Fabric Engine EOL

[Bullit]

i think either i have vietnam flashbacks replacing my recent memories or someone edited my previous message

It was me by mistake. Apologies clicked wrong button. Should have been quote.
Please verify if it is correct now.

[Bullit]

Certainly, introducing more criteria will yield a more accurate result regarding any approximation. Do you think that s/w industry fares better than startups in other industries when it comes to success rates? My intuition, which is of course not data, tells that it's quite possible that s/w startups have a lower chance of success. The only upside for s/w is that investment is relatively low compared to others. Anyway, it's a non-important side point.
Regarding your point in bold font, I'm not sure about the artist-friendly aspect of Fabric. One has to just look at Mudbox, which is, not even arguably, a lot more artist friendly than say, Zbrush.
My very take on this is that their target user-base was(is) already congregated around one or two solution already established in the industry. Pretty much the most difficult thing to overcome for any startup: get market share from the current companies. If it's not revolutionary AND extremely well marketed, it's definitely DoA.

I think Mudbox failed because of development inconsistency. At time it appeared it could give a run to Zbrush it stopped.

Note: bold part was my mistake and responsibility.

[wesserbro]

Please verify if it is correct now.

bold part was ok, but me myself commenting about how my point is important sounded kinda douchie
its back to normal. no hard feelings

I think its better to delete those unrelated messages to not distract people from the topic (including this one)

[CafeNight]

FE good idea or no?
for what should I use it. Thats what I say when its was out some time ago
apparently not for anything

[Draise]

I wonder if their business model was unsustainable. Free for independent use, 50 free licenses for small studios (which still is a pretty large studio) and then charging unknown costs for larger studios. I don't know many studios larger than 50 technical seats that would need FE everywhere to have even had the need to purchase that software platform. The cash market was... super small, strong yet, but a brick wall to get into (with 3/5 year turnover on projects and tech upgrading cycles and monumental costs to assimualte new tech hurdles). I was pro leaving it free for small people, let it gain traction for small studios and independants, get people into it.... but even I, who kiiinda knows a little visual programming with ICE, still struggled on thinking up practical applications to invest the time figuring it out in FE. Maybe that is why it didn't stick? Maybe the tech got bought out by a larger system? A huge tech giant wanting to assimulate the tech? Maybe they walked away with bucketloads and new high paying job from their excellent work? Maybe they didin't want to announce the buyout due to whiplash? Maybe they never intended to make it an independant product in the long term? Not releasing it as opensource? (because it was essentially free anyway).... I am sure some huge animationg/vfx industry has it in their shop, now owns it, and will excell with it.

There may be hundreds of factors to it.... but it's still sad. I wish there were some kind of museum for all EOL software, free as in public domain, like a library you could possibly pay to cover maintenance of the servers/liscenses for no-long available software, everywhere... I hope there will be some kind of "Hipster" software artists too "Yeah, I made this from a 2010's software" "woaaaah, that's so rad, take my money, that style is so retro!", like there still is handdrawn art and animation today.. I would hate to see these gems only become a wikipedia article soon forgotten - leaving the world in another 5-10 year darkness till the rest of the world catches up to similar ideas... all those hundreds of man hours invested, dissappear, no-longer enjoyed, no-longer exploited or used.

I thought it was revolutionary: visual programming to build (not just process) 3D and artistic data, integrated into established systems, and it all was running and compiled off GPU speed with a syntax like Python.. The ideas were like... a canvas to build my own software and inhouse pipeline without needing a team of programmers to do it nor learn a coding language, only concepts - saving me hundreds of hours having to study a single syntax (slow to type and troubleshoot and limited to that one language too!) - freeing just be an artist more and yet still do programming - idealogically bridging the gap between programmers and artists, the logic and emotion... converging an industry to be more human. I wanted to use it, specially if they spliced it into software like Blender, also opensource - potentially making it easier to get people to push and pull code to make the integration more perfect. I didin't voice that opinion... but I like the idea of better viewport performance, data management, quick task specific dumbproof workflow development for quick production scalability (like a custom image compositor for one specific type of workflow with enough variables that an untrained low wage student could pick it up and work and process the art in record time), unification of software with a central data hub, a central data hub.... that was it. A central data hub and backend code access for theoretical artists. Still daunting, but Kraken was attractive, the frog and the house in the viewport was attractive, the waving water and hairy characters and the man running at 90fps with millions of polygons.. all of that was attractive. And more, the custom widgets and tools I could potentially use in their latest release to BUILD MY OWN FRICKEN 3D software without needing to code! That was attractive.. give it enough time and education......

And.. I'm kicking myself, I should have been more active in their forums and actually used the software regardless of my fear of programming and visual architecture design. I should have supported them more.

I wonder what Eric Mootz is going to do now!

[Draise]

Another attribute I loved.. was that I could splice in data from one DCC directly into Unreal Engine. I could load and cache Alembic without import/export, and it would be livelinked. It was a proof of concept, I never tried it.. but I kept a mental note to invest in Unreal Engine to then invest in Fabric and UE. I could add features to Softimage could never eventually have, and still have features in other DCC's.. and..... the idea was cool. Fluid. It was like adding ICE to whatever you touched, with added UI and full program production capabilities without needing to code, only program.

[Bellsey]

When a company closes, there's usually alot more information and often some quotes from owners/directors. Only this last week Rushes VFX in London Soho is closing and there was a simple press release with some info.
So personally I would wait to hear from Paul before jumping to any conclusions. If its an acquisition (and I hope it is), then they're probably NDA's upped to the eyeballs.

[jonmoore]

If its an acquisition (and I hope it is), then they're probably NDA's upped to the eyeballs.

The complete removal of all traces of Fabric Engine from the Internet in one fell swoop leads me to believe something along these lines to be true. At least the documentation site and the Git repository still lives. Having said that, if you're using Fabric Engine, I'd advise you archive the Git and Docs at the soonest opportune moment.

If I'm to speculate further I suspect the geometry engine is just as valuable as the KL/Canvas integration technologies. It's very well threaded as it wasn't built on top of a decades old architecture like most of the big boys. If it is an acquisition, the buyer could well be from outside of media & entertainment DCC technologies.

[Pooby]

It feels very sad, (if you assume it's the end of Fabric)

I have been a big fan of Fabric pretty much from the outset, as, on paper, it felt like it had been designed to solve all my woes.

I made a few test tools with it, but always encountered frustration, seeing how powerful it was but not managing to get anywhere fast enough. ConsequentLy, I never fully dived in as I was always waiting for it to introduce things to stop me jumping through so many hoops and instead found myself stuck like a rabbit in the headlights of both fabric and Houdini, never leaving softimage.
I was hoping to work with a studio to remake a lot of my character systems in fabric and I have tried hard to get a few studios to invest in it, but found responses which mirrors my own fears to some extent. Isn't it just for massive studios with Td departments?. Let's see if it takes a foothold first .etc..

My optimistic speculation in the past has invariably lead to humiliating defeat, so what I say holds little credibility, However, I also feel that an acquisition is most likely.
It's a valuable product that I can't imagine just being dumped entirely in the trash. Whether it's going to be repositioned as available to the public is another matter.

It was and is a very good idea. Like Softimage.

[wesserbro]

If it's not revolutionary AND extremely well marketed, it's definitely DoA.

well, from what Draise wrote you can decide if it was meant to be a gamechanger. I think it was, on paper.
But when even Pooby (being so productive with ICE) couldn't adapt it to his pipeline proves that FE creators took the wrong route at some point.

[McNistor]

The risk of investing in small products developed by small teams does not emerge only from the high probability of failure as a business, but from being bought out as well. Usually the original company loses its control over the direction of development or any other important decision regarding the future of the s/w. Especially if we're talking about buying companies like Autodesk - we all know why is that. If Fabric was indeed bought out, let's hope it was not by Autodesk and whoever it was, will stand aside to let the original vision of the creators unfold itself however it may.

[Pooby]

well, from what Draise wrote you can decide if it was meant to be a gamechanger. I think it was, on paper.
But when even Pooby (being so productive with ICE) couldn't adapt it to his pipeline proves that FE creators took the wrong route at some point.

I don't think they necessarily took the wrong route. I think they were concentrating (quite rightly) on the underlying engine first , but it took too long for the user friendliness aspect to arrive, for me personally.

I found that tools I were making in minutes in ICE, took at least 5 times as long in fabric even accounting for my inexperience. And I was constantly finding that you had to roll your own things like 'get point normal' or 'get point reference frame ' etc and after a while it got tiring and annoying coming from ICE where every component fundamental you need is one click away.
(I find the same in Houdini, to be honest)
However, I had no doubt all these would have come in time.

The other thing that I found hard in fabric was the fact that with no modelling tools attached and no blend shape manager etc. It made my processes very convoluted with export and import taking all the spontaneity and fun away.

I was Hoping to make tutorials in fabric as I did with ICE as I learnt it, but I realised that it was never at the point I could make it look easy enough to attract and inspire people to give it a go. Especially as they wouldn't have my home-made fill-in nodes toget point normal etc.
Also, Often when I asked questions on their forum, I was met with 'code-snobbery' from other users (not the Developers, I should mention) who thought that I should learn to code not mess about with nodes. That was quite offputting.

[McNistor]

Also, Often when I asked questions on their forum, I was met with 'code-snobbery' from other users (not the Developers, I should mention) who thought that I should learn to code not mess about with nodes. That was quite offputting.

You'll find the same position/recommendation in Houdini's community: consider writing vex before fiddling with vops. It's about being efficient - one short line of vex equals many vop nodes. Of course, one should create their tool however they prefer and thankfully I don't think there's some sort of snobbery in H community like you encountered there. At least, I saw none of it. You most likely already know this, but maybe your lived experience is different.

[Pooby]

Yes. I understand that nodes are not as efficient as code. But if your role is to improve their nodal system through feedback, bypassing nodes and using code isn't very useful in achieving that end.
The bottom line is that code will never really attract people who don't know already how to code or have an interest in it. Nodes are the gateway.

Houdini put me off because there seems to be a ton of legacy alternatives to writing expressions, each with their own rules/syntax. When I asked about it, the answer was 'just try different methods until you find one that works.' I found that infuriating when trying to learn. It's like having all the pieces from 2 similar but different jigsaw puzzles, in one box.

[McNistor]

Yeah, things do change at a fast pace in Houdini, which has an obvious upside: advancement. It's indeed an issue when many, if not most, learning materials are referring to obsolete methods and tools, but SESI is deprecating lost of stuff as of late while maintaining some backwards compatibility for obvious reasons. In another version or two they'll probably write off lots of old carried away stuff and new learning materials are going to be more relevant to the newest workflows. It's probably worth considering to hang in there, but that's for each to decide of course

[kissb]

I'm hoping for the foundry, to be honest.. or isotropix.. anything but the iceberg called AD.. can't wait to see how this unfolds ( insert dramatic music )