Following the hot debate on changes to the Blender interface, Ton Roosendaal has now written a deep and insightful commentary on the position of the Blender Foundation. It refers to his base philosophy and expands on what he sees at the best approach to improving the interface. It's a large piece - it'll take you at least 15 minutes to carefully read it. But if you're seriously interested in Blender, its backgrounds and its future, this is one article you really should read.
Link
211 Comments
Thanks Ton. I have hope now.
I hope the GUI will not change to the Andrew Price "amateurs can use" version. :D
It's not an "amateur" GUI... it is a friendly GUI made for designers, architechs,... people that make the same work that Andrew. Make a simple scene (or only a model), pick materials from a preset, put two lights, make a render. Just like how he does it in all of his tutorials.
"Amateurs", You mean beginning users? For anyone who has ever used any top tier software like maya, the interface works a similar way that Andrew Price is proposing. It is a very gui based user interface that hides a very high complex user architecture. Maya's interface is meant to draw in new users as well as old and make the pathway to learning professional skills as less daunting as possible. Being pickup and learn does not necessarily mean amateur.
In any case, Ton said he did not want to make blender for users of other programs -- "Nor am I much interested in supporting Maya users, Sketch-up users, or the forum trolls who don’t use Blender anyway. Why would we in the first place? Why work for people who are not much interested in your work, if you already have like hundreds of thousands of users who are?" Yet why go through the process of creating a splash screen where you can choose maya mode or 3dsmax mode? Why go through the process of 2.5 anyway? The aim of that was to draw in new users? Was it not? If you had no interest in drawing in new users, why go through the process anyway? Just stick with 2.49 which was for traditional blender users. I understand you do not want corporate culture in your organization, yet blender's pathway is for to make the software for 3d and cg artists (engineers, sculpture artists and etc.) These are people who would be counter corporate culture, yet still want tools they can pick up and use and not worry about too much deliberation of where to look for this tool and when to use that tool.
Artists tend to gravitate to tools that do not get in the way of creativity. This is the reason why a lot of pro level artists are leaving photoshop for Mari when texturing. The sense is I just want to paint textures, not worry about the technicalities of it. Which is why Mari is dominating the texture world. The more ease of use the software gives, the more people will use it consistently and more dominantly.
""For anyone who has ever used any top tier software like maya, the interface works a similar way that Andrew Price is proposing""
WTF??? what are you talking about? Maya is a fucking nightmare of GUI for noobs with a lot of toolbars, hidden panels, hundreds and hundreds of windows, a really strange spacebar tool, that work with mouse gestures to show the basic modeling options. Where do you see the "similar" to andrew proposal?
He meant blender is not built to draw happy users away from Maya or other software just for the sake of getting them over. If you use blender you have to understand what you can expect from it and what not.
No offense, but I had to look up "Mari" to see what the hell it was, I've never heard of it before, and I sure as hell ain't a noob to the texturing world. While it's possible I just haven't heard of it, I'd wager you're stretching the definition of "Dominating", and I nearly every texture artist I know still uses Photoshop for texturing, the rest use GIMP.
Endi, you really really need a new act.
Just because someone has a different opinion, doesn't mean they're wrong :)
(not picking sides, just noticed you seem to be attacking people personally here a lot)
Sometimes an opinion can be misinformed and misleading, and thus, wrong.
You're right, I think by this point I've been beset by discussion fatigue and want to step back. It's been lively and a challenge to keep up with and I've seen a lot of good balanced contributions, had a chance to ask some good questions, wrangle through the issues and give some feedback. Something just snapped the other day, and I got tired of "toy blender... toy blender... pros were never newbs" and I boiled over a bit. Well, more than a bit. Thank you for pointing it out, it was well deserved and I thought I should acknowledge it here.
I wish I knew if there was a logical endgame for these discussions in sight. At what point have all the positions been staked out well enough that the discussion is just looping around?
Aha! Unintentionally, you gave! The "experts" and skilled ones, the few, few people that conquered the full comprehension don't want to allow paltry amateurs to be blessed by this secret treasure. If you win, Blender never will get its rightful place on the CG industry. This way Blender would be, forever, a curiosity for a few "experts".
Well this "expert" has used all major packages professionally for the last 20 years and prefers Ton's approach.
It's not a super power, it's not something to conquer, it just takes patience and study, just like all 3d apps.
I've been using Blender and doing 3D for years. I'd love a new GUI just for more directly and efficiently access to Blender's features and less fuss with the rather scattered GUI. I shouldn't have to go through one mode, two toolbars and two subcontext menus just to texture a model. That's simply not efficient, and Blender will only lose more efficiency as more features come along and get crammed into its already rather disoriented GUI.
right because any survey of Blender users or potential new users, or any Blender discussion on a related forum always overwhelmingly support the idea that the interface is so adept the way it is now.
I have done some coding over the years and I am very much familiar with the complexity of it. It takes a lot of time to make large changes like so many people want. Be patient
You are right. I am glad that this is being brought out now. I am a coder as well and this will take some time to do. But this needs to be said because I thought the point of 2.5 was to gradually move blender toward a more broader user base.
I'm not a coder (yet), but I can imagine the dimension of the work. Anyway, it must be done, patiently. A great job was already done on the transition to version 2.49 to 2.5.
Thanks Ton. I Totaly agree with you!!!!! ..just my opnion..change blender interface to sell some beginners tutorials will not help it at all!!
like!
This is not about selling beginners tutorials, It is about bridging the gap between artistry and technicality. Andrew Price would probably actually lose out on a lot of money on his tutorials if this ui scheme took place because he taught things like the Nature Academy and the Architecture academy inside of the scheme of the current interface. Andrew is proposing this to improve and modernize the workflow.
Andrew will earn more money when more people start using blender. Most of the stuff in his tutorials is about the 3d work. This is even more true for cgcookie tutorials. Jonathan's tutorials are mostly about mesh work and topology. A simpler interface will not reduce the need for tutorials.
Actually, making a more beginner friendly menu would reduce the need for tutorials! I know I've lost people when trying to introduce them to blender simply because the tutorials they needed for simple tasks didn't exist in their language. Most never get past the right button select issue, let alone learn enough to set up rendering in cycles. Having simple, always in use tools front and center would get rid of half the tutorials, and user friendly defaults (which you can change right now but require you to know they exist) would kill another third of the tutorials.
The serious discussions on UI development cannot just be washed away with a "paid shrills" argument.
I learned blender and 3d work by working through Jonathan's vehicle modeling series. 20% of that was learning about blender, but 80% were about modeling concepts like box modeling, poly by poly modeling, topology, modifiers, subdivision, creasing, etc. The claim that tutorials would not be required when the interface was changed are simply ridiculous.
And please, don't claim anymore that most users cannot get beyond the rmb issue, it is simply ridiculous.
It's not about changing merely for beginners. It's more about improving efficiency overall. If you listen to Andrew's proposal (most people simply didn't listen all that well), he wasn't addressing just trying to make Blender easier for newcomers to use.
It was more addressing and suggesting a way to make it both more accessible to new users of Blender (who aren't already new to 3D art, by the way), while making it efficient for all users.
Efficiency isn't so much a matter of beginner or experienced users--it's a matter of designing around how people think and best catering to keeping us thinking less about our tools and more about our goals in using the tools.
Of course, nobody's said that Andrew's exact proposal was necessarily the one developers might would settle on (it was an idea of something), but it was a good step in the right direction in that, to one degree or another, we need to rethink Blender's user interface.
It's simply not efficient as it stands, and longtime users are confusing their adapted to Blender's UI as being the same as it being an efficient-enough UI. It doesn't necessarily mean Blender's going to undergo a total reconstruction, but if that's what proves to be more efficient, it should be given a chance.
I personally think Andrew's on the right track with his suggested design. I've been using modo, which uses a similar nature of tabbed UI, and I think tabs work brilliantly--try the 30-day trial of modo. But where modo could take some points from Andrew's design is in not keeping all its options out in the open, which makes modo look like a wall of options. Keep primary/greater-used options open while keeping secondary/lesser-used options as drop-down menus is a nice idea.
Currently, you have to navigate between one mode, two sidebars (Toolbar and Properties sidebar), two subcontext menu (Materials and Textures), and two windows (3D Viewport and UV/Image Editor)...just to start simply start texturing.
This was just one example. There are several issues like this in Blender, all due to an un-synchronized and ad-hoc way of developers contributing to Blender. There's no singular, cohesive manner towards overall development with Blender to keep it focused and organized in a direct and efficient manner.
A new streamlined UI can change that around. And as new developments are heading to Blender, we're going to need it sooner rather than later--it's already getting crowded with the current UI, which, by the way, wasn't truly proposed as the "permanent" Blender UI to begin with (it was done as a step-up from Blender 2.4x, but it was always a work-in-progress).
People complained about the current UI when it first arrived, and people adapted to it. This current UI was more efficient than the 2.4x UI in many ways, but it was by no means a final solution. I think people should at least give a new UI a chance with Blender, because this time, they might not have to adapt merely because they have to--they can have something they can adapt to because it's efficient.
Totally agree.
Andrew Price never proposed a "back to the drawing table" option. His propose is continuing the advancement that started with 2.5 version. This will be good for new Blender users and skilled ones.
As you say, new Blender users aren't necessarily 3d newbies.
Thanks Ton. Thesecalm and reasonable speeches are what makes you a great open source figurehead. Much better than Linis' rants! I'm really part of being a small part in the blender community.
Damn touchscreen keyboard. That should be Linus ofcourse.
That was really informative and inspiring. As blender artists we are really lucky to be privy to the internal development and vision of BI. Its just not possible to have the same sort of relationship to the software made by a corporate entity. I may not be able to make it to the Blender Conference, but I want you theres one artist here cheering you on!
I would also add that think some of Andrews suggestions have merit, and would be compatible with the current UI scheme.
Excellent to see Ton give such an eloquent response. Helps us all to feel part of the larger discussion about Blender's direction.
I do worry from his (and others') comments that Andrew Price's suggestions have all been taken as dumbing down the interface. Just because it looks a bit like a ribbon interface, doesn't mean that every idea in it is about making things look nice but be less effective. There were some valuable ideas in Andrew's proposal that could be implemented as part of the current UI that wouldn't take away any control/complexity from the user but would make the system more efficient. Don't throw the baby out with the bath water!
It's all good constructive discussion about my favourite software, so hurrah for that.
And by the way: please consider becoming a Blender Foundation contributer. A bit of money can only make it easier for the BF to develop all its ideas, including the UI.
Well, this pretty much sums up the problem with Blender.
It comes right from the top: "I’m not much interested in “getting more users”. "
By all means, stick with Maya. Your copy IS legally licensed, isn’t it?
Do you think the problem is that Ton care about the current users that like blender and use blender in their works?
The problem will be that Ton won't care about the current users...
I agree, the focus really should be on making Blender better for those who actively use it. When you make Blender better for current users (which does include UI improvements), it becomes more attractive to potential users, I think.
IMHO, fixing inconsistencies within Blender's UI is part of helping current users, as is cleaning up some of the cluttered areas (but not by simply hiding the clutter in a dropdown menu). These are improvements that newcomers are likely to benefit from as well.
You are right! Yet what was the point of 2.5 with the maya and 3dsmax mode. If you are going to overhaul the ui scheme to bring in more users why not do it some more to streamline the process for artists.
I totally agree with Ton, the UI is already good like it is, changing the interface is not a priority at all, there is plugins and preferences for that !
That's a terrible attitude to have. Blender's UI is broken. It has always been broken. It's less broken now than it was before but ti's still pretty borked. It's not about dumbing it down or getting new users to me. To me it's about not having to fight with the UI when I'm busy trying to create.
Do you have any examples of what is "broken"? If Blender's UI is broken, I would not expect to see so many great pieces of artwork being put on the BlenderNation header, on a daily basis.
On the other hand, I do agree that there are some improvements to be made, which should hold at least some priority. But broken? I really have to disagree there.
When people accomplish great artwork in Blender, it is in spite of the UI not because of it. The UI is disorganized and unintuitive. Mind you, I'm not some insult-and-run griper. I'm coming at this from the angle of a professional artist. I use Photoshop, Zbrush, and a bevy of other programs. I used Maya for a long time before I switched to Blender.
I love Blender. It has become integral to my workflow. But I would not mind seeing a better UI that facilitates creativity rather than hindering it. If my choices were leave Blender or deal with it's brokenarse UI, I would stay with Blender. I honestly like Blender that much. But as I said, I wouldn't mind a change, specifically I wouldn't mind seeing it be better organized and more streamlined.
Fair enough, I think we can agree that good organization and a streamlined workflow are things to work toward. My main concern is that streamlining work doesn't focus on just one particular workflow (which is, in my opinion, one failure in Andrew Price's widely known UI proposal).
Perhaps it is simply because I spend most of my time plugging together compositing nodes in Blender, but I rarely have issues with the interface, outside of cluttered panels and dropdown menus.
I am also a professional artist, but the other apps I use are 3d. All of them tackle the same issues; they are trying to do many things. I use Blender whenever I get the choice *because* of it's super fast UI.
Calling Blender's UI broken is just trolling. If you don't like it, use something else.
@ Beorn:
Currently, you have to navigate between one mode, two sidebars (Toolbar and Properties sidebar), two subcontext menu (Materials and Textures), and two windows (3D Viewport and UV/Image Editor)…just to start simply start texturing.
If that isn't a broken flow of efficiency (it's literally broken up all over Blender), I don't know what is. It's the reason why, though I want to do my 3D texturing in Blender, I just can't rely on it repetitively as a workflow.
The current UI is rather broken, in the sense that it's not at all designed around efficiency, but rather around just finding a spot to put things. This UI is only getting more "stuffed" as new feature continue to be added.
Scrolling horizontally to access a vertical Property window? I can't even place the Property window below the 3D Viewport without it taking up a third of my screen--space better used for the 3D Viewport.
I say that as both a longtime user and someone starting to develop for Blender.
By the way, you're carrying a terrible attitude. One can call Blender "broken" and still like Blender. I love Blender, but I have no problem.pointing out its issues. The best thing you can do for something you love is to be honest about it.
But hey, if (generally) no one wants a friendly and long-due consideration about the issue of a new (or, at least, improved) UI for Blender, that's fine. Just nobody complain about the industry's reasons for not adopting Blender more.
I don't know how I'm supposed to reply to Beorn in all this jumbled mess but I'd like to point a few things out:
1.) Just because I have a differing opinion than yours doesn't mean I'm trolling.
2.) Not only did I already mention that I like Blender but I also said that if it came down to dealing with Blender's crap UI or going elsewhere, I'd stick with Blender because I like it that much. So you might actually want to try READING people's comments before smacking your face into your keyboard reflexively.
@Brian Lockett:
Horizontal scrolling to access vertical panel? I really don't understand where you would have to scroll horizontally
@ Nomis Animations
Go to your Render settings. Are you there? Good. Now, go over to your Particles settings. Did you have to scroll-wheel horizontally to get there? Yeah? That's what I mean. Vertical space, horizontal navigation.
It's not always a problem, but sometimes, it is. Never mind that it's completely foreign to a new user--even as an experienced user, there are times when that layout is in my way.
Most particularly, when I'm using Cycles and I'm forced to widen the Properties window in order to read my texture file names better. Or when I'm in the UV Editing layout, and the Properties window isn't even available anymore because there's no room for it among the split screen.
You can customize the layouts, but then you're trying to place that whole Properties window elsewhere in your Blender layout. Sometimes, I need it out of that vertical slab to the right, but it's a struggle to have it much anywhere else.
Of course, some can just duplicate the application window of Blender, put that second duplicated Blender application window onto a second monitor, and just set the Blender viewport into being the Properties window, but not everyone has a setup with two monitors readily available.
Plus, after many hours a day on the computer, that'd more mousing around between two monitors that my wrist cares to handle regularly.
But how can you call it broken when it works. I find it downright insulting to call it broken, especially when you use blender professionally. It is alright to critiize blender, but calling it broken is not constructive.
@ PhysicsGuy:
Just because you can ride a bike with a loose seat doesn't mean it's not broken. "Broken" doesn't always mean "out of commission," but rather, "needing to be fixed better."
Though, I for one haven't called it "broken" to insult the developers or the product. I have called it as such to address the nature of the problem.
Not only does the current system not work as efficiently as the software could have it, but it's quite literally broken all over the place with many things. Set up a simple texture for texture painting, and you know this immediately.
I have stated elsewhere that (for, at least, my comments) this is in no way an offense made against the developers--in fact, if anything, the current ad-hoc way of developing Blender's features and UI haven't let the developments of the developers shine in the best light.
Dynotopo, for instance, is a brilliant development that doesn't get its true shine among others beyond Blender users because it's hidden behind software with a "broken" UI.
i wouldn't call it broken since it does still work but it is a cluttered mess that could certainly use some better organization.
It is not broken. Broken means something doesn't work. When my mouse is broken, it doesn't work anymore. When a car is broken, it doesn't drive. The blender ui is simply not broken, because it works. It might not work the way you want it to work, but calling it broken is simply wrong and it is fud.
Obviously I was using the term "broken" hyperbolically, genius. Blender's UI might not technically be broken but it does not facilitate creativity, it hinders it. From the perspective of someone who works professionally with A LOT of programs, I can say that Blender's UI is not very good in my opinion. It is disorganized, cluttered, and unintuitive. If they didn't already know this on some level, they would never have overhauled it before.
I love Blender. I'll stay with it even if they don't change their borked UI. But quite frankly, I think they should change it. While I don't agree with everything Andrew Price has said, he definitely had some good ideas and he brings up some good, incontrovertable points.
i have a feeling that you read what you wanted to hear rather than what he actually said.
In sum, Blender is not interested in appealing to the Poser/Daz/SketchUp crowd, because that way lies mediocrity.
And I agree.
Whoever said a new UI was proposed in the interest of appealing to the likes of those programs? A new UI is about improving efficiency for any user--not about appealing to making Blender "for Dummies" ready.
Man, I don't mind if folks disagree with the issue, but sheesh, I wish they wouldn't rely on fallacies as their counterarguments.
Have you actually read Tons article? He did not say "we don't want to change anything", he just said that a full UI-rewrite could not be done in a reasonable time, even if they wanted it.
And Lawrence didn't even mention the UI in that post.
a) The entire issue is about the UI proposal. That's been the context here.
b) Not even Andrew Price's suggested UI design was a full UI rewrite. Things like the Node Editor and Compositor still functioned the same way as it already does--it was merely the accessing front-end features that got the UI rewrite. Much of it was more a re-skinning than a full-out core rewrite.
c) You can do a redesign in a reasonable amount of time, esp. if you summon the likes of a focused crowd-funded effort towards the issue.
d) Lawrence was in fact referring to the nature of the UI. In referring to the likes "easy-to-use" solutions likes of Poser, Daz 3D, and SketchUp, he is by both implication referring to the nature of how he feels the UI proposal trying to appeal as.
Well spoken: it's almost all about the skin of the final-delivery AND NOT ABOUT CORE. If Blender team just organize tool groups, uniformize the language and provide better progress-show devices, it will be more than 80% of users needs.
Even so, it's a lot of work. But Blender team seems not interested...
everyone here agrees with Ton, i don't.
I can appreciate what he says, on the philosophy of the core of blender, but IF the ui is really so linked to the core... well, a huge redesign has to happen real soon.
After reading Ton explanation i want to examine the source code to better understand it. I really consider a gigantic error link two separate part of a software, the ui and the core, in a mortal agreement between them.
Core is core, ui is ui. The team that handle ui should be able to change it without even consider to touch the core of the software.
This has to be true especially using a side language like python.
Other options means that the intrinsic design is a total failure.
So, if i have understood well this point, the blender team should reconsider as a priority point a separation, a total one, between core and ui, because usually this kind of mixed concepts tend, by nature, to complicate themselves as time goes by, so it can only be a bug creating monster code.
If i have misunderstood this point, so well forgive me.
Anyway, a good re-design for the UI is not a bad idea, and having more user should not be considered to have a bunch of idiots around a good software. one of these idiots could be the next james cameron (not many people knows that one of the first james cameron film was made in italy and was a very very low budget film that became a disaster... was he an idiot? i don't think so. maybe it was simply a noob).
A today noob can be the artist of tomorrow, if you give him the tools. Ok he can study, he can search youtube video tutorial and tons of usually old document on the internet, but a good ui can save hours, and this without being stupid, either for the user, either for the software.
“ i want to examine the source code to better understand it”
Maybe you should have done that first before commenting...
maybe i have also something else to do.
Maybe you shouldn’t say you want to do something when you have no intention of doing it.
maybe i have intention to do it as soon as i finish my queue of works. or do you want to judge also my works?
I completely agree. The UI needs to be fixed. It is broken.
I am curious how much you know about programming. When you say that "Core is core, ui is ui," are you speaking from experience coding things, or is that your opinion?
The OP is correct, and I do program using .Net languages. A UI control such as a drop down or value box just returns a float, an integer or a string to the underlying code. Andrew Price's ideas in Part1 / Part 2 about re-grouping UI elements more logically *should* be easy, as it should not affect any underlying code that uses the values returned from those elements. Yes, extra coding would be needed to design new UI elements that worked differently or took less space, but the underlying code is unaffected.
for a software with the goal and resources of blender, "changing the UI *should* be easy" is not (and must not) be a requirement. If blender has been designed with this in mind, the code has to be ten times bigger in term of lines of code and the maintenance would be way too high for very little gain. Today the UI in blender is data driven for settings and command based for operator, and this is good programming. In ten years of UI programming, i have seen numbers of software and the more powerful ones are those which have not be rewritten for the shake of "better design" or "better UI" and it is paradoxically more easy to add new features or correct bug in those old softwares with UI tight with the core than other where it has been artificially separated.
data driven doesn't mean that you can't separate the two things.
even better data driven engine are used to separate effectively almost everything. and if the underground engine is well structured and reflect a plugin based approach than changing a ui is definitely a "joke"
definitely my programming experience, c++, c#, objc, asp.net, html, sql, OpenGL, windows, iOS, osx programming...
you have to separate in a dramatic way the core functionality to the ui interface.
Or maybe you are assuming that the basic operations on vector and matrices need to be linked to a particular button on the screen?.
you have specific classes for them, and from the UI you call that methods. this in a well organised code.
if instead, for mysterious reasons, you link math code to ui... well, your problem.
i will never do a nightmare like this.
to better explain
lets suppose you have a math class like this
class Point3D
{
double x;
double y;
double z;
public:
void setPoint(double tx, double ty, double tz);
etc. etc. etc.
// and the a nightmare method
// where you create a stupid dialog box to ask the user
// for values for the point... this is a idiot thing.
void setPointBasedOnUIDialog();
}
things need to be separated.
ui need to comunicate to core code, not to be part of it.
have you read this http://wiki.blender.org/index.php/Dev:Source/Architecture
and http://wiki.blender.org/index.php/Dev:2.5/Source/Architecture/Window_Manager
and see any line of blender's code?
yes i have downloaded and i'm analyze it.
But the fact is:
it is well designed the ui and it is separated from the core? ok so Ton has made a bad statement and changing the ui should not be so hard
the ui is not separated from the core?
so there is a bad design and changing the ui is a nightmare
Very simple, very straight
I'm glad Ton stays focused. That's not blender's problem it's blender's strength. His answer makes me feel even more embarrassed to be part of the userbase that had this pointless discussion instead of developing our own "mediocre" visions with this fantastic piece of software.
It is good to know Ton Roosendaal, and Andrew Price. I feel that no matter what happens blender will always be awesome because of it brilliant people involved.
Why is that the problem? Ton says that his first priority are people who are already loyal blender users. That is a good thing. It has worked for blender all this time.
Having the increasing of the user base as a priority is logical only for software that people pay for. And even then, it is a weird priority, because if having many users is your thing, why not make software that more people need.
I think it is rather arrogant to 1) claim that there is a problem with blender. Sure, it is not perfect, but it does its job better than a lot of software with much simpler functionality. 2) that you think that you have the authority to question Ton's leadership. If Linus Torvalds or Guido van Rossum would want to question it, they would have a right to speak up, as they have proven that their leadership works.
But neither you nor I nor anyone here has any right to pretend that we know better how to run and finance an open source project in a sustainable way, with a dedicated team of excellent programmers. And in fact you should be ashamed of yourself, because remarks like this might actually frustrate the people who make this possible, namely the contributors.
There is a saying in the Open Source community: “He who writes the code makes the rules”.
Any fool can have an idea—just look at some of the comments here and in related BlenderNation items for the proof. Anybody can claim their idea is good, and who is to say otherwise? But the proof of the idea is in the coding: if your idea is so good, why not show some actual working code to prove it?
In sport you have the term “armchair quarterbacks”, for people who think they can play the game better than the actual players, but who in reality cannot even get out of their armchair. Similarly we have “armchair developers”, who think they can come up with better software, but who cannot even find their way around a GCC command.
It reminds me of a column by Dan Simmons. He saus he never admits to being a sf writer when he meets new people at a party, because people always start telling him about this idea they have for a trilogy they still want to write. He says any writer has ten of those ideas for breakfast. Then problem is that a book contains thousands of those ideas and they all need to be consistent. Kind of like writing a UI in a way. We have seen thousands of ideas by what you would call armchair software developers and none of them appreciate what goes into actually implementing them in a way that would actually be an improvement. And the worst part is people who start talking about a rewrite of the code. Rewrites are what destroys software companies! A complete rewrite is almost impossible to do commercially, because all the people working on the new code will not bring in any money until the damn thing is finished. So how on earth should the blender foundation finance it? A refactor to pull the ui from the rest of the software would be beautiful, as I would like to recycle the UI, but I understand that in a project like blender, a separation between UI and operations is not so practical, especially when dealing with mesh editing operations.
Anyway, I try to write all my software as command line tools, preferably not more than a 100 lines of Python,, because UI design scares me.
What a bizarre response PhysicsGuy!
Put down the Cool-Aid for a minute and think about this.
Would Linus Torvalds or Guido van Rossum or Steve Jobs or Bill Gates or anyone else running a large enterprise say something as patently ridiculous and arrogant as “I’m not much interested in “getting more users”. “ Obviously if that is your attitude then you aren't going to listen to what people want, or more importantly what they need.
Sorry you can't handle the truth, but that attitude is one of the problems with Blender. Too bad some of you fan-boys can't see that. Oh...by the way I'm a Blender fan-boy myself, but I'm not blinded by this weird cult-like following you guys have for Ton.
I started using Blender in 2011. In this short amount of time, I've seen more resources become available and met more and more people who've begun using it. This makes me suspect that Blender's user base is both healthy and growing considerably. If this is true, then it means that Ton & Co. do not need to make radical changes to attract new users.
I also find it odd that the term "fan boy" keeps getting slung by those who seem to want to Blender dominate the CGI industry - I think of fan boys as the type that brags because such and such major movie used Blender etc.
Thank you (@speakingmute), all very good.
By the same token, it's not as either/or a situation as some make it out to be. Yes, Blender is for the artists, and some of the more serious artists do request changes. When the artists find that something could be done more quickly by making it take fewer steps, it is made to happen fairly quickly. When a complete rewrite of a tool is needed, it happens sooner or later and we get improved topology, knife and bevel functionality as a result. When the panels need reorganizing, this happens as well. It's been going on since version 1.0 and will continue to do so.
Everyone needs to take a step back and stop imagining that one person is going to impose a change. All changes have always been a result of artists working with the developers. The proposals and comments this past month are only meant to provide raw brainstorming for the mill. It's been discussed for weeks through nearly a thousand comments and replies, and all the positions have been staked out well enough that we don't have to keep trying to talk over each other anymore. It's all out there, now it's on the coder/artist partnerships to sift through it all and work something out.
I think what he says it's not he doesn't want new users at all. But he also doesn't want new users just to have new users.
If you're a freelancer , what do you prefer : 10 regular clients , that understand you well and are reliable. Or 100 clients that don't care about your work and ask silly things ?
I think he'd prefer quality user over quantity of users.
And that's very clever !
Indeed. New users who just want this new shiny version of blender are maybe not the users you want. The improvements to blender should be those that people already using it need. New serious users will find blender anyway.
Be careful! When Steve Jobs or Bill Gates say they don't care about new users, they are crazy, because people pay for their software. When van Rossum says he doesn't care about new users, or makes sense. He is making decisions for python in the way that make sense to him and the coders around him. He doesn't have to care about gathering more users,because people are flocking towards python anyway. Why? I don't know, but there are enough people that are trolling python forums about the indent thing and Guido doesn't care!
When Torvalds says he doesn't care about new users, it is false modesty.
he doesn't say new users would be a bad thing just that he isn't motivated to do things for the purpose of getting more users.
i'm pretty sure he was saying he'd rather do what he can for current users than users that may never exist anyway.
here's a quote from his post (it's towards the end that it seems most people didn't read to...)
"A lot of suggestions as have been done – also to make Blender “easy to learn” – really fit well within a powerful 3D creation tool too. Power users like toolbars or custom shelves too. Power users want efficient workflow and a UI that works with them and can be configured well. By making it work for them first, it will be something any beginning user would appreciate too, or will eagerly accept to learn."
and
"So! UI changes will happen yes. It’s a matter of improving issues gradually, tackling them one by one."
Agree with you, NightShift.
No software can underestimate the need of be adopted by a expressive number of users. The survive of a software depends on it.
A opensource and colaborative software like Blender have even more need of users. The reason is only one: money. We are on a capitalist world. Blender is free (no cost). But Blender maintenance and improvement has costs.
So, Blender Foundation need donations, but Blender can also get money with CDs, TShirts, and the experts (like Andrew Price) can put a lot of tutorials on the net with short advertisements to make cash. Andrew has his "Nature Academy" and now "Architecture Academy", both based on Blender. Other veteran users have their sites, blogs and 3d products.
The math is simple:
Blender users > consuption > money > maintenance of Blender team.
So, it's arrogance and folly despise the need of new users. Because there are a rotativity: all the time old users (for several reasons) leave Blender. They get a new job that impose other softwares, they change their areas to another which have no need of 3d software, they retire, etc.
And, let me say one more time, old users also want to save time. UI should be almost imperceptible, cause artists want to focus on their jobs, and not on discover on the internet "which is the damn shortcut for this operation".
No one is saying that having more users is not a good thing. It is just a bad motivation to change things. The BF wants to make blender more usable for the people who are using blender now. This tends to mean that more features get added. But it also means that sometimes there needs to be a timeout to change a few things around to make it more consistent. But thisshould be done for current users, meaning us, not new users. And when blender gets a little more awesome, new users will come anyway.
totally agree on this
Bravo Ton!
A most reasoned and thoughtful article.
Chasing the lowest common denominator always leads to mediocrity.
I have learned Blender over 4 years and fought with learning both 3D modelling concepts and the blender interface.
Your thoughts on keeping people who are committed to learning Blender and using it, rather than a simple hit counter of users is a great insight.
This is the difference between special ops training and human wave military assaults.
Despite Andrews obvious good intentions, his main interest seems to be in producing still type pictures, not full movies, which is surely what is the main purpose of Blender.
There are many other options for this activity (Photoshop?), so why not let Blender run to its strength - making movies and animation?
In the end - there is a great urge to learn everything in double quick time, while we grow impatient of having to put effort and brainpower into training ourselves well.
Does anyone really think that changing the UI is going to relieve the newbie of having to learn about the myriad principles and concepts involved in 3D modelling?
Lets face it - learning 3D modelling is hard - only a poor craftsman blames his tools rather than himself.
If you are going to say that, then you need a better video sequencer because it needs drastic improvement to be a capable one.
I am user of blender for 5+ years and I believe that Andrews points have merit.
Interestingly enough, a good deal of the points were originally brought up by William Reynish in 2008 when Blender 2.5 redesign was in the works, just with added usability ideas that were not yet popular at the time (and apparently still not popular with some). Progress bars for one, those would have made 2.5 so much better ran some random numbers (that switch between frames and percent depending on how the tool feels at that point.
Was wondering if anyone else remembered this... He had another look in 2011 as well (see his blog) where I think he politely suggested a "clean-up". The main point from that post was dealing with the amount of clutter. Worth a look.
I don't agree that blender is mainly meant for video. Stills are equally important and a source of income for many people. Commercial artists making cover art that scientists send to nature of science or other journals usually cost about 2500 euros. In the netherlands it's always the same guy making it, so there is a market for blenderheads there!
For the rest, I agree with you. 3D modelling is hard. No reason to pretemd it is not. We need to find all the little inconsistencies in the UI and get rid of those. That is what is important to me. I don't mind learning a complicated UI. Hell, I use VI. But I want it to be as consistent as possible.
"only a poor craftsman blames his tools rather than himself."
very true but a well organized work space that has what you need where you need it also makes work easier and faster.
it's one of the staples of lean manufacturing.
no i don't think a better organized gui is going to help the newbies learn 3d any easier but it's certain to not complicate the process like trying to learn in a cluttered mess will.
I think 3d software should be very easy, cause we leave in a 3d world. Years ago I worked with Amapi, and it was quite easy. Nowadays I work with Blender and Vue. And Vue is quite easy for me, although it has resources that Blender doesn't have (yet). And Blender have resources that Vue only dream of having (for now). And I really consider Blender UI very unfriendly. It was a lot worst before version 2.5. As Andrew pointed out, the current version is a lot better than before. Now is time to - patiently and progressively - improve and organize the UI even more.
After Reading Ton's Answer, I felt that he is sincerely deeply touched by the fact that people are not thankful for the work already done so far, All those hours of tears and sweat spent coding to make BLENDER a state of the art, are not valuable against a -some how- weak UI design.(this is personally what I understood).
Dear Ton, if this is really what you wanted to express, then let me inform you that you missed the point. In fact the Blender Users Community are so PROUD of Blender as a package to the point, to ask its maker to push-up Blender to the next level.
Really this is a fact, nowadays artist from around the world using Blender achieved a level of perfection only achievable by commercial packages. Statistics showed that Maya and 3dsMax users are starting to switch to Blender because of this reality, just follow on YouTube the number of followers that watched video tutorials about "HOW TO SWITCH FROM Maya/3DsMax TO BLENDER" and it's going up.
Dear Ton, let me share with you a secret: Know that around 75% of Maya and 3DsMax users on the world Basis have a cracked version of those software, Autodesk knows this, but it leaves it as it is because those 75% of people are a user base that promote for Autodesk 3D softwares indirectly. It's because of them that 3DsMax came from the shadow an started rivaling with Maya. The clever idea here is when those 75% became addicted to Maya or 3DsMax, they found themselves obliged to buy a legal copy to produce something sell-able, or they influence their employers to buy a copy because it's only what they know how to use, and they have spent so much time to learn them then the idea to change them with other Soft seems out of this world. This is a marketing ruse used by Big companies.
Now let's comeback to our beloved BLENDER, Keep in mind my friend Ton that the people started the question of the UI change are all "artists" not "coders" and the only way to them to participate in developing blender is on an artistic point of view, and the only artistic side of Blender directly clear to users is the UI !!!!
So their intention is good, but without knowing they touched a sensible point, in fact as you said the UI is so tied to the internal code to the point of changing just a little portion can scramble the whole Core code.
So for me they are right and at the same time they are wrong. They are right because Blender really needs a Good UI kick on the Ass;) , and they are wrong because it is not the perfect time to start it. Now how to fix this without making anyone angry???
THE SOLUTION IS WITH ME.
I have spent a lot of years on/off (since Blender 2.49)trying to hack on Bender code in the shadow, my goal was not to participate in development, but to understand how you cast your UI design idea on this kind of huge project code (the MVC model). Over the years the Blender code became so complicated and uses a lot of languages to the point that Only the GHOST library is the perfect example of the CHAOS theory. The code takes several branches and proliferate from leaf to leaf, it accumulates code upon code...and so on and so fourth.. Really Hat Down for Bug fixers.!
So with this knowledge my proposed solution is as follow: MAKE THOSE ARTISTS BECOME CODERS !
likewise they realize what coders are facing. But the question is: HOW????
how to make an illustrator or modeler without any programming background become a C/C++ or python Coder???
My solution is: ALGORITHM. Yes Algorithm, instead of loosing time teaching artists how to write ideas in C/C++ or python, teach them how to write it in Human language in diagram form. Rectangles for declarations, Diamond for 'IF' branches, Square for 'Switch statement'...etc, I will take the burden to elaborate this system just for this sake. Then translate the actual Blender code to this form of language by taking each module apart, after that, produce videos and podcast by Veteran Blender coders to explain the idea behind each algorithm and how to chain them to each others. I will be in charge of translating the coder's way of thinking to the Artist's way of understanding using my developed way of teaching. At the end my goal is to try to separate the UI code from the Blender core code by the help of those newly grown 'Art-Coders'.
By knowing the tough task the Blender foundation is dealing with, People starts to weight their words before starting any Discussion about any subject. Because they can take a pen and a paper and try how hard is to integrate that Idea to the present state of Blender just by using Algorithm.
Dear Ton, the space here is not a much to explain what I think of about this situation. But be sure that I'm offering you a new way to change this World's BIG Blender Community and may be others!
Just let this idea boil in your mind and contact me if you are up for it.
And I Invite every Artist or Blender User concerned by the question of developing BLENDER's UI to show me if they are up to participate in this project not only by words and votes but by actual Algorithm and Code.
Yours
ORGANICOMAN
would love to learn code. Is there a good webpage or tutorials that would help me? I am a 2D guy interested in making games. I want to get my ideas out but know i need to know code.
Google something like “learn to program” and you will find lots of links. You should find detailed courses at MIT Open Courseware, Khan Academy and that kind of thing.
Be prepared to put in a few years getting properly skilled. Don’t expect to dive into the Blender source until you do so: at over a million lines of code, it’s not for beginners.
thank you Lawrence D'Oliveiro.
the new boston go to video tutorials. very good site or code academy
Thumb up for codecademy! If you are looking to use the either the Blender API or the Game Engine API run through their Python course. I did it in a week with no prior Python knowledge and I haven't looked back when using the BGE and have even made a few decent scripts (no UI tie ins yet) for automating some tedious tasks when modeling.
I would suggest http://www.processing.org. MIT-based, and very good as a learning tool towards programming. It was actually made specifically for that purpouse. :)
I do not think Ton is touched by the lack of gratitude. He has enough people around him to know what he does is greatly appreciated. He didn't receive an honorary doctorate for looking good in a robe.
I think he is communicating that appreciates the opinion of everyone who is seriously using blender and is willing to improve blender for them, but that getting more users is not his priority. It is that simple. Andrew and Jonathan have a different priority. For them, more users mean more people on their website means more revenue from the website. That doesn't make them bad people, it just gives them a different priority. As Ton says, he has a very well defined priority, which is the mission statement of the Blender Foundation. This mission statement is legally registered, which means that he, as chairman of the foundation has to work with that mission statement in mind. And that mission statement is not compatible with a blind drive towards more users, but rather with making blender a fully functional 3d software, with a full pipeline. That means those are the things that need to be addressed. This is what the open movie projects are for. They expose the weaknesses in the features and usability of blender in an accute way. If blender is good enough for project Mango, one cannot seriously claim that blender is limiting your creativity. Every 3d pipeline is struggle with the tools. That's why all studios actually have inhouse developers to make glue logic between the tools they use for mesh editing, texturing, sculpting, animation, color correction, etc., because there is no commercial tool that does it all. There people could easily learn to use blender, because their whole patchwork of internal glue logic is also not a particularly intuitive UI. It is not the usability of blender that is stopping the studios from using it. It is the fact that they have this whole pipeline carefully tweaked and they have invested a lot of money in it. The thing blender needs is a feature length project to show that within blender, you actually do not need all those scripts and external tools, but it is all in one package. That means better asset management. That is what Gooseberry will be about.
To come back to your point of having artists write pseudo code to lighten the load of the developers is plainly naieve. Code needs to be read and written!
People, People, let's stay logic and squash this problem reasonably.
From one side TON wants Blender to stay Rigid and result focused, on the other side Users want Blender to be flexible and attractive Power-horse (=pleasant+robust).
If we weight this two thoughts we find them right and wrong at the same time, But Why?
Because "Result oriented" is good but "Rigid against change" is bad, we know all that a stiff rod is exposed to be broken. The same for the other idea, "Attractive" is good but against what cost? if it cost me to sweep all the code and start from the beginning to make it attractive this is craziness.
So logically this is the problem, The two of them Are WRIGHT and WRONG at the same time, so if we keep arguing about it (like what people are doing right now in the forum) we will spent the eternity, because it is a closed loop.
Then My task, as it is yours (BLENDER USERS) to make this situation beneficial and not a traffic bottle neck , is to find a SOLUTION.
Let me start first by exposing my solution and its goal (in fact it is an explanation of what I said above, in my 1st commentary). To make you understand me better I'll use an analogy (example) I will take the CPU as my Troy Horse.
A one single CPU core uses a Technic named MultiThreading to simulate a shadow-clone of himself (like in Ninja art). So by using this technic a one single CPU core will be shown as Two CPU's in the TaskManager (in windows OS). [Just verify it by right clicking on your "TaskBar"( Win OS) and choose "TaskManager" then go to "Performance Tab", you will find the double number of your real physical CPU cores].
MultiThreading makes CPU double his capacity of treating operations with the same cost of time amount.
So how can we explore this idea and use it to:
1- Develop the BLENDER UI to some thing friendly like Andrew's suggestion.
2- Keep the development of BLENDER robustness steady and continuous.
The answer is by gathering THREADS ( Blender users) from the community and switch them to coding, then ATTACK the UI coding challenge.
Another question rises here, How can an Artist become Coder????
Let me tell you that it is the most easy transition that can happen in any field, here is my proof to consolidate my saying: Take Jonathon Williamson as an example, he started first by using Blender as a Modeler, then Texturer, then a sculptor to finish by a BLENDER Python Addon Developer. check his Web site at Blendercookie.com,
anyone can say: Yeah, his Python Addons work despite he has no coding background( Himself confirm that in one of his video tutorials).
So the transition is possible. From Artist----> Coder, [the other way around is already proofed on the field, just check Ryan Kingsley a developer at Pixologic Zbrush, he became a quiet talented sculptor after being an excellent coder].
So maybe some artist agree with me and ask me: Hey ORGANICOMAN, I'm a Blender artist and I want to learn coding to develop BLENDER UI, Teach me!
My answer is: How are you good in USING BLENDER????
yes, because here IS the hook. Your competence in using BLENDER qualify you more or less to understand the Blender CODE behind, it is like explaining a car's mechanic system to a VERY good DRIVER.
DO WE AGREE ?
So stop BLAH BLAH, and let's gather around a strong will to make this happen.
FOR ME I ENGAGE MY SELF TO SWITCH ANY VERY GOOD BLENDER ARTIST TO A QUIET PSEUDO CODE DEVELOPER, my only requirements are:
1- Be a very good Blender user (used almost every module in blender, from 3DView to the Compositor).
2- Have some understanding of Math and Logic.
3- Be really dedicated to change the UI look of blender by Writing his Code (or pseudo code precisely).
I'M NOT CONTRIBUTING IN THIS PROBLEM BY CRITICS BUT BY SOLUTIONS.
BE LIKE ME !!!!
Noobs will never become pros. Why? Just read the forums. Most of them are stucked at "why right click" and same "big" problems haha.
3d is not a toy. These guys will never make any serious with Blender. They will watch millions of video tutorials and stucked in "tutorial-land". Andrew Price needs them haha. Sorry. :)
When you started with Blender, you were by definition a new user. Does this mean that since you started as a noob, you will never be a worthy pro? Your logic needs work.
It seems that different people imply different meanings with the word "noob".
I my understanding (and most of the people i know seem to have a simiplar one), a noob is someone who does not want to learn anything but complains about everything.
However i have seen people writing on forums "I'm a total noob, but i want to learn this properly". So language and it's development over time may have made the term "noob" a bit to diffuse to be used correctly.
We all can agree that we don't need the "i don't want to learn anything" guys.
And me all agree that every new "I know i'm not a pro, but i'm willing to learn" guy is welcome to the Blender-community. However, and that is what Ton said, i may be that it is not the highest priority to get new users, and especially not the highest priority to implement features for people who don't use Blender.
If i think about that, it would be almost impossible to develop for them:
- How would the features be tested -> existing users
- who would design the interface -> developers+existing users
- how would the workflow be improved ->ask existing users
So, improve the UI (Andrew pointed that consistency thing out very good, and the developers are already collecting inconsistency on a wiki-page (they may have even been doing that before, i dont know)) for existing users, and the new users will have it easier too...
Unfortunately, this is so true!
How many of these guys who want to make a dumbed down interface do so because they are stuck in the basics, like the flying dutchman forever sailing the seas of right clickism.
My recommendation to escape this cycle of mediocrity is to design and produce a short movie or tutorial using Blender.
This will quickly make you an expert, or realize that you are not intended for the 3D moviemaking world.
I am continually amazed at how many people simply do not go the full 9 yards, and make a short movie, but stay stuck at making a still life scene or a ten second animation.
As endi says, if you want to think like a pro, take on a big animation project, then see if RMB is going to kill the project, or whether you have forgotten all about these little UI hurdles.
Really, if you are so obsessed about the UI, why?
How about thinking about the storyline, character development, and all the other things which make or break an animation.
This discussion reminds me of two sculptors arguing about whether one chisel is better than another instead of actually using them to create a sculpture.
Completely agreed regarding "two sculptors arguing about a chisel instead of actually using them to create a sculpture." That's exactly why I wish he would get back to continuing the Dead Cyborg game development instead of posing non-starters, such as that gem of an idea that pros never started out as new users, which is a concept that just plain beggars belief.
There are entirely too many false claims in general going around regarding this topic, with no clear end to it in sight.
"This will quickly make you an expert, or realize that you are not intended for the 3D moviemaking world."
What about the thousands of us that have no wish to make a 3d movie? I personally use Blender to make reasonable renders of robots I build in order to provide better graphics for my presentations than AutoCAD would give me.
Sometimes they should argue whether a fine tip chisel is needed or if they should start out by first learning how to get the right shape with a power tool before worrying about cutting in fingernails. In this case, it's more like deciding whether we should skip to the chisel instead of digging out the block first! Who knows, you might get stuck with granite instead of marble (and then your marble chisel will slowly break down under the pressure of incorrect use)
Blender is a toy according to many people. Some people laugh at Blender. Part of the problem is the UI flaws. Just because you have learned to be a contortionist to use Blender doesn't justify why others should. There are sensible things that could be done to improve the UI for everyone's benefit. I don't appreciate your tone to Andrew's well intentioned initiative. Be thoughtful but not insulting.
The reason why Blender is seen as a “toy’ is purely down to marketing. Look at all the professional work done by properly-skilled Blender artists for the proof.
What marketing? I think the reason it is seen as a toy is because it lacks care and attention to the overall user experience. The previous Blender UI made Blender look like a piece of junk. Over time stuff was squeezed in and chopped around any old how and it was almost inaccessible to a new user. Now we have a better base for the UI but stuff still gets put in without much oversight or finesse. Apparently Ton and a fairly large number of people want to do nothing about this preferring to keep adding new functionality that doesn't really blend all that well together. Eventually we end up back where we were, a piece of inaccessible junk that people rightly or wrongly derisively refer to as a toy. Even people who are fluent users or happily wallow in any given mess should welcome some general house cleaning. The UI is not all it can or should be.
Ton said Blender needs and will receive house cleaning. He just said that overhauling the UI isn't going to be the focus. I don't recall anyone arguing against Andrew Price's first two videos where he was pointing out specific issues with the existing UI; it was his proposal for a brand new UI that so many objected to. What a lot of critics here don't seem to grasp or acknowledge is that many people simply don't find Blender's UI frustrating or difficult - and those of us who don't aren't necessarily "memory gurus" or fanatical partisans for Blender.
People don't call blender a toy because of the UI, they call it a toy because they don't want to admit that something free can do the job just as well as an expensive piece of kit. As I said in other comments, pipelines in studios are not that straightforward and need a lot of internal work. A studio using commercial software still needs developers to glue everything together, because this software doesn't solve all their problems. Looks for instance at the Mocha planar tracking software. Their own tutorial videos show how you have to click a button to paste the tracking data to the clipboard and then paste it somewhere into the timeline of your compositor software. That is not particularly user friendly. Ofcourse, when studios started to use it, they demanded it be converted into a plugin so that they don't have to go through that routine anymore. But the original product did not have a good UI. It was the feature that sold it.
This is why I don't believe your argument. Usability is a big thing for amateur users moving into 3D. I am such an amateur, but a stubborn one.
I want to add a little thing, the famous "industry standard" of the last weeks, a standard really similar to the proposal.... Could somebody tell me what are these industry standards? because a lot of people speak about it but I don't see any standard in the rest of software. I don't know all the 3d suites ¿Ok? but could somebody tell me what software have this UI? Because Maya, Max, Houdini, XSI, Modo, Lightwave, Mari, Nuke, Zbrush, Mudbox... don't follow these "industry standards" with a "Task orientated" GUI and a ribbon toolbar.
Only Modo have a "task orientated" GUI but really different to the proposal because in modo this tabs make the same that blender's GUI presets, THE SAME. You can convert the model tab in any other tab in five minutes.
Where is this famous Industry standard that Blender GUI broke in all points?
Maya uses a ribbon bar and a side bar like Andrew is proposing
Do you know the difference between a Ribbon toolbar with task orientated and a toolbar with tabs? Because I'm seeing all your answer and I see that you don't know what are you talking about.
Sorry, in my country is 0:40AM. I cannot spent my time talking with somebody that don't see something obvious.
From past Maya user:
Maya's tabs are waste of time eating screen space. Try learning living without them and your workflow will speed up.
Amen brother!
I have kept out of this discussion because I wanted to make sure I understood where people were coming from before I comment.
That being said, I've been using Blender since 2.03 and the last UI redesign was a nightmare for me. I had to relearn things I had spent almost a decade learning.
The current UI is intimidating and full of inconsistencies. It's the main reason that I stopped experimenting with a lot of Blender's tools and use it almost solely for basic modeling now.
I don't care about it being "beginner friendly" or anything. I don't care if it attracts new users. I care if it's straightforward. I care if it makes sense. Andrew Price's version does just that.
I wish that the majority of the naysayers would just come out and say, "I'm afraid of change," rather than bashing a very clearly thought out UI concept which was backed by plenty of supporting information.
I would pay to have the option of a cleaner, easier to understand interface, rather than the cluttered mess the current UI has become.
I'm afraid of the change. I know how to use Maya and can afford to purchase it if I so desired. Maya vastly exceeds Blender in cloth simulation, skinning, and many other areas. The reason I use Blender? I love the UI in 2.6 - I don't have to fish through menus, use widgets, or deal with toolbars cluttering up the screen. I'll take the right click over the alt key any day. I won't say I don't want Blender to be more like Maya/Max, because I certainly would like to see more of Maya's functionality in Blender. But, aside from modeling and (Increasingly) sculpting, Blender's UI is one of the areas I think it's much better than the major commercial suites.
How much would you pay?
Say a skilled, experience programmer is available for $100/hour. How many hours would you pay for, and what would you expect to get done in that time?
I was being facetious.
I certainly have no idea. It would entirely depend on what would be changed (are we sticking to Andrew's recommendations or something else?) and whether I'd be able to apply that interface to new versions of Blender when they come out.
Would the new interface be a one-time thing, or would it be continually upgraded? How many people would be involved in crafting it? How many other people would be purchasing it?
Too many questions to easily answer in a short response, but there you go.
The UI has been redesigned once already from within the Blender Foundation. I simply find its implementation lacking.
I'm certainly aware of how much work went into it (okay, I can only guess, but it must have been a lot), and while I appreciate the hard work, the end result is less than ideal from my perspective.
I see no issue with talking about potential shortcomings as they pertain to the software we love so much. Clearly, from looking at the comments, other people have real issues with constructive dialogue so I decided to post a straightforward, "from-my-experience" reaction to the current UI and Andrew's proposal.
I really have little else to say. :)
Though I loved Andrew's new UI proposal with fanfare (and still do), I felt that it was good to see a balance of the community's usability as a whole -which programmers are a part of. They need usability as well. They gotta fix the stuff we break and structure it so that it's maintainable.
I felt Ton might have been better to let the conversation sit a bit longer before making a public response. His response reads defensive, dismissive, and derogatory of the criticisms and character of those making the criticisms the past few weeks.
Here are a few quotes that made me raise my eyebrow:
"Uncertainty and doubts lead to stress and it can make people become poisonous – usually without even knowing."
"The Blender team ... really isn’t dying for ideas or suggestions, nor do they spend time twiddling thumbs waiting for a brilliant request to be passed on."
" they’re not incompetent idiots either."
All of these responses seemed odd to me, but I do generally undestand some of Ton's frustrations. He is just being defensive where we were just trying to open up an honest conversation.
My biggest frustration; however, was: "EASY TO LEARN vs EASY TO USE" section. It reminded me of the couple who wants a divorce and asks, "Is it better for the children to have us separate or for us to yell at each other and not seperate?" and the answer is neither, stop yelling at each other.
Easy to learn and easy to use are not mutually exclusive; in fact one strengthens the other. Blender for me is neither easy to learn nor easy to use. I have been using it for about 8 years now, almost daily, and find myself having to look up online references often to remind myself how to perform simple useful tasks. (Alpha textures anyone?)
I find myself gravitating more and more to different software to replace different aspects of Blender for this very reason. Marmoset for rendering. Topogun for retopologizing and baking, photoshop for the combining of AO maps and Color maps (Can do it is Blender but it isn't half as easy), Zbrush for not only sculpting but for texture painting. I have even used Photoshop for texture painting, and Photoshop's texture painting is HORRIBLE.
I again understand Ton's point of view, but I think he was a bit dismissive of others points of view.
Just my two cents.
Definitely I think ease of use takes precedence over ease of learning. Remember Blender was originally created for in-house use by a professional graphics house that lived or died by the quality and timeliness of its work. So smoothness of workflow and high productivity were paramount, and I think in these respects Blender succeeds admirably.
Lawrence, Photoshop was also originally an in-house project at ILM (based on work for displaying images by a PhD student as most great programs are), yet the program is one of the easiest for beginners to learn, far more intuitive than GIMP. Hell, the graphical user interface idea itself was an in-house project at PARC, and we all know how that story ended. Just because something started as an in-house professional product doesn't mean it needs to stay that way. It also doesn't mean it needs to lose sight of it's original goals, but those two are not mutually exclusive.
Oh, but it wasn't :) Please re-check the history of the project.
"Oh, but it wasn’t :) Please re-check the history of the project."
By the same technicality Blender itself was never an inhouse project either, only the tools that would eventually become it were.
On the subject of "easy to learn vs. easy to use," I'd like to point something out. While the two aren't mutually exclusive, and I don't think anyone said that they are, you run into difficulty when you try to do both and still pack in the necessary advanced tools.
Take Microsoft Word for example. Word has a "user friendly," "easy to learn" UI, but the cost is hiding a lot of advanced functionality. Did you know that it can actually do some good-looking (though simple) curve-based 3D modeling? I didn't know about it until just recently, because it is very, very hidden.
And there is the balance Blender must try to find. How can we keep the UI as simple as possible, without compromising on the tools? I for one don't want to see all the advanced features that make Blender such an awesome piece of software to work with hidden in obscure menus, for the sake of keeping the rest of it "easy to learn."
There are many ways around the issue, but as in Office, the advanced functions are best kept hidden but with good documentation. There should be no reason why simple selecting and moving an object is behind two different menus with a total of three levels (technically there's a "translate" button in the object tool, but everywhere else it's called "grab/move", so it might as well just be a different function), while cell fracture has it's own button. Just like Office lets you make your own custom toolbars for the ribbon, Blender should strive for default simplicity with customizable complexity. After all, Blender has one up on word processors because you can have an extra 50-60 keyboard shortcuts for advanced users to work with (since there's no typing involved, practically any key press is a shortcut).
I remember once Microsoft admitting that something like three quarters of requests for new features in Office were for ones that were already in there.
Trying to use Microsoft Office as some kind of exemplar of good UI design is, shall we say, a trifle optimistic?
Nobody said Office was good design, rather that they had good ideas for the reasoning behind their design. Taking their claim out of context is a clear sign you wish to promote one argument without reasoning behind it. I went ahead and looked the quote you mis-quoted (actually 90%, but read it all):
"PressPass: Why has Microsoft decided to change the user interface in this Microsoft Office release?
Numoto: While users have generally been very satisfied with the existing versions of Microsoft Office -- and the consistency of the user interface over the years has been a big part of that -- we have heard from customers that the current user interface and toolbar system had grown difficult to negotiate. As we’ve added more and more features to the Office applications, it’s become a challenge for people to find the tools they want to use. For example, Word 1.0 only had about 100 commands and users could simply click through the menus to see all the features and tools. Microsoft Office Word 2003 has over 1,500 commands, many of which are admittedly difficult to find.
In a recent customer survey, we asked users what features they wanted from the new version of Microsoft Office. More than 90 percent asked for features that were already available in Office. This made it clear to us that we needed to address the issue of “discoverability” in the current user interface. In addition, we wanted to provide new features and found the old user interface couldn’t easily and clearly expose some capabilities, especially those involving diagramming and art. And finally, we wanted to give it a redesigned look."
Yes, Office redesign was to REDUCE the number of "wait, that was in Office?" responses, and the 90% was due to old menu and toolbar design. Blender needs to do a similar thing and ask OTHERS what is wrong with their program. Long time users can't see the failings because they tend to have worked around them, just like new users can't see the benefits to why some things are they way they are. There is no easy answer to the UI issue, and claiming that Blender is good because it works for some is not a valid answer, ESPECIALLY for the development team.
“Nobody said Office was good design” — then what is it supposed to be an example of?
"then what is it supposed to be an example of?" It is merely an example of hiding inner complexity and feature fullness with a simple user friendly interface. It is most certainly not a bad example of UI design as you are imploring, nor are any programs based off the same usability principles.
But is that an example of “good design”, or not?
“The Blender team … really isn’t dying for ideas or suggestions, nor do they spend time twiddling thumbs waiting for a brilliant request to be passed on.”
” they’re not incompetent idiots either.”
This may or may not have to do with the current discussion, but if you look on the mailing lists, every two weeks there is this guy who knows best how to change the whole Blender for his specific task or just suggests features that would take years to develop with the expectation to see it in the next release. Or the other guys that suggest features/complain about missing functionality that is already there (apart from improving the UI, improve the documentation and the visibility of the documentation). Somehow, i understand these statements by Ton very good...
"Or the other guys that suggest features/complain about missing functionality that is already there "
This is actually the number one reason for UI overhauls. When Microsoft did a survey and found out users were asking for "new" functionality that was already there at a rate of 90%, they decided it was time to rethink how their UI worked. Blender too should have a fairly extensive survey to better understand the programs strengths and weaknesses. I bet even Ton would think twice about his stance if the survey shows what some people fear.
Unfortunately, Microsoft’s idea of a UI overhaul was to introduce the Ribbon, which fits in quite poorly with modern widescreen monitors.
Compare the OpenOffice Sidebar, which works much better. Or even Blender’s moving of the default Properties window location from the bottom of the screen to the side.
These are examples of UI overhauls that work.
I have one major question on one of the core points of his argument.
"The misconception here is that the “UI” is some kind of separated entity of a program, which can just be peeled off and replaced with another interface. However, a really good UI (and certainly Blender’s) is not just an abstract independent layer, it’s the reflection (and sometimes even the core) of the program’s design."
So, what is the program design? From the current set of UIs (can't even consider some components to be the same interface), Blender has no one design goal. The biggest issue Blender has right now is not ribbons or vertical tools and definitely not professionals or beginners, it's a unified design language.
For example, you select with right click in displays but select with left click in menus and popups (save, preferences,etc). Why do some simulation functions show up in the modifiers section but have no way to delete them, yet others are entirely in that menu, and still others like rigid bodies don't even show up? Hell, zoom is wheel in 3dview but control-wheel in console and control-middle button in menu panes!
"On the other hand, our project is flourishing and highly successful already. Here’s some of the actions we take now:
...
Expand the ‘foundation’ team (developers supported by donations and Development Fund) with people who can code on UIs and usability"
We currently have coding layout language, so why, after a decade of development, is there still no unified user interface design language? Perhaps rather than throwing more programmers at the issue we should step back and decide what the core functionality of blender is, and how to best improve usability without sacrificing features. Only after we know how Blender should work and understand how to apply that to a unified design language should we throw more resources into developing the UI. More than any other thing you can possibly do, a unified design language will improve usability for all users, from the first day to your last big project.
Go watch Sebastian König’s video explanation of how, in fact, left- and right-click in Blender are applied a lot more consistently than you might think.
König only mentions that it is clever, but occasionally does say it's not consistent, just check his conclusion at the end of "About RMB".
While I agree considerations must be made for blender, some actions make no sense in the context of the primary goal and therefore inconsistent even if it's consistent on a technicality. In timeline for example, the primary goal is moving back and forth through time in your scene, so while you can think of the time indicator as a cursor, the rest of the buttons move you through time with the left button while the middle is right! I am not fighting for or against RMB-select (I have one machine set up for left when showing others blender, my personal machine has right), I am concerned that Blender has no discussion on how to define what is the meaning behind certain functions that are taken for granted by longtime users but drive new ones insane. König makes that clear by defining what those functions do for him, but there are just as many uses of blender as there are users, and the most important thing to do is discuss why things are the way they are and if there's a reason why they can't be any other way (and no, "because that's the way I learned to do it" and "because it'll take me time to get used to a new way" are not valid reasons).
Ton say : "BLENDER IS FOR ARTISTS"
I think there are some small points in the UI and the workflow that make this not fully true.
Sometimes you are stuck because you try to do something but it does not work and you don't understand why.
For example : in the compositor window, to use the backdrop , you have to click 'use backdrop'.. but you have to add a view node to have the backdrop displayed.
How to guess this as you have no warning message when you tick it ? So you loose time because you think you just have to tick the option.
Is it good for creativity ? To my opinion no, because first you think blender is bugged and it breaks your creative workflow.
I agree that once you know it, it is ok. But you have to learn this type of small things that are not obvious in the GUI (and if you don't use it often you may forget it).
Andrew pointed out this type of issues (discrepancies, lack of 'error' or warning messages). This is the type of interesting points he mentioned about the GUI.
So no need to rewrite the full GUI. Probably just doing some reorganization of some panels with the existing code, and progressively improving the small things I mentioned above.
I don't think anyone suggested a full GUI redesign, just useful improvements and a better defining of just what Blender GUI means. Blender currently lacks a well defined (or even poorly defined) design philosophy, so technically any solid definition of what it means to be Blender will be a "full rewrite"
I agree. Blender got a better UI since 2.5 version. But, now is time to prepare more improvements. It is not a "back to the drawing board" propose, but an organization and a "friendlyzation" (lol) of the UI.
A painter don't want to know about the chemical composition of his inks. On the same way, a lot of digital artists don't want to know about codes or mysterious shortcuts like "Ctrl-Shif-Alt-Tab-Z" or something. They want to get absolutly concentrated on the work, and not on the tool. The best tool is that one which can stay on the backstage.
There is no need of any special actions towards "friendlyzation". Improving things as they are planned to be will make blender easier to use already.
As you said - concentration on the work is what's important and many recent changes go towards that goal.
As for painters - a bad example :)
I know a lot of painters and am a painter too myself. Even the most humanitarian serious-hobby painter knows something about chemistry of their paints. A real professionals for most part know almost everything not only about paints they use, but related to their craft at all.
I agree about the painters: at least, some chemical knowledge is necessary for they. But if they get in trouble because some chemical difficulties, they simple will think: "what the hell. I'm not a chemical, I'm an artist. And probably they will change the ink brand.
“A painter don’t want to know about the chemical composition of his inks.”
Yes he does. You get very different results with oil paints versus watercolours versus acrylics. Any skilled painter learns to understand the difference, and how to best use them.
Art depends on technology.
Perhaps many who complain about the UI are those trying to use Blender on a notebook? Blender, like all Adobe applications, are a pain to use without a second monitor to hold open all the panels you need. My second monitor is entirely filled with panels in both Blender and Adobe CS apps. The main screen remains uncluttered. My sub-panels are all rearranged to fold up the areas I never use, leaving only the controls I need. And of course, I use several different Screen layouts all beautifully setup for different tasks. I guess many complainers are just too lazy to customise Blender to their requirements.
Incidentally, I've recently been using InDesign, and like many Adobe apps it has a far worse interface than Blender. It consists of hundreds of microscopically small toolbar icons to control many functions, and the most absurd menu system in existence, with 197 items listed by default... and that's before moving onto any > link to expand a menu to the right! I spend more time reading the menu than reading my document :-)
It's always delightfully refreshing to spark up Blender, where the UI gets in the way of my thought process far less than any Adobe app.
I don't use blender on two monitors. I only use one. The second one is used for reference images.
" I guess many complainers are just too lazy to customise Blender to their requirements."
Not lazy, inexperienced. Hence the DEFAULT setting changes. Blender was built on customization, and an experienced user will quickly be able to customize the settings back to what he wants regardless of the new UI elements to make the program more user friendly. This is the place where most "advanced users" complaining about change get it wrong. Just look at AutoCAD and how their UI changed to help new users while the pros replied with a big "Meh, I just use command line anyway" and "I can always customize it back to how it was in 5min", there was very little backlash and none of the "our way or the highway" snobbery we see with blender
>Go watch Sebastian König’s video explanation of how, in fact, left- and right-click in Blender are applied a lot more consistently than you might think.
This myth again. You can use every argument that he uses for RMB Select in this video also for LMB select.
If even then this video is a video that shows the power of working with two mouse buttons and their corresponding functions. No matter if you select with rmb or with lmb. I work since years with LMB select in Blender. And miss exactly nothing. The benefit from rmb select over lmb select is exactly zero.
But the drawback is definitely a bit more than zero. LMB select is since a long time a standard out of the time capsule of Blender. I have seen and still see generations of new and even advanced users become stopped by this default rmb select in Blender. Even i as a long time Blender user could never work with RMB select. Take me away the LMB select and i am gone. The current rmb select by default is a loose loose feature.
To modify a quote from Ton (if everybody does it it must be stupid) :
Just doing it the other way doesn`t mean to be more clever.
This may be not about LMB or RMB to select, but also about 2nd MB for context menu or alternate selection mode...
A modern mouse has three buttons, you can easily use left select, middle cursor, and right context menu. Pros can easily change it to left cursor, right select, and middle context if they want, or practically anything else. To long time users, it doesn't matter what new defaults are, especially if they can import their old settings.
I am really suprised by this debate and some of the opinions I am reading.
I personally agree with Ton and really appreciate all the hard work that Ton has put in to Blender.
Regarding new users and the way some people refer to them. Everyone was once a new user of Blender. I personally used to program in machine code, assembly language, BASIC, etc on the old Acorn Archimedes and first used Caligari which became Truespace on the Amiga and Lightwave. When I first encountered Blender 2.3 (I think), I gave up and did not open the program again for another month. At first it was very difficult, but eventually I started to be able to use it quite well. Learning Blender is the same as learning anything else. The first steps are always the hardest. There are quite a Blender few resources for new users today.
Yes Blender is very difficult for new users and so what. Most programs are difficult for new users. I understand that Blender has an UI that is difficult to adjust to if you have not seen a program like Blender or Lightwave before. I would challenge anyone who would claim that the first time they encountered Maya, 3DS, Lightwave, XSI or Houdini that they did not have serious problems at first.
Andrew Price came out with an interesting suggestion. It is an idea like the thousands of other ideas for Blender, that have to be assessed and researched to find out if they are even workable, which quite often they are not. There may be a lot of aspects of the idea, maybe only some or none, that might be practicable. In the process of creating or evolving software. What can seem like good ideas, might not work and what can seem like bad ideas can fox lots of problems and vice versa. Evolving a program as complex as Blender is far from easy. I really admire how well Ton and his people have coded Blender. The code really is a work of art and I really wish I could code 10% as good as those coders. Blender is also an extremely flexible tool. It is not just good for modelling. Blender is great for special effects, compositing, architectural design, animation, etc. It covers so many areas, it is incredible. When I think of all the functions that are available in Blender for such a small program, I am amazed. Blender is clearly efficiently coded.
My personal point of view is that I prefer to have all the functions available. I think Ton is on the right path.
If I wanted a Maya or 3DS UI, I would use those programs. I prefer Blender.
"My personal point of view is that I prefer to have all the functions available. I think Ton is on the right path."
I would suggest you both download the free trial of Autocad 2013. It has over a thousand functions you can access and perhaps ten thousand options for experts to use. But it still has a simple UI that even beginners can learn in a matter of minutes, not weeks. Why limit your experience to other artistic 3d modeling programs when there are many great examples of user friendly but professional grade UI's in the CAD/CAM world.
I used to be a registered user of 3DS and Autocad. I left them and there expensive yearly upgrades behind a long time ago. as for the Autodesk EULA, lets not even go there.
You enjoy your Autocad. I prefer Blender and I am staying with Blender. It really is that simple.
The article must be partially a result of me bluntly shooting Ton in the irc by saying "Blender has a figure but not a face." It is harsh, but it has to be said.
Because every time I'm in incognito mode in the irc, I see some devs saying stuff any creditable devs wouldn't say. I know some are jokes, few are brainstorming and opinions, most of the time problem solving. But there is a culture in the irc that shouldn't be there (some famous studios have it too), devs might not see them, but outsiders can see clearly.
The problem is some devs can "hide" behind irc and talk about stuff that any user won't agree with. Then slowly sneak into making it happen (either good or bad). Just bad culture. I'm grateful for Blender but bad culture is bad culture, no matter how great the software is.
Adrew Price UI isn't bad - some says it's really good. But have you used it? I don't think so. It may looks nice, but how can you now it's better?
if someone wants to leran blender, he/she wil learn blender. If it's too difficult for you to learn blender, you propably won't learn it with diffrent UI either. And there are lots of new users too - people who have motivation to learn, who wants to learn.
Of course few changes in user interface woul be really nice - like history, but making blender interface less blender and more "adobe" and "microsoft"? Are you sure this is good idea?
In my opinion changing interface will be really bad for those more advanced users.
Nicolas, even people what wish to learn 3d modeling need to start somewhere. Most aren't set on Blender if they can get free trials to other programs. But once a user learns the basics on a program, they generally stick with it until they find the program doesn't satisfy their needs. But when they come back, they probably won't go to ones they remember were too difficult the first time.
I am one of the few I personally know to have started on blender and dropped it as soon as I got a free trial to 3ds Max. But what actually got me back wasn't even that program (interface was better than blender, but still not friendly), rather PLM's Solid Edge (free licenses while in college were nice, and layout is almost identical to NX) and Autocad. After that I was much more open to blender, but this was still the 2.49 days, and I dropped it in favor of just doing stuff the old fashion way in Photoshop and a sketchpad. You know why I preferred the CAD programs to blender? Goal oriented layouts. No more of this "memorize the shortcut table for everything since we don't care about you" attitude, every basic tool I would always use was right there on my toolbar, every setting easy to see and context sensitive. I recently got a license of the newest Solid Edge to do my actual modeling in and retired blender to rendering duties.
And guess what? ALL of the professional CAD programs from Autodesk, SolidWorks, and PLM changed their interfaces to one with points Andrew Price suggested. Guess how many professionals (remember, some of these guys make "professional" Blender artists look like the amateurs in terms of experience) went to the boards and complained about the ribbon system being an affront to experienced users... Practically zero, because they just said "well, not like I care since I just use shortcuts for everything anyway".
In fact, the point that Andrew Price was making was that true advanced users, those who have actually mastered the program, will have absolutely no problems with a UI change because they don't need the same type of feedback. The rest of the people, except those that claim they have mastered the system when all they did is memorize the broken paths they needed, will benefit from well thought out UI language and optimization.
I just don't understand how in the world anyone in the Blender community keeps complaining about how the 3D industry doesn't adopt Blender more and developers do more to appeal to professional use of Blender, but aren't even willing to meet the industry and professionals halfway on the issue with a focus on efficiency.
One could take some time out to learn Blender, but you're not just learning Blender's UI--you have to learn its entire philosophy. You can't even sufficiently incorporate Blender into an existing pipeline all two well.
I do game development and I can't even export animated .FBX models--a staple for game development now. I also can't just drag-and-drop .blend setups into Unity without encountering a problem almost every time. Meanwhile, Maya, 3ds Max and even modo make it all a breeze to do.
Blender's main problem is that it tries to master so much, it never truly masters any one thing. It tries to be the every-answer of the FOSS universe, and while that's a nice sentiment, it causes Blender's development to result with incoherent developments. And then its development is not unified--it's collective, and thus, its UI grows rather scattered and incoherent as a result.
Just why IS "Seed" next to "Clamp" in the Cycles' Sampling subcontext, as if they're associated? Why do I have to flip between two modes (Object and Texture Mode), two sidebars (Toolbar and Properties sidebar), two subcontext menu (Materials and Textures), and two windows (3D Viewport and UV/Image Editor) just to simply start texturing a model? Why are the Grease Pencil options split between the two sidebars?
Blender runs counter to practically every industry-standard software on the market, and heck, it runs inconsistently with its own self in places. Even ZBrush, as crazy and unintuitive as its UI may seem, has a method to its madness by offering specialized functionalities unavailable anywhere else as a result, and so people find it worth the trouble learning. But Blender has no such advantage, except that is free. But merely being "free" isn't enough.
Sure, some (many) artists out there are using Blender professionally, but frankly, Blender's still a hobbyist majority, and most of the Blender professions started as Blender hobbyists. If a non-Blender user freelancer wants a tool they can trust, many will just look at modo these days, which is not only more affordable than Maya and 3ds Max, but caters more to efficiency than merely trying to please its predominately-hobbyist user base. Even Maya and Max have become affordable now, thanks to Maya LT and Autodesk's available rental plans.
Some people can bring up the *price tag* of Blender as a reason to use it, but frankly, a working professional is willing to spend money to save time not fussing with their software and that plays well with other software.
I know this because while I use Blender myself (and I truly do love Blender), I've been looking at buying modo, just to help make my game development workflow as streamlined and no-fuss as possible. I would love to just rely on Blender 100%, but I just can't. I was hoping this UI proposal would spark some improvements with Blender (both in UI and workflow efficiency), but I guess not.
Corrections:
"...an existing pipeline all too well."
"...than merely trying to please a predominately-hobbyist user base."
"I would love to just rely on Blender 100%, but I just can’t. I was hoping this UI proposal would spark some improvements with Blender (both in UI and workflow efficiency), but I guess not."
I too was hoping that those in charge would finally consider the meaning of UI as being as important as the actual code behind it. I've actually turned outside of artistic 3d modeling and into the realm of CAD to supplement my needs. There is only so much time I can devote to learning the program, even if I have plenty of time to master it.
As a professional you should know that "FBX is a proprietary file format (.fbx) developed by Kaydara and now owned by Autodesk" (Wiki) - Maya, Max are Autodesk products so naturally there's no problems with export while others havo no rights or need licenses to do that. You may try to reverse engineer FBX if you like, but Blender format is Open and everyone - You are free to ask Unity Developers do use .blend file as import source with all animations and ton's of other stuff. We use .blend format in our interactive WebGL applications with no problems.
Oh... i'm so tired of these all "profesionals" and "industry standards". Art is created by artist and software is only a tool to do that.
Modo manages the .fbx importing just fine and it is a Luxology product, NOT an autodesk product. Lockett specifically included Modo in his list. I'm not sure why you chose to ignore that.
People reason only when they want to reason, and hear only what they want to hear.
1) You've refuted absolutely nothing I have stated. Absolutely nothing.
2) Scroll down and read the whole article, sir:
"Autodesk provides a C++ FBX SDK that can read, write, and convert to/from FBX files..."
Just because FBX is a proprietary file format, it doesn't at all mean you can't develop support for this file format elsewhere. The proprietary nature of FBX hasn't at all stopped every other software in the industry from adopting it with full support.
Even Blender now has better FBX, which, while is something, it's not full support. As I pointed out (and you ignored), I can't import animated FBX files (actually, I said "export" by mistake, but I meant "import"). I'm aware that the FBX format in Blender is a work-in-progress, but that's another issue with Blender--I'm too often waiting for basic features to arrive.
We'll get new camera-tracking features, 3D printing features and new simulation models added (which are nice and all), but then it takes years just to see a simple things like full FBX support. This happens because development is rather scattered instead of focused together to ensure the basic features are done with design cohesiveness.
3) You're pulling a fallacy in claiming that "As a professional you should know that 'FBX is a proprietary file format (.fbx) developed by Kaydara and now owned by Autodesk.'"
First of all, I was already fully-aware of the origin of the FBX format. Second of all, no, professionals don't have to know that FBX is a proprietary file format developed by Kaydara and now owned by Autodesk--they only really need to know how to use it.
But yeah, pretty much anyone working in the industry in some way are familiar with the fact that Autodesk owns--or, at least, emphasizes more than anyone--the FBX format.
4) I don't have to ask Unity developers to use .blend file as import source with all animations and other aspects of the file.
My issue isn't that Unity doesn't take it at all--it's that Unity sometimes runs into these little hiccups with doing so. It doesn't happen all the time, but it's happened to me enough times where I'd just rather use a FBX file and be done with it...which I can't completely do in Blender.
5) You might be tired of "these all “profesionals” and “industry standards,” but I guess you and I have different issues and needs, now don't we? You might not find such issues as limited FBX support as that big as deal for your line of work, but it's completely game-changing for my line of work.
And you claim that "Art is created by artist and software is only a tool to do that," but you clearly don't care about the side of the argument that, yes, sometimes the tool matters. Very much so.
You want proof? I'll bet you probably use Photoshop instead of the GIMP. Why would that be? If I am right in my assumption, it's because you know that the GIMP isn't capable of even half of the time-saving, convenient and customizable features that Photoshop has.
* * * *
I don't mind a respectful disagreement with my point of view--I love a good exchange. But I do get sick of people often coming at me with nothing but fallacious arguments, most particularly on this issue of improving Blender. If you don't have a relevant argument to actually refute something I've said, please don't try with red herrings.
Autodesk does provide an SDK for FBX. Unfortunately, the SDK is licensed in a way that's incompatible with the GPL, so an importer/exporter that bundles the SDK cannot be shipped with Blender. It may be possible to get around this by wrapping the SDK in a Python wrapper and having a Python script make calls to that wrapper, but even in that case, the SDK could not be shipped with Blender; the user would need to download that on his/her own.
Modo (and other commercial software that supports FBX) has licensing that does not conflict with the FBX SDK's licensing. Blender does not have that luxury, so the importer/exporter must be meticulously written from scratch.
Yeah, you're 100% right there. I'm aware there's a difference in licensing there going on, but still, we've gotten this far with implementing FBX in Blender and apparently, it's still the intention of developers to further include support for armatures and animation. It'll take some fancy doing, and yes, it's not always easy, but it's still on the agenda, meaning that something is still possible..
Though my greater point overall was that the fancy doing should've been addressed earlier. Before working on camera-tracking and 3D-printing tools first (which are nice and all), perhaps they would've been better to focus on some of the more incomplete basic features first.
It's too often the little features, I find, that hold me back from relying on Blender 100% as I would like.
Anyways, thanks for addressing this more in detail. It'll clear up some confusion for folks in this dialogue I've started. You're right--it is a doozy of licensing. It's just a doozy we probably should've seen worked on first.
Blender artists don’t care that other artists don’t use Blender; they do care that they get looked down upon when they reveal their work was done with Blender.
Blender doesn’t get taken seriously because it doesn’t have a multi-million-dollar PR budget behind it to extol its virtues to all and sundry, like the proprietary packages do. That’s all. Everything else is an excuse.
" they do care that they get looked down upon when they reveal their work was done with Blender."
I've seen plenty of Blender artists looking down at Maya users too. In fact, there's plenty of cases when Blender users look down on other, newer users too! But it's usually not a "oh, well that looks bad then response" when people reveal they used blender, rather more of a "must have taken a long time to get proficient enough at blender to do that".
"Blender artists don’t care that other artists don’t use Blender;"
The point is that they should care, care enough to learn why others decide not to use blender. If you find that most people avoid blender because of the UI, you know what you have to work on. If they avoid it for plugins, find out which plugins and how to give them the features they need. The backwards, and flat out elitist, thinking that Blender should not care about those who use something else is going to cause the end of blender. Most GPL programs end up running into that wall at some point in time, and that's when they get railroaded by something new even if it reuses 90% of the code (see the slide of megui and rise of handbrake).
Everything else is an excuse? Well, if that isn't spoken like a true biased fan. I think you're purposely ignoring some standing issues here.
I think I would better say that Blender doesn’t get taken seriously because it's a waiting game with Blender, for the most part.
Not that at all to say that commercial software are always on the ball with development, but nobody using Maya, Max, Cinema 4D or modo had to wait nearly a whole decade for n-gons or several years just for simple FBX import, like the Blender community has.
Blender doesn't get taken seriously when it goes on to develop 3D-printing tools, camera-tracking tools and simulation features galore (which are dandy and all), but still doesn't focus on refining the essentials first. It tries to do so many things that it never truly masters any one of those things.
We have 3D texturing that's just now being addressed with features others have long had (not to mention, setting up texture painting isn't streamlined yet). We have a great advantage with Dynotopo in sculpting, but we're still missing simple things like sculpting layers. We've needed split vertex normals support for OBJ and FBX long before just now. We only just recently added the true Knife tool's functionalities. We need more options with the Undo/Redo feature. And Cycles, while a great renderer, is still waiting on features that are already completed and far excelled in the likes of V-Ray.
Meanwhile, we do have an advanced game engine, camera-tracking tools, 3D-printing tools, an ocean simulator, Dynamic Paint, and non-photorealistic renderer, which while are all beautiful efforts and useful features, they're not first-priority developments. We needed a true layer system first before any of these. We needed to work on FBX long time ago. We needed the option for mode-independent chronological Undo/Redo.
And to make things worse, we've got scattered documentation that often doesn't explain the features well enough. You have to hope someone made a decent tutorial on the feature you want to learn, half of the time. Good luck learning Blender without taking about a couple of weeks to go through searching through ad-hoc learning material.
I'm not at all knocking the hard and continuous work of these fine developers of Blender. On the contrary, I say that the way Blender's been developed is undermining the work of these guys. Blender is so inconsistent and ad-hoc with much of its development together, that it's hard to let these developer's developments truly shine with their best light.
Blender's simply not trying to address the industry's needs, but rather, the community's interest. There's nothing wrong with this. It's just becomes a problem when people like you start calling others who don't consider Blender industry-ready as simply "making excuses."
There are many things I could say about this issue, but the main reason why many don't take Blender seriously is that they're too often waiting on Blender to catch up to what's already standard and ready elsewhere. Nobody has time to wait on Blender. Except Blender users who have no choice in the matter.
It's not all an excuse. There are very practical reasons as to why Blender isn't widely-accepted. It's got little to do with PR. It's got more to do with practicality. Professionals will pay for a tool that's going to do the job they need without fuss. Blender needs to stop resting on its "free and open-source" laurels, because nobody cares.
Correction:
"...If Blender users want Blender to be taken seriously by industry professionals (many freelancers included), Blender needs to stop resting on its “free and open-source” laurels, because nobody cares."
You seem to think the agenda should be set by what proprietary packages consider important, not by what the Blender community considers important.
This reinforces my point, that your thinking has been swayed by the multi-million-dollar PR budgets behind the developers of those proprietary packages.
The entire topic is moot and has been overtaken by events.
The presentations at the conference today provide clear proof that nothing drastic will be done with the UI and that all involved have been conversing, listening, learning and re-evaluating their preconceptions.
To the vocal segment here who wish to continue sniping around the edges of the topic, pointing fingers and making accusations of ill intent, feel free to do so. Everyone else, the BCon videos are online and it's a good showing so far. Lots to view, make of it what you will, and everyone have a great weekend.
I agree with Brian.
Sooo many trolls in such a small place, one can hang an axe midair :D
The old "It ain't broke so why fix it" response. What a surprise!
It's very hard for me to understand how someone capable of programming an application with so much functionality can't see that separating the GUI from the rest of the application would actually help new functionality implementation, bug fixing, scripting integration and so on.
Maybe it's time to fork Blender.
Hey, that forking Blender idea sounds like a great solution.
Though, it'd might prove an even greater challenge than the original Blender development. It's already hard having development of Blender with the few full(er)-time developers we have. I'd imagine it ll might be even harder finding some dedicated developers to develop a fork.
But hey, if we somehow could, a Blender fork's actually not a bad suggestion. Might just be the best solution of all. That way, those who want the original Blender can keep it, and those who want an attempt at what they view as a more efficient Blender could likewise have it.
Maybe it's time some developers in favor of a newer Blender design organize a special Blender team for this effort, and possibly even look into organizing a crowdfunding towards kick-starting the effort.
I love when I see 3D artists with "years of experience" that don't know make a basic model, texture or something (In blender or any other soft) talking about the industry, the problems of blender,...
It's like movie critics that never made a movie. ¬¬U
Yup! Exactly.
"t’s very hard for me to understand how someone capable of programming an application with so much functionality can’t see that separating the GUI from the rest of the application would actually help new functionality implementation, bug fixing, scripting integration and so on."
Because it doesn't. Blender is a very modular system. Shuffeling around functionality so it better suits the workflow makes it harder to maintain, bugfix etc.
You're right about the forking though. If someone thinks it would be feasible to get a better UI around the current code it would be best for them to get a team together and try their luck instead of trying to force the blender developers to move in a direction, that doesn't fit their vision of blender.
There has been an android branch and I would like to see a branch that tries to integrate touch functionality better. If someone has the resources - why not?
"Because it doesn’t. Blender is a very modular system. Shuffeling around functionality so it better suits the workflow makes it harder to maintain, bugfix etc."
There are two types of "functionality", one being the tools you use and the code behind those tools. The other functionality is the ability to use the tools you need without needing a 1600 page manual listing what tools are where.
The function of tools in blender are mostly hidden out of sight, with only the tips sticking out in the form of python bindings that can be used to control how the tool presents itself (and the result of the tool of course, but aside from progress bar discussions that is irrelevant) . A UI overhaul can be as superficial or deep as you want, and can start with some thing as easy as rearranging current tabs (manually create new properties menu items and delete the old ones). Throw in some tool icons that just map to existing keys (move, rotate, generate specific shapes, etc) and you can have a working example without needing to rewrite how the panels are displayed. That can improve functionality without changing function at all. However, that requires developers that know what needs to be changed to be feasible within any reasonable amount of time, and it is doubtful any existing developers would want to devote even a few days worth of time to do it.
So far this discussion hasn't been flaming for too long, and perhaps someone or group will take it upon themselves to give the UI overhaul serious thought (soon).
all my friends that use 3dsmax or Maya when i try that learn blender they say "No" and the reason is the UI, in the beginnings of Blender the UI has very unfriendly, and many time and money after this not has changed very much, I remember when the developers say "Undo function in Blender, Why?".
Sorry, but this is tale stories or your friends are totally dumb. I am teaching Blender schoolkids and students at University for 5 years now and they master Blender basics very quickly - no complain about UI at all.
you are right, your schoolkids dont use 3dsmax or Maya.
I can tell you from first hand experience that kids can pick up a foreign language much easier than an adult who has never been exposed that that language.
I think that is apples and oranges. Blender should be as intuitive as possible given the restriction that you can't make radical changes to existing code.
The Blender UI cannot be radically transformed, but that doesn't mean that it can't be improved.
Now they say: "alert the user when the file was not saved? Why?".
Correction: they say "Alert the user when the file was not saved? There's an option for that in User Preferences. Enable if you would like."
Yes, some tutorial gave me this tip and I immediately turn it on. It works if someone try to quit Blender. But don't work (yet) when you try to open another file.
But, this should come from factory instead of be an option that you must activate.
Anyway, thank you!!
Indeed. You'd think the most obvious features such as the quit prompt would just be factory default from the start. Why make anyone new (and maybe even not so new) look for it?
If someone didn't come to using Blender when the feature was first announced, one could easily go a long time without even awareness of the software.
And some folks claim that people should just "learn" Blender, but as if ANYONE goes and explore every nook and cranny of their robust software.
But what are you going to do? One will often find themselves surrounded by a "fanboy" mentality around here. Let's just keep to our esoteric ways and blind optimism--never mind if an honest dose of reality could just be the best thing for the community.
Correction:
"...even awareness of the software's feature.".
Some people on Blender team and Blender fans do a confusion between developers and users. Developer must to know a lot about codes, languages. Some of them even dream in python language.
Users are different kind of people. For example, I make book covers and book ilustrations, mainly children's books. For me, the ideal UI should be unnoticeable, like the brush is unnoticeable for a painter: the important is the flick, not the brush.
This confusion about the two "tribes" is behind the current discussion.
You would be surprised that the most skilled developers are not interested in the language itself at all. There is a common saying in the programmers world: "Its not about how many languages you know. If you have learned two or three of them really good, you will master nearly all of the others without any problems in small time."
I think developers think more about functionality and data-models. If they could, they would get a piece of paper that is the size of Amsterdam and draw "the perfect blender" data/ui/node-graphs on it. Well, it would take a few hundred years, or more :-)
You are absolutely right about the problem: How to get Developers and Users together? I think the devs are addressing that quite well i the last few moths by trying to get skilled artists who use a specific part of blender very often to work together with the devs of this part of blender. Sounds like a very good way to me...
Yes, your last paragraph sounds very promising. I think this is the right way.
If you're painting without an awareness of the brush you're using, I'd say you're not a particularly serious painter.
That aside, you might be surprised to know how many of Blender's developers are also users, artists, and designers (or started as such). The developer/user dichotomy exists... though it's arguably quite a bit less in the Blender community.
We can all agree he meant that there should be no jarring aspects to the UI, just like your brush doesn't suddenly turn to stone. Jumping between the menus and buttons and keyboard shortcuts needs to be smooth, rather than the disjointed mess we have in some parts of blender (some by necessity, like motion tracking, others by careless UI element placement like with the rigid body system)
Basroil, I like the way you think. Hope you'll stick around.
After see your answers and your portfolio. Do you really think that the problem in your relationship with the 3D is the UI of blender? Because I think that your problem is that you don't know nothing about 3D, art and software. Blender's UI is not guilty. Your problem is that you don't want to learn nothing, and you will have this problem in blender, maya, max, modo, houdini and all the 3D suites.
You won't do better works with proposal, you won't do faster, for the work that you make you can use the 3D options of photoshop extended. Do you really think the basic things that you make is the only target of blender? must all people suffer a kid's GUI because you make kid's book covers? you are selfish.
I don't want tell that the people with few skills in art or 3D must shut up... but at least don't speak in all the fucking post like yisus.
Ah,and this message also goes to brian Lockett.
so when you have no points the only one is to attack others works?.. good point indeed...(Sarcastic mode on)
No, the point is other... It's really insulting (specially for devs) that two guys that don't know to use the basic tools of blender speak like they have ten years of experience, criticizing developers, plannings, design... If your works are worst than a average amateur kid at least be modest and don't speak about the industry, art, code,... in every post.
Ton writes: "The misconception here is that the “UI” is some kind of separated entity of a program, which can just be peeled off and replaced with another interface. However, a really good UI (and certainly Blender’s) is not just an abstract independent layer, it’s the reflection (and sometimes even the core) of the program’s design."
I have been involved with software development and testing with the US Department of Defense high technology programs since 1989. Ton is right. You can't just peel off one UI and slap on another. The UI is integral to the entire code and not subject to significant modification without great cost.
The UI should be part of the initial concept design because changing it down the road becomes increasingly more difficult--unless you rewrite the code with a new well thought out UI design.
But Andrew Price is also right in that the Blender UI can stand improvement. But it's not realistic to expect Blender to adapt such a radical design change.
I guess I'm in the middle. While ribbon bars may be a bridge too far, I do not think it unreasonable for Blender to make minor improvements in the UI with each new release.
The purpose or function of many of the buttons, sliders, data fields, etc. could be improved. Perhaps a link to the Wiki for that option and the documentation on the Wiki could be improved as well. That by itself would be a big help without changing the current design of the UI.
There are areas where a user needs to create something that has standard steps in the process. Currently, a user must discover those steps on his own. In such cases, perhaps Blender could prompt the user to the next required step in the process. That does not require a UI design change.
Blender could make some improvement in the location of the options so that common related options are grouped more logically.
I'm not trying to tell the Blender Foundation how to suck eggs. It's their software and design and programming culture and while outside suggestions are welcomed, the Blender Foundation is not likely to make radical changes any more than Maya, ZBrush, or any other application that feels they have a good UI design.
I'm only suggesting that it is possible to make Blender easier to use without adding a ribbon bar. As Blender Foundation continues to improve the functionality, I hope the allocate some level of effort into making UI improvements where they can.
Hey Ton,
Thanks Ton
Do not change UI. I am happy with Blender UI.
I love it.
Do not follow "Andrew Price". I don't meant to rude.
Blender UI is just perfect .
I think You're genius.
Sorry for bad English.
Because I am deaf.
Keep minds open, that's all we need.
As a amateur Blender user who has spent my evenings learning this software through the many dedicated users uploading tons of tutorials into Youtube and Vimeo, i must say, this is really depressive reading. 'And they do wonder why wars start....'
Is Blender really that bad? Do the development team deserve this by its own users? Are those who throw in the biggest rocks and stir hardest in the water in such a pain?
Andrew is a dedicated creator, a good one too, and this was his way of looking at what he needs, point taken, case closed, lets see what happens next.
I met Maya and Autodesk users that wont touch Blender...why? Because they are accustomed with what they got, they can do what they want and is satisfied in whatever they do. And some of them do it really well too. There are amazing artist out there. And i seen amazing artwork created in Blender also, dazzling!
I used to have a Canon camera i was very proud of, with the best of lenses and equipment. And i took fairly good images i thought myself. Then i met a guy that had a cheap standard Canon with the stupid plastic lens, which i never thought much of, its the basic that is delivered with every Canon camera. But his images was out of this world beautiful. Some can make gold with the simplest means and others can't make anything with the even the best equipment, its the mind behind, not always the tool.
In my daily work i use a 3D software for pipe, hull and electro engineering, 5000 pounds pr. lisence, love it or leave it, hardly any import and no export option. Terrible 3D component modelling capabilities, the worst, but the alternatives is so bad you just have to stick to it. Import a model created by any Autodesk software and your stuck in limboland. Just too many verticies to handle.
Then i got another 3D software for private use, for animation purposes, wonderful for the pros, i'm sure. If you want to know how it works, pay for knowledge, you get nothing for free, no free support, no Youtube tutorials, hardly any user groups. And the creators of the software don't answer any request, they run in their own pace, still people use it. Fantastic tool, super hard to learn.
Most of the developers in Blender is doing this because they love it, unpaid, got a family that complains if they spend too much time behind their screen, need to relax, watch tv, go out. Ton said they are just like us, humans like us, i dont think they are..... they are artists, creators, thinkers, we...most of us are the complainers, the one who get grey hair and scream if we dont get what we want. Like my kids.
Things will change, things are in development, calm down, keep making fantastic art and dreams, you dont need to have it NOW, not everything. And if you do, you know there are other software who will meet your needs.
I love Blender because i just do, i have tried many software but this one i just like. I dont need a pirate copy of any other software. I can't afford the super expensive Autodesk software with all its thousand of plugins which adds even more to the bill to perform yet even more which would have been cool.
Throw as much petrol and fire bombs into the discussion as you like but concider if it gains anything. The dev team got a fair idea of what is needed now and what some want, they dont ignore you......
Love your story about the cheap plastic lens, and how it doesn't matter! :)
This debate is good and healthy. But some guys insist on disqualify Andrew or whom agree with him. This kind of reaction sounds like: "bad news? Lets kill the messenger".
Thankyou so much Ton, you've summed up the position of the Blender Foundation perfectly, and it's the right one, you have the balance just right. Ignore the naysayers!
I started working at a company about 6 weeks ago, I've been using Blender for my job to create 3D models of furniture that they manufacture, Blender effectively is paying my bills at this point. So, I'm going to give back and donate to Blender right now, send them a nice chunky donation as thankyou for all they have given me in this great application. If you've ever earnt a dollar from Blender, you should consider giving back too!
Just made a 60AUD donation, and bought a training DVD while I was there too. It's only fair considering right now I'd be flat broke without Blender.
Blender's current Ui schematic is a blessing when it’s compared to version 2.49 which I hated with a passion (2.49 has the worst Ui scheme ever regardless how stable it is). Having said that I feel that the current interface could use some work by rearranging things just like Andrew Price suggested. Adding a feedback dialog box would be helpful too.
Honestly, I think the new proposal would be great for Blender but at the same time I understand how restructuring the core-interface from a programmers perspective would be a nightmare. There really is no way to achieve something of this magnitude in just one step. Making smaller changes over a period of a year would be more realistic.
The way I see it, the easier the software is to use the faster I will get the job done. Please support a new schematic.
This discussion reminds me of that old Aesop's fable about the boy the man and the donkey.
After taking the advice of every stranger on the road to market, the donkey is eventually drowned in a river, and so it would be with Blender if Ton and the design team took the same road.
Blender is NOT trying to be Maya, or any other commercial proprietory program - it is trying to be Blender.
In this world where the pressure to conform to the herd is so strong, it is easy to see why some people are more comfortable with commercially produced interfaces, even though they are simply just another form of symbolic metaphor.
Blender does not ask to be accepted by commercial users, or by anyone in particular - it is the solution to the problem for those who need it and are willing to work with it.
It is also not a fossil, but a living creature which is on a journey, changing and evolving as it travels.
Its purpose is not to make money for corporations, but to make solutions for real people, regardless of their financial status.
Those who are trying to force Blender into a corporate box, are simply not looking at Blender in the right light.
I understand that we all have to make a living, and modern society has been absolutely saturated with the "time is money" , and "you get what you pay for" memes, but it does not mean that is the right way.
Blender is like a free fountain on the road of life - you can drink or not, and it won't cost you a cent, but don't pollute it for other travellers.
Mr Price has made his point, and it is his perogative, but to call Blender "broken" is both inflammatory and untrue.
It belittles the hundreds of volunteers who have given their time and talent to producing this wonderful free gem for all of us to use as we see fit, and to the limits of our talents.
Ton is very perceptive and his observations are both insightful and measured.
We are very lucky to have him at the helm of Blender, as it weathers the storms unleashed by its own users.
I am a user of Blender since it's beginning when it was edited by NaN. I don't use it professionnaly (my main job is programming) but for fun and I make pretty beautiful pictures in my opinion. And for the years I use it, the user interface hasn't changed a lot. For sure, the 2.5 version was something different but still has the same usability.
An user interface is like an identity. Blender has it's own identity, and I think that trying to imitate other software is not a good idea : Blender will loose it's identity. And changing radically an interface people are used to work with can of course introduce a lot of complains (simply look at Office when they started with ribbon).
The proposal of new interface is not evil but simply let users to have the choice.