I do not think SideFX has anything to worry about here, as they have had at least a 20-year head start on the idea of everything being driven by nodes and they also had the concept of the multi-solver almost since Blender was still That little free app. that does not even have undo.
The potential advantage for Blender though is not overall power (that is simply not a realistic prediction to make), but rather ease of use, as there is a more traditional DCC environment attached to it and the devs. have been working to make it friendly enough for artists while providing enough for those who really want to get into advanced math (to the point where they even ditched the original design for the current one that feels a lot like Cycles and the compositor).
Though part of the result of Everything Nodes so far is not even related to what you can do with them at the moment, but in how it incentivized the development of long-awaited UI/UX improvements in the node editor (colored connectors for instance).
Ace canāt understand anything outside of the context of competition. Itās always a zero sum game and a battle to be won. Itās exhausting and heās been told to tone it down dozens of times over the years, to exactly no effect.
Just to be clear, I wasnāt saying that Blender is going after SideFX at allā¦just pointing out that if someone is wondering where things are headed with GeoNodes, they should look at Houdini.
If people do not like my feedback (despite my best effort to articulate and bring useful detail), I am fine with that, but there was no need to lay bait on the hopes it will get me kicked out of the community (and it is pretty egregious in my opinion that 7 people clicked that heart icon).
This is about Geometry Nodes, not the people who use them, I am not going to claim how closures work for instance or how node functionality everywhere will work (though with the latter I think I remember something Maya using nodes under the hood and hidden from the user, is there any similarity)?
I would just love to have a community (with you included) that doesnāt center itās entire being on us vs them ideology.
The post I was critiquing this time has a baseless claim at itās center that a multibillion dollar company is competing with blender, but in this instance they donāt have to worry because of their 20 year head start.
They donāt have to worry because they arenāt in competition. Competition doesnāt need to be the center of every single conversation. Sometimes, people (or organizations) are just doing their thing to the best of their ability and it doesnāt need to be evaluated and judged as to whether or not itās good enough, or going to beat the other, or if they need to worry about what another organization is doing.
The core of the posts of yours that I take issue with is the adamant centering of adversariality that simply doesnāt exist. The constant hyperbolic hypothetical situations do not help the conversation at all.
Anti-fanboyism is just as bad as fanboyism, and you simply donāt need to hold everything up as a fight. This isnāt reality TV. Itās software development. Itās ok if itās boring and you donāt need to try to make it dramatic by continually inventing conflict where it isnt.
Perhaps 7 people liked my post because they also donāt want to see every post turn into a MMA commentary thread about who is besting who, and who doesnāt stand a chance, and who is up against the ropes now. Thereās enough real conflict in the real world, we simply donāt need to drum up more where it doesnāt exist.
If you put Ton and the CEO of Autodesk in the same room, Iām certain they are 10,000 times more likely to have a positive conversation than a boxing match. Can you please carry the same level of decorum into our conversations here?
And no one wants to kick you out of the community. You are a fixture here and you are more than welcome. But please please please try to be more considerate of how much you are empowering narratives of conflict.
And it turns out the majority of people I know in real life make the same statements, so I kind of find it an annoyance when I hear the exact same things said to me by my parents and others from those online (even to the point where I actively fear speaking to them at times).
At this point I just donāt care anymore. If literally no one believes I made any progress, then why even bother? I will just chalk it up to a cognitive inability and just forget about reputation and trying to improve. I can spend most of my days now just alone with the thoughts I have accumulated and with the love of my beagle who is literally the only one I know who seems to truly get me at times.
Sure, I can reduce my activity here and just ādeliver the goodsā when I create something in Blender, if not for peaked interest in its development and the existing inertia resulting from 20 years of activity.
Hello ! Iād say what you think is fine, itās more how itās communicated ! Or how itās presented. I think itās worth exploring and try to understand what could sometimes upset people in the way you express your feelings or ideas. I can understand that it can be quite difficult but itās also part of how we could live in society. As we all have small or big social problems to solve and like in every practice we sometimes need to go back to basics to improve ! Some people are more socially skilled than other but no-one is perfect either, itās natural to struggle and eventually improve !
Nothing is really settled unless you decide to give up !
Anyway I can see that you make efforts and adapt to state things differently sometimes. By getting feedback from peers you might see what could cause some issues. Sometimes what we think is a good way turn out to be not that good Iām sure everyone could relate to that !
Still trying to figure out exactly the difference between closures/bundles and the existing repeat zones, simulation zones, and for each zones. Does this also mean the concept of the physics world and the simulation UI in the properties panel wind up obsolete?
It almost sounds like a closure will be a solver in a boxed-up form, but is it right then to assume the system would be too complex to allow the user to simply open it up and make precise edits according to need (so not at all like the community particle simulators you can find now)?
Well, I think up to this point repeat, sim, and for each zones are relatively easy to understand ! Itās different forms of loops and itās one thing !
Bundles is a way to pass different elements through a single wire.
Say I have two meshes, a point cloud, I can join geometry that and separate those latter but itās not super handy.
Bundle allows to pack things and you can unpack them latter. The cool thing is that it supports geo/curve/points, but also valuesā¦ Itās really a way to pass multiple stuff without having to plug 10 wires if that makes sense !
Now closure itās the most abstract concept above all. Itās a function that you can passā¦
Letās imagine I have a nodegroup A that is a particle generator. And I have another nodegroup B that handle gravity. For now the only way to add that gravity function is to open group A and add group B on it.
closures allows to pass that gravity functionality just like if I was plugging some values. Without having to open the master group, In shortā¦
Anyway, the catch here is that those things are for super power users . Itās when youāre building very complex systems that needs to communicate between each others. I donāt think itās going to be that useful even for pretty advanced users. But since theyāre building physics systems within GN those tools make sense but itās going to be completely obfuscated when weāll use those physics systems. To put it differently Iām not sure we really need to care about those. Compared to current loops that are great tools for advanced GN users.
I think this is another problem they need to solve is to be able to pass global data into GN. Indeed youāre right for now it would be possible to have different simulations in different objects all with a different gravity and no unified parameters. There is no way to create a global parameter like wind force, and pass that to all systems.
They are currently looking into how to address that with new upcoming toolsā¦
Well , itās hard to tell. I think since closures is a way to inject some behaviour in a solver. Think like gravity, collision, forcesā¦ It might be possible to create custom behaviour that can be passed to the big solver without have to fiddle with itā¦
But yeah I think itās likely to be quite advanced for the common mortal anyway weāll have to wait and see, and anyway itās fine to try to change values here and there and try to break the system in a instructive manner !
Chances are the studios that use Blender now will likely have people who will handle the noodling while the artists just enjoy the results, not to mention that once the use of bundled assets really gets going, we will have a combination of things provided by the BF and sold on the various markets.
I have developed a reputation of being a jack of all trades (in part because I would rather not be dependent on a third party when it comes to projects), but I never actually took any kind of advanced math class, so things like matrices/transforms still feel a bit foreign. No, I would rather not run to ChatGPT either.
Yes I think itās gonna be really cool ! Basically we will be provided physics solvers that should be easy to use, at least at the same level of current system.
But nothing would prevent someone from implementing another algorithm, or for studio to build specific tools, or as you said someone selling assets based on those tools, itās a super clever move !
Yeah I totally agree there is a point where itās not the same skill set anymore.
We shouldnāt have to build our own physics/particles solvers to do art stuff but itās amazing that still itās possible !
Anyway, houdini is build just like that , a lot of the tools are in fact nodegroups but I donāt think a lot of people tweak them or build their own. I expect just the same with blender !
I think any physics world is going to be defined via behaviours in nodes, using closures. Just my barely educated guess.
Itās not totally clear to me how deep theyāre going with atomic nodes (as opposed to blackboxes/ubernodes). But it should be possible to tweak future solvers, assuming you have the know-how, otherwise what would be the point?
On the topic of the criticism that was made to youā¦ wrt your tendency to prophecize things, itās true that it is a bit tiringā¦ but I personally donāt want you outā¦ I have trouble picturing a BlenderArtists with no Aceā¦ take the criticism constructively, and donāt despair right out. We have enough of those users already heheh
I didnt understand what closures are supposed to be until i saw the following video. But it is actually a really simple concept which is only difficult to describe with words but easy to understand with a video:
Why do you think that? I see closures as a great way to adhere to the single responsibility principle as it is in programming which makes code easier to understand and maintain.
Iād love to put certain things in my existing node trees as closures, just so I can easily refactor or improve that specific part of my node tree. I have some trees now that are so embedded and so complex I hardly dare open them again.
Closures will alleviate that to a degree. So I actually expect the opposite to happen.
People that are afraid of the added complexity also forget that the end goal of geometry nodes is not to require that everyone has to build them. Closures also open up the way to eventually having color ramps and float curves as input elements in the GN modifier panel for example.
The end goal is to make it so that everyone can build modifiers (or tools) that act and are as easy to use as any native modifier or tool is currently. You will be able to pick your GN group just like a modifier if you make it an asset and all you have to do is play with the settings in its modifier panel to use it. It just takes a while for them to get there and these are necessary steps to make it so.
That doesnāt mean the eventual end goal of GN modifiers has changed.
Iām sure everyone could relate to that !
anyway weāll have to wait and see, and anyway itās fine to try to change values here and there and try to break the system in a instructive manner !