
> > > some (hopefully) nice abstractions that fit exactly what we need. What
> > > would be helpful from a Python GL point of view?
> > >
> > > Mike Erwin
> > > musician, naturalist, pixel pusher, hacker extraordinaire
> > >
> > > On Sun, Apr 17, 2016 at 11:30 AM, Blendergit <blendergit at gmail.com>
> > wrote:
> > >
> > > > Hi,
> > > >
> > > > Only to be sure about Blender 2.8 and OpenGL version.
> > > >
> > > > 1) Blender 2.8 will be >= OpenGL 3.2.
> > > > 2) All Blender code using old OpenGL version must be migrated (if the
> > > > module is not removed or recoded).
> > > > 3) All add-ons using OpenGL must be updated to >= OpenGL 3.2.
> > > >
> > > > I need to be sure about that because I have several popular add-ons
> > using
> > > > OpenGL and some ideas to improve grease pencil, so OpenGL
> 3.2provides a
> > > > wide new range of tools.
> > > >
> > > > Thanks,
> > > > Antonio Vazquez
> > > > _______________________________________________
> > > > Bf-committers mailing list
> > > > Bf-committers at blender.org
> > > > https://lists.blender.org/mailman/listinfo/bf-committers
> > > >
> > > _______________________________________________
> > > Bf-committers mailing list
> > > Bf-committers at blender.org
> > > https://lists.blender.org/mailman/listinfo/bf-committers
> > > _______________________________________________
> > > Bf-committers mailing list
> > > Bf-committers at blender.org
> > > https://lists.blender.org/mailman/listinfo/bf-committers
> > >
> > _______________________________________________
> > Bf-committers mailing list
> > Bf-committers at blender.org
> > https://lists.blender.org/mailman/listinfo/bf-committers
> > _______________________________________________
> > Bf-committers mailing list
> > Bf-committers at blender.org
> > https://lists.blender.org/mailman/listinfo/bf-committers
> >
> _______________________________________________
> Bf-committers mailing list
> Bf-committers at blender.org
> https://lists.blender.org/mailman/listinfo/bf-committers
>