You add an instance node to your graph where you can pick/reference the other particle system or object to instance…? Seems reasonable to me, but not sure what @jacqueslucke is planning in this respect.
1 Like
This particle tool looks a bit like the Blender Particle Nodes
https://youtu.be/aXse8XQ1IQQ
3 Likes
Talos way of managing Particle nodes sure seems interesting, I hope @jacqueslucke can look at it briefly for inspiration
2 Likes
(Also posted in another thread)
… methods of implementation such as the basics of adding instanced or animated geometry offset by the age of the particle / event are omitted or not documented (from what I can tell by going through the docs) And that’s the primary issue… there’s little to no supporting (official) sample files for people to bug test or get to know how to use the system properly and how to visualise something while designing your system (which was) Something we did during the 3ds max development of particle flow which lead to the decent documentation and use cases that accompanied it, however here it’s seriously lacking which is only going to impede growth or adoption of the system. As an FX artist of x-years I was staring at the system and just banging my head on the desk trying to figure out how to get something renderable, even an animated (character) object offset by time. Things like this will deter a lot of people without proper documentation and official sample files. Tyflow has abridged documentation however as the entire system is based on pflow’s terminology and basic function, it’s exceptionally simple to pick up. As this new blender system is bespoke, it’s a totally different scenario and needs to be addressed at the very very earliest opportunity - even providing sample files from the developers for us to pull apart and see how it ticks which is what we did during the pflow development back in 2003 and it was adopted very quickly by the community.
WRT debugging I’d say unreals blueprint system is a pretty good example.
Nice breakdown on Houdini’s power from a artist
Is the refactor already usable? I saw that it is merged into function but I get a lot of errors trying to build it on windows ^^
I can build on windows without errors. Can you build the master branch? There were some updates to the way things are build on windows, see https://lists.blender.org/pipermail/bf-committers/2019-November/050305.html.
3 Likes
I’ve been testing the new system lately. It seems that to keep things as clean as possible, the different particle systems are best to arrange vertically. I still do find myself wishing for the particle system node to have an output so I could neatly order e.g a trail system horizontally, a bit like this:
It just feels more natural to me. I do recall this being discussed already at some point, though, so if it’s already decided then I’ll just comply 
Another thing. I remember you saying somewhere that particle collisions are not a near future goal, but how about a per-particle force that affects other particles, but only in a given radius? So that although the particles wouldn’t be “colliding” per se, they’d still sort of repel/attract each other inside the radius. It’d be immensely useful. I have no idea if that’s excactly what a particle collision system would entail, so excuse me if it’s a dumb question 
(btw, why is the functions branch download at blender.org so old and doesn’t seem to update even though there’s been changes?)
1 Like
It feels more natural to me as well. Unfortunately, there is one fundamental issue with this approach. How do you represent the case when the trail particles influence the middle particle system as well. With this solution, you’d have to create a link cycle, which should not be allowed.
I’m still thinking about how the general structure could be changed to make it easier to organize the tree. In the last two weeks I checked other existing node based particle/simulation systems in more detail again. With the knowledge I gained during the development of the particle system, it became much easier to extract and understand the key concepts in e.g. Houdini. Now I’m thinking about how these concepts could be adapted to Blender. We almost certainly cannot use them directly, because we have to work with the Blender we have.
Until then, I’ll not change the structure of the node tree. The current design with the Influences which end up in a single input is on purpose very “simple”. It helped me to understand what is required to make such a simulation work. I noticed, that my concepts have a fairly direct mapping to the core concepts in Softimage ICE and Houdini. A lot of what Houdini does in DOP networks could be considered syntactic sugar to achieve a better tree structure (that is also what made it a bit hard for me to understand their node trees at first).
In summary, while I’m quite happy with the current structure, I am asking myself all the time how it can be improved. This is a hard task though. The hardest part of it is probably, that it has to work within Blender without huge changes.
Yes, I think implementing a per particle force so that particles repel each other is very useful, and should be comparatively easy to implement. I’ve never actually tested this though, so I’m not sure if it really works the way we want it to work.
The functions branch from the buildbot is so old, because for a long time there were no functional changes. This week I worked on some changes users can see though, so I’ll run the buildbot now.
9 Likes
Intuitively it feels like combining this per-particle force with a bit of drag should make it possible to produce stable particle “masses” and crystal-like structures, which would be awesome! Though it’d probably also be quite resource-intensive… I hope you end up testing the idea
Thanks!
3 Likes
Is it possible yet to instance an object on the particles yet or can we only display the particles as either points or Tetrahedrons?
For now you can just display them as points and then use Blenders existing duplication system to create instances on every vertex.
6 Likes
Forces ( Turbulence, Gravity…) are missing from the 2.82 Functions Branch Linux 64bit builds. Is this going to be replaced with something ?
1 Like
Yes. The forces have been unified to make them more flexible. You’ll be able to build node groups that mimic the old behavior.
Also see my weekly report.
5 Likes
Awesome …
Thank you for the reply.
1 Like
Can you update the windows build too, on blender.org? It seems that only the mac/linux builds got updated.
1 Like
I have updated my windows build over at https://blender.community/c/graphicall/Qcbbbc/
5 Likes
Hey, thanks a lot!
Oh, somehow I hadn’t realised that the “points” option is actual vertices! It’s nice to now have proper spherical particles for all the tests
2 Likes
What way to use new unified forces? It is no possible understanding how it works by clicking without a guide… It needs some axamples.
1 Like