Hate, misinformation, hackers and threatening

Hi everybody,

I’m the coder behind E-Cycles. As many say their truth, I though I would bring mine.
Buying E-cycles with a 980Ti costs 250€ gives about the perf of a 2080 to 2080Ti in many real use case scenarios and much more with the quick settings. So in fact artists with low income get more for their bucks and can spare money. This money is then used to make rendering better, which is a successful example of targeted funding if you will. So the argument paid addons gives advantage to rich people is at least with E-Cycles wrong and in fact the contrary. People with money can afford a 2080 or 2080Ti with very expensive external renderers which contribute 0 to Blender. E-Cycles enable poorer people to compete for an affordable price with the big and rich ones.

Now, normally I don’t speak publicly about such things, but it’s a bit too much currently. So some facts in chronological order:

  • after several patches I did for free which are included now in Blender, I wanted to work more seriously, tried a fundraising for a new patch. I got 400€ and was accused of code hostaging although the code was available publicly and the builds also.
  • I proposed to work for the BF. Ton answered he preferred Cycles dev to be paid by external companies and send me to Insydium. I worked for them, got never paid. I proposed my patches to Brecht and he told me maybe some render farm would like it but their was no interest from the BF.
  • I started E-Cycles, got loads of people saying to be carefull because it may give bad image quality, break in animation, the foundation may not want from my patches, etc. All of this information are now proved wrong, even the one saying the BF may not want them (see next points) but the first image was there.
  • I got several very heated debate bringing a lot of fear about feature that may not work, may not be finished, etc.
  • after some user asked how it was looking for the possible inclusion in master, I decided to contact Brecht to discuss it. Long story short, he although he could include all my patches instantly, he didn’t to respect my work and because my users are very happy with my work. But as I said (although it was clear from the begining) that I would make the patches available publicly in late december, he said he wanted it all for 2.81. He also noted his opinion was not reflecting the one from the BF, but added at the end of the discussion that he can add similar features if I don’t do that. Inclusion in 2.81 or late December makes 6 weeks difference, so it’s not really a time reason. So in the end, he is ready to use fund money to do duplicate work just to show… I don’t know what.
  • after the discussion with Brecht, the computer I use to post here and debug user files was hacked and the virus targeted only Blender files and source code files.
  • Now I get some haters in my thread who say E-Cycles is much slower (like 2x) than vanilla Cycles.

As much as I would like to make E-Cycles open source now and don’t care about builds anymore, uploading on the different markets, etc. The BF refused my offers several times and now that I took another way, I want to respect my users and do what I said I will do.

Although I’m sure my work is beneficial to this community, I’m here to cooperate not to fight against some fanatics/haters or idealist who don’t do what they teach.
So is it possible to find a way to properly be mutually beneficial here on the long term?


Hi, I use E-Cycles for several month now it save me a lot of time and I get an advanced feature set for very little money.
There was hundreds of tests of E-Cycles user over the time prove the speed of it.
If a scene is slower with E-Cycles than vanilla build there is something wrong with the scene, settings or hardware.
About development plan, stay to it what you say at first time, period.
I saw many many patches molder in the patch tracker over the years and never understand they not accepted or reviewed even lot of user and often professional user bet to include it. One example was OIDN patch from swerner.
Easy to implement, add not much code to the C code base and maintained from Intel.
“No we don´t need it, is not good enough” bla bla.
@bliblubli, I would provide one small patch in December of your code to the patch tracker and wait for BF reaction. If it would accepted next patch.
You can provide your fork on Github or something late December and people can build it they can use it.
As you give very good and fast support and many user would pay for your work anyway if you still publish builds on Gumroad for your price.

Cheers and thank you for your work and support, mib


Is it werid that i have a hard time picking the craziest part in this story!? Whoha! best of luck man! hang in there!


Hi @bliblubli,

I am sorry you feel unappreciated. First of I don’t know about the specifics of your case other than what I have now read here. I’ll ask Brecht too. But until then can you please write me in perhaps some more detail about your case? I have been following your work for a while already because of a separate interest, it looks impressive.

Clearly there are users who very much like your work, which is a great thing.

I think we should be able to work out something together without anything escalating.

I am fine with PMs, I’ll leave you my address in one at least.


Very sorry to hear about this. If you see any hate, intentional misinformation or trolling, please always flag such posts for moderation. We take this very seriously.


I’m not sure which specific patches or discussion this is referring to. Maybe the OpenCL speedups with more adaptive compilation?

I think this is a misrepresentation of the e-mail exchange we had (which I’m happy to be made public if it helps clear up any misunderstandings).

What you first proposed to me was to keep the main E-Cycles 2x speedup outside of future Blender releases when you make the patches public in December/January. And instead it would only be available as part of your (free?) course to keep the incentive to buy E-Cycles.

And my reply was that I think this is not a business model that the Blender Foundation should participate in, as we don’t artificially restrict what goes in the Blender source code.

As far as wanting it all for 2.81. My point was that if you were being funded by the BF, then from that moment on all development would be in public. And if there was a patch to speed up cycles ready for the 2.81 release, we wouldn’t delay that.

Further, I didn’t say we would add similar features if you don’t contribute your changes. You asked me if I was going to integrate E-Cycles patches before christmas. My reply was:

No, I’m not planning to do that. We may end up doing similar optimizations before the end of the year, but I have no intent of looking at your code so I wouldn’t know.

So what I meant was, we’ll be looking at Cycles optimizations too, but if some of them are similar to yours I don’t know, because I don’t plan to look at the E-Cycles source code before you release it.


I know we had a lot of misunderstanding during our exchange, but there is an important detail. I didn’t ask the BF to keep the main speedup outside of future Blender releases. I was discussing the different possibilities of how the patches after release to the public could be used. One of the possibility was indeed to keep the patches as patches, for the course indeed then available on the BF channels (cloud or free as you would have wished then), so that:

  • more people get an incentive to both learn how to improve Blender and see it’s not that hard and everybody can do it. It empowers many to be the only one to decide what goes in or not, allowing a perfect fit for very different workflow. The current model of one build to rule them all with architects, visualizers, scientific people, game developers and of course animators, etc. make it hard to satisfy everyone’s need and put some pressure on you to include patches quickly, create disappointement on user and developer side. The learning way (kind of instead of offering a fish to a hungry person, teach him to fish) fits the current situation much better, the course show it’s actually easy to learn and maintain, also for artists.
  • and also to keep my funding independent from the BF by offering regular builds and support for open source code like several successful Linux distros do (RHEL for example). Although I still think it would be a good Idea to motivate more people to participate in Blender development, I was ready as I said for other variants, including working for the BF and making everything open from day one. Independence from the BF is also still good I think as it offers alternatives to a very big user base which not always agree with one person’s decision.

But as soon as we started to discuss the other variant (including all in Blender), the pressure went up. If it’s ready it should go in immediately. As you know it is ready as several professional use it on a daily basis for production work, so basically that condition implied to put everything in for 2.81, ignoring my current user base and what I engaged myself to. Then the possibility of adding similar optimizations, the past shows some correlation is starting. Filmic log was discussed and asked for year even during BConf, within a month after the addition in E-cycles, it is in 2.8. I changed the image display mode to glsl (which in some scenario remove a big bottleneck and can bring substantial speedup), some weeks after a patch goes in 2.8 to do something similar (better implemented than what I did by the way, well done).
And even if you wasn’t planing to integrate more stuff like E-cycles, I came to you to make a cooperation. Now that you have more than a million funding, you can indeed spend time on stuff already done, but what’s the point when there is so much to improve in Blender and this improvement can come for free anyway? Why don’t you want to look at my patches? To tell the world you can do that? I think nobody here needs a proof one of us is competent. If it’s about the delay, I think the community can understand the one year shift. It was said so from the beginning, respected by everyone here and the reason it’s not already in 2.80 is because you and Ton refused in the first place. Now you change your mind and I’m happy to go forward, but the way “okay let’s ignore Mathieu and his users completely, just get it in Blender asap” is not a good way for me and it still isn’t.


Thanks to all for the support, I think it’s good if it now goes public. Ton is also welcome to give his point of view on this. I think it’s better for everyone to get a better picture of what happens under the hood on the human level. I hope a better way of interaction between the BF and independant developers can come from it. If PM help keep things cooler, I’m ok for it.

1 Like

I’m happy to look at them at any point you release them. The reason I’m not doing this yet is explained in my email:

If I did request that source code from you right now and integrated it into Blender, that would hurt your business. I don’t want to do this, clearly you have a good thing going and your users are happy.

I don’t think I said or implied that I wouldn’t look at your patches when you release them?


I’m happy to give you access to all the code. But I want to be sure first that you aknowledge the current situation is also due to your and Ton’s decisions in the past and you guarantee that my current engagement and users will be respected = one year after release for public availability. A clarification on what the inclusion in master (as it seems to be the prefered way) would have for consequences would also be appreciated. I’m the first addon developer to offer my work for inclusion after a year (only BSurfaces did it in the past?). As you know, crowdfunding doesn’t work for serious development outside the BF. So I would like clear statements form the BF on what happens beside taking/receiving.

1 Like

If you don’t want code to end up in master before a certain date, don’t send it to me before that date. I don’t want to read code and then somehow have to keep it in the back of my mind when working on Cycles, so I don’t accidentally make similar changes.

If you contribute code it’s the same as any other patches for Blender. You release it under a compatible license and from that moment on we can work to integrate it, depending on available time and release schedules.

(By the way, Cycles itself was also developed as an add-on for a little less than a year and then contributed to the Blender project, but it was just me working in private.)


For clarification about the exchange with Brecht and the following acts:
I don’t think the BF is stupid enough to act like that. But I think an overall ambiance in a small but non-negligeable part of the community, including some devs to show me as a bad person (Ton calling me a code hostager or some saying true devs work for free or that I’m just doing some git apply and asking money for it) encourages some people to act radically. So a statement from the BF that:
a) what I do is legal, real work and with the course very good for people to really understand and be able to change, which is the spirit of open source, make people more independant.
c) the current situation of E-Cycles behind a paywall is a results of the BF decisions too
b) the offer for inclusion after a year is welcome and a good practice (if you think it’s bad then tell it too, think what free Blender would look like today if all paid for addons did the same)

that would help discourage some less minded people to do stupid things like posting hate message, false benchmark and bad hacking.

1 Like

We go a step further, but you still put the bad decision on me, you being the nice guy who release everything for free to everybody without delay and me the bad guy puting code behind pay walls. Could be an epic film maybe, but I would like to see you and Ton take more of responsability regarding past decision which have consequences today.

The Idea of Ton with Cycles devs working for and paid by external studios worked well in some case, not so good in other. I lived with the consequences already = being under the poverty line with social help until december last year because I was never paid by the Cycles4D guys. It was also my decision to accept to work for them, trusting Ton, yet I lived the consequences alone. Now you didn’t want the work I did either, I thank you for at least pointing me to render farms (which didn’t want because they sell hours an dthey would sell 2x less hours with E-Cycles). So I and my users were the only one believing in my work. Now that you see you can actually benefit from it, you change your mind and I should abide instantly?


about knowing about the changes and the consequences for you: there are loads of area to improve which are 100% sure to stack 100% with my work.

  • better tile managing with multiple GPU and denoising
  • use a real GI cache instead of the world lightning when using AO simplify
  • add RTX
  • optimize texture memory usage
  • etc. etc. etc.

Add on that all the reviews that could be done on the tracker… Some of my patches like the one I did with Lukas Stockner for full atomic SVM compilation can bring huge speedup for OpenCL.


It is not easy to be included in official Master, you see Fracture Modifier branch case.
My request, you all try to reconcile and reach an agreement…
If an agreement is not reached, you could do what Fracture Modifier people have done, a separate branch/project. That way you manage your times as you please.


Considering the desired business model here and the community-based/free as in freedom model of Blender itself, I don’t see why you couldn’t have held on to those patches for a year and only then submit them to the patch tracker. I fully understand Brecht in why asking the patches to sit in the tracker and then requesting a monopoly over the code (for financial gain) would rub the BF the wrong way.

More likely, keep the patches private until a year after they end up in E-Cycles and there should be no problem, as Brecht and co. can’t copy the implementation of something not available for public consumption on their own site. This would apply to your own code only though, not patches from others such as Lukas and Stefen.

I didn’t read the whole story, but my eye catched the above quote. Just want to note that there are other countries where 200 dollar per month is the standard.
I mean 100 dollar is not for every-one the same. Some people could never afford it. That’s all I wanted to say.


I don’t think I’ve ever said or implied what you’re doing is bad. I don’t have a problem with it in any case. What other developers might have said, I guess those were conversations that I haven’t seen, I’ll have to ask about that.

We’re happy to get code contributions, and we’re thankful for the developers making them. The only condition is that when the Blender Foundation gets involved, it’s based on fully open source code.

We’ve definitely lacked time to review and integrate Cycles patches and I hope we can do better after 2.80, when I’m no longer tied up coordinating Blender development as a whole. I would have loved to integrate something like OpenImageDenoise earlier, but we had to make some choices to get 2.80 actually released.


How do you intend to answer the obvious observation that you are basically freeloading on an open-source product in order to sell a desirable set of patches to it – then belly-aching in public about how the community is responding?

If Blender weren’t out there, and if it wasn’t free, you would have nothing to sell and nowhere to sell it. If you’re not “breaking” the rules here, you’re certainly “bending” them. So far I think that the community has been remarkably indulgent.

“I’ve got this great thing called a ‘Bevel tool.’ But if you want to use it in connection with this open-source tool that I didn’t write, pay me $30.” Huh?


I feel there’s been a bit of misurnderstaing from both sides. I really appreciate biblubli work, I really could benefit from it in a lot of occasions. It goes without saying that I enormously appreciate the bf work as well. I make a living thanks to blender.

My feeling is that the 2 sides are both working to improve blender. For this I just hope that you guys can work out any misunderstanding that might have happened and find a solution that is beneficial for everyone.