- This topic has 15 replies, 8 voices, and was last updated 2 months, 3 weeks ago by
.
-
Posts
-
Anyone else see this project?
https://posestudio.org/Isn't that really trying to re-invent Studio & Poser? :0/
If it specialised in specific areas, there might be some use, I suppose.
I am not thrilled about it being open-source.
https://en.wikipedia.org/wiki/List_of_formerly_open-source_or_free_software
@sandeep: You do have a point. In this case, I don't see any clear outline/roadmap/framework except 'We'll provide the foundation, you build the house' and 'Fund me' or 'Code me'.
So say after a couple of years of funding and coding other business interests make an offer: Who's 'in charge'? Who 'owns' the assets? Who makes the decisions?
I don't know if you do any FOSS work, but even if the license is GPLv3, it's still Copyrighted by an individual or other entity who can decide to change the license or sell the property at any time, going forward.
So after all the free coding and artwork by others, the 'foundation' boys could walk away with the prize. Sort of like vibe-coding, but with humans instead of AI.
But I'm probably being overly pessimistic.
Who's 'in charge'? Who 'owns' the assets? Who makes the decisions?
So after all the free coding and artwork by others, the 'foundation' boys could walk away with the prize. Sort of like vibe-coding, but with humans instead of AI.
The answer to those questions is usually hidden in the project's CLA (Contributor License Agreement). Some founders (using your term) go to great lengths to legally "sanitize" third-party contributions before they're merged into the codebase. That often means requiring contributors to sign a CLA that gives the project certain rights over those contributions.
The interesting part is what rights the CLA actually grants. If the founders intend to keep the project under a single open-source license indefinitely, they may not need much more than a DCO (like the linux kernel) or a fairly limited CLA. But if they want the flexibility to dual-license the project, change the license later, or incorporate community contributions into commercial products, the CLA will often grant them those additional rights. That's why it's worth reading before contributing.
GitHub supports CLA workflows through various bots and integrations, so the mechanics are fairly straightforward. The legal substance, however, comes entirely from whatever terms the project chooses to put in its CLA.
A good example of a mature contributor agreement is the Eclipse Foundation's. Eclipse hosts projects under several different licenses (including the EPL and GPL, depending on the project), and its contributor agreements are designed to accommodate those differences. The Eclipse Public License itself is a weak copyleft license: modifications to EPL-covered code generally remain under the EPL, but independent modules can remain proprietary. That's one reason Eclipse has been widely adopted for commercial development while still remaining an open-source ecosystem. It also gives you an idea of how much legal thought can go into a well-designed CLA.
Disclaimer: I took some feedback from chatgpt to validate my statements (since we are talking about licenses) and fix the language (since english is not my native). Otherwise, the content, examples and explanation are entirely mine.
Azureus is a real-world example of how this can play out (and since many of us here may have used it in the past). Azureus started as a GPL open-source BitTorrent client developed primarily by a small group of maintainers. Over time, the company behind it shifted its focus toward commercialization, rebranded it as Vuze, added proprietary features, bundled advertising and content services, and the community gradually became less enthusiastic about the project's direction.
The important point is that the original GPL code didn't disappear. Because it had been released under the GPL, anyone could fork the last GPL version and continue development. That's exactly what eventually happened with BiglyBT, which continues as a community-driven open-source project without the commercial baggage that accumulated around Vuze.
So, in a sense, both sides of the open-source model worked exactly as intended. The original maintainers were free to pursue a commercial strategy, and the community was equally free to continue the open-source codebase in a different direction.
Did Azureus use a CLA for accepting contributions? Did sourceforge support CLAs back then? Did BiglyBT achieve the same heights of popularity and success as it's parent Azureus? Well, these are questions for another day.
Disclaimer: as before.
That's why I think it's important to distinguish between ownership and control. A CLA may determine who has the legal right to commercialize or relicense future contributions, but an open-source license like the GPL ensures that released code can't simply be taken away. Azureus becoming Vuze and later being forked into BiglyBT is a good example: the maintainers controlled the project's commercial future, but they couldn't prevent the community from continuing the open-source codebase.
The AI may have some better examples in the ecosystems of Qt, cloud computing and JavaScript frameworks, but those are not the hills I wanna die on 😉 Also, I cited the examples of Eclipse and Azureus, since Azureus is built on top of Eclipse SWT to provide consistent look-n-feel across multiple platforms, like Qt, but using Java.
I'm desperately hoping that this pans out. I'm on their Discord server giving advice as someone who's been in the hobby since the first version of Poser. For now, I'm assuming they mean well.
We already have Blender. Why would anyone need this? Blender's only problem is the relatively steep learning curve. But once that is overcome, it can do almost everything. So unless this can build a Blender replacement that is as easy to learn as DS, I don't see the point.
Looks like they are going to give us a look see....
https://discord.gg/smFvkaaT?event=1521863378392322200I checked out the Discord and github. From what I get from the early posts, it looks like there are five principals involved. They have provided some scaffolding: QT, Vulcan, SQL and suggested an interface with Claude. It's all C++, so is intended to run natively and cross-platform.
On Discord, some users have posed the idea of using Blender as a back-end, @ihatzi (one of the principals) doesn't seem entirely 'sold', but is keeping it open for now.
Currently, it's intended as a unified workflow, something like Blender, but with a more user-friendly interface. If they did use Blender code, it would of course be subject to Blender's GPL, they couldn't spin-off a dual license for commercial purposes.
That would give me a little more 'confidence' that it wouldn't get 'sold out from under' everyone.
Now, they do check all the boxes for a well thought out initial project specification, but even that is pretty much 'up in the air'. They say they need coders, managers, lawyers, artists, moderators, etc., but mostly coders and money.
If it gains/maintains interest and shows some incremental progress, the project may survive, but with Daz, Blender, 3DMax, iClone, Maya, etc. out there and with a lot of us already invested in those workflows, it's going to be a (long) steep hill to climb.
What @ulysses said :0)
I work more in Blender than I do in DS. But I've also been using it for the last 12-15 years so while there are some things DS can do much easier, I still prefer rendering things in Blender. I can also customize some details in Blender once I export stuff from DS. I use plenty of different 3d tools, but they all seem to orbit around Blender as the final stop. The only thing that doesn't do that great of a job in Blender is the painting "feature". It's so bare bones you have to create your own brushes before you can even try to paint. If it at least came with some practical brushes or was easier to install them, it would be better.
- You must be logged in to reply to this topic.
