I've always wondered why niche specific software such as Blender doesn't have a ton of industry backing. Any medium sized graphics shop could have a full developer on the payroll for a fraction of the repeated licensing costs of proprietary solutions, that is, someone who works full time on Blender and whom you can directly approach, in-house, for features and fixes.
And it wouldn't even interfere with their competitive edge, since the software isn't their business. They don't have to care about the GPL as long as the software does what they want.
I think it has something to do with appealing to an employee's vanity, where getting a very expensive software package "for free" to do your job makes one feel appreciated.
Disclaimer: Worked for them as a Blender developer on Next-Gen
It has also been used for many of the VFX shots in the TV show "Man in the High Castle".
"Normal days" changed during the course of the project. In the beginning, it was work on new features for months if necessary, towards the end it were overnight patches for "frame x in shot y looks wrong/crashes/takes forever".
Yes, almost everything is made in blender except for some textures which were painted in Substance Painter and volumetric effects like explosions which were generated in FumeFX and Houdini before being imported, rendered and composited in blender.
- motion blur render times. (They patched Cycles to use Embree as its ray engine to solve this)
- memory consumption
- lack of UDIM
- clunky compositor
Looks like a good movie. And Netflix dubbed it to Swedish! Will probably watch it with my kids some time.
"Yes, cycles was used for everything, though our version of cycles was modified (with stefan's embree core, and crypto-mattes which were beyond valuable for compositing). The version of blender we used was the studios own dev version (which I believe was using blender 2.78 as it's base?)"
The reason why it’s not used more: as good as it is Blender is still not Maya or Cinema 4D. Each 3D application has its own strengths and Maya has a LOT of tools that are application specific and that other applications might struggle to copy because they are so intertwined into its UX/UI and of course lots of proprietary R&D.
There’s also the question of pipeline - .fbx file format is the standard in the industry and for good or for bad it’s ruled by Autodesk afaik. Other applications have to adversarially replicate it (i think?) a bit like .doc files going onto iWork
And on your comment of having employees working on Blender - I believe Disney has a few people who contribute to Blender. But lots of the other applications like Maya and Katana are also extensible and studios like ILM have extensively modified Maya and others so to answer your question - they do extend other applications like Maya, Katana and NUKe and they have people on staff working on those extensions.
The thing those extensions or plugins turn out to often be excellent competitive advantage against other studios so it’s actually helpful to keep them in house. All the big VFX studios have multiple proprietary technologies that they don’t share them with anyone at all - unless it becomes advantageous to sell it to others, in which try often spin it off or sell it to The Foundry.
It learned 3D modelling with one of those "very expensive software packages" and it made me feel appreciated at the time. The knowledge however is now completely useless for me because the software, while still existing, is totally niche now. Blender would have been a lifetime investment...
Getting to learn skills that you can't use unless you continue to pay for the expensive right to use those skills should be valued negatively by a rational economic agent.
only someone who hasn't really used the tool in anger and mastered it would claim this. I can't even switch to eclipse from intellij and not have a huge productivity hit for several months. and that's just typing code! A real 3D modeller would have so much muscle memory for their modelling app of choice it't not funny.
I can only imagine what it would be like to switch from something that you've used a flavor of for your whole career to something entirely new, and 3D modelling for animations is way more complex than what I've ever dealt with.
Keeping it in a programming paradigm, CAD is more like assembly. It's different from platform to platform and even if you know one there's no guarantee you can make another work. Contrast that with 3d modeling packages which mostly use the same concepts, and order of operations rarely matters.
I understand why people use Blender as a CAD package a lot more now, thanks!
As someone who had a prior career in VFX using mostly Autodesk and Adobe software, there are people who spend their whole career in one of these packages and have a lot of difficulty becoming fluent in a second one.
They are designed with completely different metaphors and workflows and yes, the end result is always pixels, but learning multiple high end creative software is nothing like learning different programming languages.
Just like a programming language.
I'm sorry you're so shit at your job that switching IDEs screws you up. Not all of us are so incompetent.
Have you found this to be true in your experience?
That’s exactly it, developing a high-end plug-ins is very expensive; to give that away for free is completely illogical. It’s a restaurant giving away sandwiches and selling napkins. Selling “service” and “support” or “hosting,” is ridiculous for the plugin industry. Most people in the business know how to use most of the important plugins already, so what kind of support would be needed at a level that users would pay? In all of my 15+ years of owning Waves audio plugins, I have not once contacted their support. If I need a reverb, I pay someone for a reverb. I am not going to pay someone to provide “support” for my reverb. I’m a pro, I don’t need to pay for help with a tool that has been almost industry standard for years.
I love open source, Postgres, Rails, Sidekiq for example. But the GPL-religious side of open source? Not so much. It seems too culty-dogmatic-restricted. Kind of the opposite of “free.” “You are free to do anything you want, except anything we don’t want you to do. The MIT license is the one that is actually free.
GPL's ideals are good. It's just that the ideals are hard to achieve when there's a need to also make money. I would describe MIT style to be 'free to exploit me [the software] as you please, even if you don't intend to contribute anything back'.
So if you got it free, it is fair to redistribute free or charge for it or whatever. But if the original author sold it to you for $40, you must sell it for at least $41. There's no royalties; all that money is yours, it just means you can't compete with the original author on price alone.
idk tho i gotta think about it
Also, this is assuming your initial software is worth "stealing", it has to be good enough first, which Blender is nowadays.
Most developers know this, but company managers that forbid open source contribution don't understand this. That's why I'm personally not too much against source available licensing. It's the only language old companies understand.
Don’t fall into that trap. The freedoms provided by both Open Source and Free Software licensing are well defined and mean (1) the right to use the program for whatever purpose and (2) the right to fork.
Anything else is unacceptable.
As for what language companies understand, many of us don’t give a fuck. If they are good citizens and distribute software under FOSS compatible licenses all is good, otherwise we’re not interested, because the availability of the source code is not the point, the right to use that source code is ;-)
Looking at source code without the license to copy from it ... is a legal minefield ;-)
Anyway, so then, when someone else somehow makes money using their code and they don't get a cut, of course that seems like "abuse" to them. If they had any idea what they were doing they would have realized that that's not "abuse" at all, because other people using their code for any purpose whatsoever is literally the point of the license.
It's just as bad if not worse when anyone reflexively GPL's their work without understanding that, because there's a freaking holy hell of a lot more legal consequences to understand about the GPL than there are about say, 2-clause BSDL, which is basically just "this is copyrighted by me and you can't remove my copyright" and "if you use this and something goes wrong, you can't sue me". If they didn't understand such a simple license, what hope do they have of understanding the GPL? None at all whatsoever. People who've been arguing about the GPL for 20+ years rarely understand the GPL, at least not in its entirety.
There's nothing wrong with someone making their code proprietary or GPL'd or whatever- it's their code, and it's their choice to make. But it's also their responsibility to understand all the consequences of that choice, and it's their responsibility to understand why they want to release their code in the first place.
Why anyone would use all third-party proprietary software, I have no idea.
Here's where that procedural texture approach started: "Road to Point Reyes". First "photorealistic" render, 1983. Today, that's not even an acceptable game asset.
I think the notion that people (and companies) pay for proprietary software because there is support or some other reasons are complete BS. They pay for software because they have to in order to use it. It's just a cost of doing business.
As free software has become viable (high quality) in some areas, business folks are more willing to allow it. The primary concern is "can it do the job". The second is how much will it cost me?
Sure, sometimes they buy closed sourcr licenses, but as a corporation providing services to a third party you don't care about the license - you care very much the software adds the value to the output of the employee who uses it.
I don't really see the difference between this and "accept its drawbacks", except that if your developer turns out to be productive, you have something a lot more valuable on your hands, and if the developer fails, you can replace them (whereas for a given proprietary product, you typically can not replace the vendor while keeping the licensing agreement and investment in training [this last bit being a big part of why in-house developers make sense for Blender, as they always have for 3D animation studios]).
The difference is one is known, and the other is unknown. When you're evaluating existing software you can evaluate multiple competitors based on their finished product. You can't do that when you're hiring a developer.
>and if the developer fails, you can replace them (whereas for a given proprietary product, you typically can not replace the vendor while keeping the licensing agreement and investment in training [this last bit being a big part of why in-house developers make sense for Blender, as they always have for 3D animation studios]).
I'd say this argument supports using only existing established successful products, to avoid being burnt.
Only if the software you're buying meets all of your current and future requirements, and is maintained indefinitely. If new requirements come up, they're at least not any more helpful than in-house resources (unless a lot of people suddenly have the same requirements at the same time); and if they stop maintaining it, there's a good chance you literally can't do anything about that.
You just have a different set of unknowns, one of which seems more manageable (which is probably why it is more popular), and probably is more manageable in some subset of use cases.
Nobody expects 3d design software to be maintained indefinitely or expects 3d design software to satisfy future requirements magically. You buy a 3d design software to make you money by creating content. Either it makes you money or it doesn't. The easiest thing to do is to simply buy what everyone else is using because it makes it easier to hire talent from a pre-existing pool of experts, or if need be, train them on a software for which training programs already exist because its popular.
>If new requirements come up, they're at least not any more helpful than in-house resources (unless a lot of people suddenly have the same requirements at the same time);
Well, wait a minute. Before you even get there, you not only have to be good at producing 3D art/content (or atleast good enough to make a decent chunk of change), you have to now hire dev/test/pm folks and successfully run a software development team in-house. I don't find this to be a realistic proposition.
>and if they stop maintaining it, there's a good chance you literally can't do anything about that.
A vendor for any software that you rely on can go out of business. Do you plan to run independent software development teams for your OS, accounting software, browser, IDE, etc ? The answer obviously is no. And the same reasoning applies to your 3d design software.
>You just have a different set of unknowns, one of which seems more manageable (which is probably why it is more popular), and probably is more manageable in some subset of use cases.
I don't agree at all. I think it actually is more manageable.
Do you? Depends on what you need done.
> I don't agree at all. I think it actually is more manageable.
Well, at that point it is a matter of opinion.
I don’t think the popularity of proprietary solutions is vanity, it’s rather that switching between 3d modellers is really hard. It might be months before you’re even close to productive again. Also interoperability between packages is a huge problem. Blender only really started to get a good UI this decade, and I think you’re more likely to see it get picked up in new shops rather than old shops migrating.
COMPLETELY ALTERING YOUR PRODUCTION PIPELINE TO ACCOMMODATE ONE PACKAGE gets the point across a little bit better.
There are studios out there that have decades of in-house scripts and programs built around a given set of software. If you do a lot of architectural rendering you may have your own outdoor lighting systems. If you focus primarily on title sequences you may have certain effects that you developed. Having to completely rewrite all of that in-house code would be a MASSIVE undertaking.
DIY is time-consuming, difficult, and doesn't scale. Small companies often start out with as cheap a solution as they can, but as they grow they'll find they're doing way more work just to support the DIY solution, and features are difficult and time-consuming to produce. If you have the money and you need a feature now, licensing makes sense.
Because prior to Blender 2.8 it was really really bad software compared to commercial alternatives.
Blender is becoming quite good. Which impresses the hell out of me. But it’s taken a long, long time to get there.
If you are not a software shop you really would like to outsource those risks.
Hence, they pay the capital cost to get a fixed, known deliverable.
Developing new software is hard and risky. Sure, if you can get the right team for the job it can work out wonderfully but this just creates another level of risk - you need to find the right developers.
Even then, Pixar's advantage was the people they hired and allowed to thrive... not Renderman.
Conversely, if you're a software company and have media production needs, you generally don't in-house it — you focus on your core competency and outsource media production.
Software people tend not to question outsourcing when it's done by a software company outsourcing things outside their core competency.
If the returns to the particular company that employed the developer was higher than the opportunity cost of expending that money elsewhere in the company, then you'd have a strong case for that. However, most medium sized graphics shops are probably concerned that they'd effectively be subsidising their competitors' production costs, since the 'exclusive' returns to the shop employing the developer are likely minimal compared to any other user of the feature they develop.
The free rider problem / tragedy of the commons strikes again.
What causes most usage of big proprietary software is that the software has 10,000 features and every customer only needs five of them, but for each customer it's a different five. That makes it hard for a free competitor to get users because adding five features only gets one customer; it scales poorly without revenue.
But as the customer yourself, having your developer add the five things you yourself need is completely feasible, and doesn't help your competitors that much because the five things they need are different.
Comparing paying a FLOSS developer to paying for commercial software doesn't make sense. Companies don't see themselves as funding development either way.
And yet they still want them, and things take non-zero time to implement regardless of who is doing it. It's not as if proprietary software appears from whole cloth in zero time containing every feature requested by every customer.
> If they have work to do, they need a solution NOW.
If they have work to do they can do it the same way they did it last week. But when the question comes whether to upgrade to the new version of the proprietary software, you should ask whether it makes more sense to pay the money to do that this year, or stay on the existing version for a year while you use the money that you would have spent and instead get the free software equivalent into a state that it satisfies your needs, then never have to pay for the proprietary software again.
> Then once the purchase has been made, the question of paying someone to work on it is a separate issue. Most rational people would say "why would I do that?"
Because you have twenty or more full-time people using it and if you can make them each even a single digit percentage more efficient on a permanent basis, it becomes highly profitable to hire the software developer that allows that to happen.
> Comparing paying a FLOSS developer to paying for commercial software doesn't make sense. Companies don't see themselves as funding development either way.
What did they think the proprietary software license money was funding?
I'm curious as to why the community working on it hasn't taken steps to change this- the software is equally if not more functional than alternatives like Maya and making it a bit more familiar to industry professionals could go a long way
I bought plenty of commercial software, specially developer tools, that had source available on the installation floppies.
If anything we are turning back to those days, after many are starting to realize how hard it is to keep a business living from donations and occasional consulting gigs.
The real question being: was the comment inaccurate? Can we just discuss that? I know how Foundations and Donor Advised Funds make their decisions, people that aspire to support open source dont have the discretion.
Everytime I ask for clarity I just get immediately shadowbanned like "this guy questioned our authority thats the last straw"
Let's look at it the other way. Let's say your intention is to contribute to thoughtful conversation and that you indeed know a lot about "how foundations make their decisions". I'm quite willing to believe that. In that case, though, you should make your contribution in a 180-degree different way. You need to explain what you know in terms that aren't grandiose and provocative, but rather, scrupulously accurate. You need to explain how you know it. You need to tie it to the topic at hand in a way that explains why it's relevant. And you need to somehow include the limits of what you know, to leave some oxygen for anyone who might know different things than you do. All that is not hard to do, but it requires a different genre of comment: more information, less grandstanding.
There's a simple reason Blender doesn't see much use in film production: Support. There is no way the Blender Foundation could provide the level of support that a company like Autodesk does. When you run into problems you call your enterprise support team and get it fixed. This is why you pay them the big bucks. Their support staff does nothing but fix problems, where anyone that you could hire to do the same would necessarily spend most of their time sitting on their ass doing nothing since stuff just doesn't go wrong all that often. That alone will keep Blender out of a LOT of shops, especially those that are just large enough to be working on major projects.
The idea that a company should just hire a programmer they don't otherwise need just so they can make some software work is asinine. That's like saying a home builder should hire a full time auto mechanic instead of taking their work vehicles to the shop when they break down. No, hiring a guy will not offset the cost of purchasing licenses and support. This is especially true when you hit one of those moments where things get so shredded that the cost of enterprise support seems like a bargain.
There are other, more technical, issues that keep Blender from being used much. Chief among them is that it isn't really exceptional at anything. It's not uncommon for things to be modeled in one program, textured in another, then imported into another for animation, and then brought into a final one for lighting and rendering. There are better programs for modeling than Blender. There are better programs for rendering, texturing, animation. Blender is oatmeal. It's there, it works, but it's not really great at anything and as a result has never found a niche.
Because, ultimately, that's what it's all about.
Weren't you just saying the magic ingredient was "support"? Do you believe a first line help-desk is going to provide better support than a guy whose livelihood depends on it and you can call to your desk?
>That's like saying a home builder should hire a full time auto mechanic
In this specific case, it's more like fairly large transporters having their own mechanics, which they have, because it's cheaper, and better, and faster, and generally very convenient all round.
I can guarantee you there are no top-level VFX houses or game developers that ever speak to first-line support.
If you're not top level you have no need to waste money on a programmer that will likely be sitting on his ass 23/7. No, that isn't a typo. The vast majority of the time these systems work as intended. When they don't, who do you think is going to respond better? The multitude of professionals that are paid to work on this code day in and day out... or Steve, who spends most of his time pretending he isn't shopping for waifu pillows? WE KNOW IT'S YOU, STEVE. CLEAR YOUR BROWSER HISTORY.
When RMS is giving a speech, he has an opportunity to expound on what's now wordplay about "free software", to a captive audience that's perhaps already receptive or prepared to listen.
But low-level grassroots advocacy opportunities often happen in contexts in which people are talking for some other purpose, and if you only say "free software", and they don't already know what you mean, you're actually working against your goal. People who don't know what "free software" means naturally assume you mean software for which they don't pay money. If you instead say "libre software", it's not misleading, and if they don't know, and they care, they can ask you about it, or look it up.
(I suspect it would've been better to fully embrace the "libre" term before an office suite was branded that. Now we have a new potential source of confusion, such as "Yes, I already tried Libre, but liked Office better". But I still usually feel more effective saying "libre software" than "free software". And, in practice, I end up saying "open source" perhaps the majority of the time, even when I'm thinking libre specifically, because "open source" is more established than "libre", perhaps because the FSF keeps saying "free".)
If our industry as a whole managed to utter words like "cloud computing", "webscale" or "full stack" with a straight face I really think "libre software" shouldn't be that out of reach.
That is the current reality with the word "free". Pretty much every HN/Reddit submission on free software will have this discussion, with plenty of people disagreeing with the use of the word "free". Unfortunately, the usual response is simply a denial of reality: "Those who don't oppose the word 'free' have an ulterior anti-free agenda" or some similar sentiment.
You've tried ordering the frappucino at the local diner, patiently explaining the merits of this fantastic bewerage, convinced by the zeal of your fellows that this thing is the best thing to happen to coffee bean ever. To no avail.
Or let's make this even more concrete. You are an english nobleman trying to order coffee in venice in 1614. No can do, coffee arrived in venice in 1615. Unless you know someone specific you cannot explain coffee to most people in venice 1614. In 1615 everyone knows, or think they know what it is, as clergymen banned coffee only little bit after it arrived...
My point is, you can have a very good idea but ideas don't spread by the merit of their goodness. I mean, we've know decades what we are doing to the biosphere but only now the large population is taking notice as shit starts to actually it the fan.
So people who discuss libre software fervently might be right, but that still does not mean you achieve anything by prosetylizing the concept to your employer.
Last analogue: You were stolen from Senegal and now are planting cotton in Georgia. A guy called Sam claims he owns you and will beat you if you try to leave. You are convinced slavery is wrong and that no one should be able to own another human being. Most of you fellow slaves agree but somehow you've failed to convince Sam and the rest of his farmer friends.
We have a long history of using the word "free". The problems with the word are not hypothetical - it's there in the history - to the point that it divides the free software community.
It's a mentality of "eh, it works". Of the, at this point, 50+ people I've talked to about software licensing about 3 put any merit in the libre approach. That's not scientific nor representative, but to me it's highly discouraging. Whenever my friend, who recently saw RMS talk live, talks about libre software I notice other students roll their eyes.
To be honest, I don't think most people, students or otherwise, are really even aware of what libre software and the GPL is - let alone think it's something worth investing time and energy into.
I still provision my own Linux (VPS) servers, administer MariaDB on them and string it all together with PHP and sprinkle in non-trivial Javascrpt for UI and graphing. My company is profitable, customers are happy, and my software feeds 2 families well.
My methods are antiquated, sure, but if me telling people I'm a full stack developer is laughable/wrong, then I'm truly lost.
"Full stack" is a bit of a pretentious nitpick on my part, I actually hesitated putting it in there but couldn't resist. I'm an embedded developer who does a good amount of low level/bare metal work, so when I hear about engineering positions requiring being able to do basic Linux sysadmin + write some Python/PHP/JS/HTML/CSS/... referred to as "full stack" I always want to arrogantly point out that this is actually quite a long way away from the "full" stack. I guess it's just an other symptom of web technologies at large eating a huge portion of software development.
In my understanding it has more to do with a person's ability to learn, or lack of fear from learning, if you wish.
Of course I have met my share of "full stack" devs (PHP + JS), but that's the same as with "senior" devs... Having a job title and being one is not the same thing.
I've met quite a few who are happy specializing and essentially restricting themselves to a couple of technologies. When they call themselves "full stack", I get a bit annoyed, since I feel it should mean more than that.
So yeah, if you give me a month, I'm sure I can become productive in any given stack, and given a year, I'll be mostly indistinguishable from a specialist.
I know people who do that in practice.
I still can't understand why you think it's such a hilarious question that you laugh out loud about it?
- practical experience with embedded devices
- popular databases
- web frontend APIs and common frameworks
- popular backed platforms
- Dev OPs and other types of automation
A lot of developers specialize, and relatively few branch out from that. I use the term because I don't want to be stuck in a single role.
Programming languages and code are so unique in being freely available, even the infrastructure around the them is proprietary.
These days I don't think the problem is the use of the ostensibly ambiguous word 'free', so much as that people (generic) don't really value freedom (or privacy) as much as we would like them to.
(This may come across as snark, but it seriously isn’t. “Freedom” anything has been polluted, IMHO)
Perhaps bytes of freedom?
On the near zero chance they remember the penguin and ask a free software zealot about it, they'd likely get an earful about how that is a symbol taken from the 2nd Presbyterian Church of Code, and please don't confuse that with the symbology from the 1st Presbyterian Church of Code across the street from it.
The only FLOSS orgs I know that conceivably care about that generic person sitting in coach are Mozilla and Whisper Systems. And both are small time players compared to their proprietary competition.
Given that, it is extraordinary that normal people know anything at all about digital freedom and privacy.
 Just remembered that Linux is GPLv2 licensed, and that the religious zealotry would probably be about Linux not being GNU, but rather a "final" piece that was "good enough" to use in conjunction with other GPLvs licensed OS tools. So it's even worse-- this is like explaining to someone that they're reading the right book in the wrong wing of the right church. I love FLOSS so much.
Unclear about "don't use my tools for monetery gain"
It's easier to pronounce if you're a romance speaker, but still not consistent, again due to the location of that "r." "Free" is pretty easy to say for everyone, and recognizable whether your "r" is a flap, is French, or is the English "r" which sounds like a humming engine.
"my SQL" vs "my-sequel" (I use the former), NginX I still pronounce as "en-jinx", similarly GIF.
If people call it "lib-ruh" software it doesn't really matter as long as the meaning is conveyed.
Not that you are wrong. The "free as in freedom, not as in beer" trope has just gotten too tiresome. It's a losing battle. The only solution is taking control of the language, and Libre helps to achieve this.
Going forward I will give some thought to always using this term when appropriate.
I do think the pronunciation issue is going to hamper its adoption, though. If it’s pronounced as it is in French, it’s very similar to how we say the zodiac sign “libra,” but if you don’t know it’s french, it could easily be Spanish/Italian/Latin and be pronounced “leebray.”
The fact that it’s not an easy word to know how to pronounce means that people who have only read it will be hesitant to use it in conversation or public speech.
In fact, just learned now that it's supposed to be loaned from French, I had always assumed it to be from Spanish (I'm biased because I'm Spanish).
Is it? According to https://www.gnu.org/philosophy/free-sw.en.html
‘We sometimes call it “libre software,” borrowing the French or Spanish word for “free” as in freedom’
Rarely are projects of any meaningful size actually free. Someone is paying to afford the hosting/dev/etc. Sometimes that payment is in the form of donations from larger organizations. Often a mix of that and user donations. However I think it's important that we don't so easily undermine a projects funding. Especially projects as great as Blender.
I think I just long for the idea that we all recognized that many of these projects we use could do well with some basic coffee money from us all. Something somewhat like Github is setting up. $3/m to a project I make money off of means nothing to me, but from a couple thousand users it would be amazing for the sole project owner.
It's complex, I know. There's pros and cons to a lot of funding talk. Nevertheless, I just want to see great projects succeed. As Blender is.
Adopting the term Freedom Software can be done by you, today, at no cost. Everyone will understand what it means.
Messaging matters. Call it Freedom Software :)
When you say "libre software" people who don't already know what it is are probably done listening before you get a chance to make a point about anything.
Why so complicated?
No reasonable person would decide that the best way to cut through a complicated overlap of a technical discussion a moral issue with many subtleties was with a pun, of all things.
Are there calls to change Blender's license?
Are add-on developers violating GPL? The post mentions that a bridge between open source and proprietary needs to be open source but the add-on itself doesn't?
Is it just a "how dare you sell products closed source products on top of blender?"
The business model of providing support is all well and good but it's just one. If the software is super easy to use then why would you pay?
Some of the add-ons on this site are commercial plugins available for sale by their authors. All Blender plugins have to be GPL so this sort of redistribution is legal but some of these authors have been rather upset about it. So a large argument ensued on the Blender forums about the GPL, the ethics of software redistribution even when copyright law says it is okay, etc.
> All Blender plugins have to be GPL
Didn't the blog article just point out an obvious loophole of turning the addon into a bridge to an external non-GPL module, and putting some critical functionality into the non-GPL part?
Of course, developing such addons will be harder, but even moving just some pointless, trivial, easily rewritten code to the non-GPL part would force people who want to redistribute it to rewrite that part.
Or even completely subverting the spirit (but probably not the law) of the GPL by building a small non-GPL "DRM server" that is called and checked from the GPL'd module, and having the open source part refuse to work if this server isn't present? Of course, anyone would be invited to take the GPL'ed code and remove the checks, but it would shift the cost from "redistribute the version for free" to creating and maintaining a fork, which might be enough to get people to pay instead.
> NOTE! This copyright does not cover user programs that use kernel
services by normal system calls - this is merely considered normal use
of the kernel, and does not fall under the heading of "derived work".
Going by GNU's description of what does and does not extend the GPL requirement and how Blender plugins works, to me at least it seems they very clearly have to be GPL. Unless someone took the route that is common in closed source Linux drivers of having a small open source module that communicates with a binary blob where the majority of the actual interesting code is, but I have not seen any Blender plugins like that.
The Linux kernel actually provides a specific license exemption for the headers that makes software using syscalls not bound by the GPL.
Frankly if you want to be doing any meaningful computation in your addon it probably makes sense to go this route and have the closed source part in C while the addon "bridge" is python.
We should GPL the software that is calling the Api of the GPL software.
>not all software running on Linux using Linux APIs (syscalls) is GPL
This is because linus Torvold(Other contributors) has given the syscall exception along with GPL-2.0. That's why there exists all kind of proprietary software running on top linux os.
Note: I am an open source officer I deal on this topic very often
Well yeah it's a debate - I guess this is their side of the debate. Arguing for something isn't 'pretending'.
Just a side-note about the distinctions between open-source, free software, and commercial software. The US government considers any software that has a license and is available to the public “commercial”. So all GNU software in their view is commercial. It’s not a legal distinction of whether money is charged, it’s whether the software has a license and is public.
U.S. law governing federal procurement (U.S. Code Title 41, Chapter 7, Section 403) defines "commercial item" as including "Any item, other than real property, that is of a type customarily used by the general public or by non-governmental entities for purposes other than governmental purposes (i.e., it has some non-government use), and (i) Has been sold, leased, or licensed to the general public; or (ii) Has been offered for sale, lease, or license to the general public ...".
the only term that i am aware of that really works is non-free.
Only if the people between blender depot has a copy distributed under the gpl? Granted, the most likely way they'd get a copy would be under the gpl (get a copy from someone that purchased the plug-in).
But it does not immediately follow that its trivial to legally redistribute the latest copy of a gpl plug-in, unless you purchase new versions as they are released.
Recently people started saying that the OSI's Open Source Definition is restrictive and obsolete, that having the sources available is enough, that it's not a problem to restrict usage, etc... That's why it's important to stress "free as in freedom" in free software and not "open as in you can see the source" in open source.
I think the FSF has tried to squash too many semantic, technical and legal subtleties into the GPL, the words free, libre, etc.
Note: "Copyleft" as opposed to simply Free Software entails reciprocity, e.g., the "viral" part of the GPL. See https://en.wikipedia.org/wiki/Copyleft
The post isn't very clear, but they way I interpret it is that the Blender add-ons do have to comply with the GPL.
Furthermore, I don't think that's new. Look at the answer in January 2017 to this question  on Blender Stackexchange by Jaroslav Jerryno Novotny:
>Blender also includes the Blender Python API, so every piece of code of the addon that uses some Blender Python API must be also licensed under GNU. This only applies to the addon script files or binaries.
Later on, it's clear that compatibility with closed source programs is restricted to sharing files - that is linking is not allowed:
>But you can do something like commercial render engines do: the export plugin is GNU (uses Blender API) and converts scene data to commercial application (ie. renderer) which is not GNU (doesn't use Blender API) and the licenses differ. This works because the addon is not dependent on the non-GNU application.
> I expect that all add-on developers recognize and respect this concept.
"And if you think you ‘suffer from piracy’ or find it hard to do business with Free Software? Just distinguish yourselves with the proven successful free/open source business model: provide docs, training, content, frequent updates and support. Your customers will love you!"
Well, then why does Blender need a development fund? 
I honestly agree with most of what this article says and I admire the spirit of open source development and the GPL. Still, I don't think it's a viable business model for all of us under all circumstances. People need to be able to make money with their software or otherwise some software just won't be available. I guess that's effectively what this says for the Blender community. Nothing wrong with that at all, just not the way to go for everyone.
Edit: To be clear, I'm not downplaying that ideal, only pointing out that at least in the United States it's pretty unrealistic. In some other western countries they don't balk at the idea of public funding for projects that benefit society, and those would be one example of taking steps towards that more idealized world.
"Your customers will love you!" even implies that this way of doing business is somehow superior to other licensing models.
I'd say: Your customers might love you, but for most projects this is likely also going to translate to "your customers won't pay you".
Ditto for the much touted GPL “service model” - charging for support and training. Realistically, this just does NOT work for anyone except for RedHat, because otherwise there would’ve been plentful examples everytime this subject surfaces and there’s always none. Also saying this as an ISV with tens of years of experience - making living off GPL-licensed software through secondary services is an absolute suicidal pipe dream. Even the double-licensed model (which is not applicable in Belnder's add-on case) is dramatically inferior to more conventional licensing models.
If you don't admire the spirit of free software as well / instead, then you may have missed the point of the article.
To clarify: the article contains advice about how to fund operations for plugin developers. I doubt that it is viable for all plugin developers to fund their operations off of documentation and services. I am just saying that we should be honest and express that funding the development directly will be needed and probably the only way to get it is donations. That's really all I am saying here.
One example is I was recently messing around with Space Engineers code. Keen software gives opens it up but doesn't mean you can just download the code and get the game for free
On the legal side, free software and open source could very well be considered sinonyms (if that makes sense). The actual difference is one of attitude and end-goals.
You cite 'open source' as something that you admire.
It's well known that 'free software' describes an attitude that the 'open source' crowd wasn't as keen to embrace, hence the subsequent linguistic wrangling.
I really don't see how this kind of linguistic wrangling helps. Instead it kind of diverts from my initial point and drives the discussion into nitpicking.
This is not true.
The word 'free' is used six times in the first paragraph to emphasise the point being made, then they really spell it out for you :
"This freedom is what makes the GNU GPL license so powerful and it is why it’s much more than 'open source'."
Why wouldn't they have such a fund? Sponsoring is a viable FLOSS business model, and it heavily overlaps with support (which is discussed in your quote from the blogpost).
The notion that "people" aren't "able to make money" with FLOSS is just a fallacy, and Blender itself is one of the clearest examples of how wrong it is - it was failing as a commercial prospect, and then got "rescued" or "ransomed out" as FLOSS via crowdfunding - one of the earliest crowdfunding initiatives I'm aware of, for that matter; well before there was anything like Kickstarter, Indiegogo etc.
How many people are actually paid by the Blender Foundation for their contributions to the software? How many fewer employees do they have than a company like Autodesk or Maxon?
The simple fact is that FLOSS is the outsourcing of large swathes of development cost to those stupid or interested enough to make something happen for free on their own time.
What do you do when that interest wanes?
What do you do? You pay for support, plausibly via a crowdfunding-like mechanism - continued development and maintenance being something that's encompassed by "paid support", of course. What do you do when interest in a piece of proprietary software wanes and the original developer goes out of business altogether? Never mind, you answered that quite nicely already!
FLOSS is more viable than non-FLOSS in the long run. And we have solid empirical evidence of that, because we still have code from the 1970s and 1980s chugging along nicely as part of our modern Linux distributions. Try that with Windows 10, or even mac OS!
And yet, in the meantime...
> ... How many fewer employees do they [Blender Foundation] have than a company like Autodesk or Maxon?
...how is that supposed to be a bad thing, exactly? It's a lot like measuring the worth of a software project by the LOC of its source code, given a constant or even shrinking set of features.
I simply don't buy that we all should be living off donations to support our software development. It's utterly clear that only the biggest projects can pull that off, and even there — only to a certain degree, with some notable exceptions (RedHat, Linux).
Most "software development" is internal software that exists simply to solve some organization's bespoke issues, and is not open to the public - FLOSS vs. proprietary is a non-issue there. As far as off-the-shelf, non-internal stuff goes, software maintenance is actually a bigger issue than software development per se, and it's entirely appropriate to say "I will not be putting any effort into maintaining this unless you pay me some real $$$ for the trouble." So, I'm not even sure that we're disagreeing about anything of relevance - except inasmuch as "being able to work on the project - or to sponsor work on the project by third-parties - no matter what" is often very important!
Do you have a source for this?
Im so conflicted about this. I feel like my left brain is fighting my right brain. Or mom and dad are fighting again.
On one hand, I have been a huge blender fan for over a decade and personally would give away anything I developed for it. Even if it were high quality. I am all for this way of doing it. Kind of forcing a level playing field no matter if you are AAA studio or some broke but talented college student making 3d models.
On the other hand I totally get how those add-on devs feel being officially told "Thanks for all your hard work, but if someone takes your add-on and gives it away, it's fine." - These add-on devs can charge users to download the add-on, not for the add-on. I used to believe it wasnt right to charge for the ability to download a 'free' add-on, but after seeing how high quality some of the addons are, I completely feel they deserve some compensation for improving blender even more.
Basically, I don't even know what's right anymore.
Would that resolve your hemispheric conflict?
What you are proposing is a form of a shared source license.
My thought is not to restrict anyone from doing anything. Rather that, for example, any profit-seeking entity can use any code, but that that profit-seeking entity must pay a standard and proportional fee to the license-holder of that code, generally its author(s) and/or maintainer(s).
Does that qualify as "restricting anyone from making use of the program in a specific field of endeavor"? Going by the wording, I would lean toward "no".
Furthermore, while on that page I glanced at the other paragraphs, and I couldn't help but notice that according to #9: "License Must Not Restrict Other Software" and its text: "The license must not place restrictions on other software that is distributed along with the licensed software. For example, the license must not insist that all other programs distributed on the same medium must be open-source software.", the GNU/GPL don't qualify as "open source" software.
It is to laugh.
Well, so you are effectively restricting the field of endeavor. People who want to do business with the code licensed this way can no longer do so without having to pay while all others still can use the code freely. I don't think there's much room for interpretation as to whether this fulfills the term "discrimination".
What you can do instead is dual-license your code, offering more "favorable" terms in a second proprietary/commercial license to businesses. However, this usually comes with the caveat that licensees of the commercial license can now do proprietary changes without having to give them back to the community.
> the GNU/GPL don't qualify as "open source" software.
> It is to laugh.
What? I am afraid you are misinterpreting the text and/or confusing the terms of the General Public License. This says you should not place restrictions on the licenses of other software contained on the same medium. It doesn't say anything about derivative work. Why would this mean the GPL doesn't qualify as an open source license?
Here's a list of all OSI approved licenses: https://opensource.org/licenses/alphabetical
It’s true that I haven’t comprehensively studied the GPL.
Mostly I’ve observed companies using hundreds, thousands, or millions of hours (depending) of the very hard work of many very smart open source devs in building their own proprietary systems on top and making fucktons of money, none of which is ever seen by the people who made 90 or 95% of their business possible.
May I suggest you first read the license before jumping to the conclusion that the OSI is incompetent at reviewing licenses?
5. Conveying Modified Source Versions.
A compilation of a covered work with other separate and independent works, which are not by their nature extensions of the covered work, and which are not combined with it such as to form a larger program, in or on a volume of a storage or distribution medium, is called an “aggregate” if the compilation and its resulting copyright are not used to limit the access or legal rights of the compilation's users beyond what the individual works permit. Inclusion of a covered work in an aggregate does not cause this License to apply to the other parts of the aggregate.