interface

[GDP_Sabrina]

What are you hundreds of designers doing?

I guess, they all are creating brochures, movies, samples, models and all that marketing stuff.

Re-educating hundreds of designers is going to be very very expensive.

Especially if the interface is a little bit "unnatural" - please excuse this comparison, but a colleague said:
"Sabrina, you were so enthusiastic about Blender, so I downloaded it. After clicking around for a while I must say: This application looks like a tool for programmers, who like to fiddle around. The usage of Blender versus Cinhema4D is like a text console versus a windows based operating system. I am sorry, but I guess you won't find many supports at the marketing department. Big companies like us usually have demands and regularies for applications. And one must is, that applications should follow some common usage guidelines like menues, copy&paste between applications and stuff like that. The more an application behaves like professional standards, the more supporters you can find for a switch."

Regards
Sabrina

[Antlab]

Hi.
I have just registered to the forum, but since several years I am interested in Blender (I also bought a couple of manuals, and made tens of gingerbread guys:-)).

After clicking around for a while I must say: This application looks like a tool for programmers, who like to fiddle around. The usage of Blender versus Cinhema4D is like a text console versus a windows based operating system.

I think that you just touched the critical point. The discussions about Blender interface are very similar to those concerning GUI vs command line. Also the arguments from both parts are analogous. For that reason I think that it is difficult to reach a sort of "agreement".
I am quite convinced that, considering the present state and the development speed, Blender with a better interface could really gain an ENORMOUS number of new users (both amateur and professional).
From the other hand, I can understand that the "hard core" users, after many years of study and work, harshly reject a too radical innovation of the procedures.
Sabrina, sincerely I agree with many of your points, and I'm really curious about the possibilities of new UI customization with the 2.50 release. One good example could be what Pixologic made with ZBrush: also in that case the interface is not immediately intuitive, but thanks to the possibility of easily modifying it, in the new ZB3 there is a button to switch from the complete scheme to a simplified one . In that way new users can more easily learn the basic concepts and commands, before being exposed to the full complexity of the program.

Ciao

Antonino

[GDP_Sabrina]

IMHO, blender is comparable to the Linux dilema:

But Linux did change in the last years. Though the visual interface was there for years, in the last versions they started to offer the configuration to end users (instead of weird config editing files manually).

Meanwhile Linux has developed further:
On the one hand Linux offers a very cute text console GUI with fast command and scripting features. And on the other hand Linux offers a visual seductive interface (KDE, Gnome etc.) for the occasional users, which allows them to do the same (though they simply need to click around; but they are able to config the system at least).

Since Linux offers most of the features over the visual GUI, Linux gets more and more a real competitor to Windows. Blender has so much to offer. IMHO the classic blender interface and a fresh look should only be a different perspective to the same thing.

Regards
Sabrina

[kAinStein]

kAinStein: Do you develop for blender?

For short: No. I tried but I totally suck in plain C. Because the greatest part of Blender is written in C and I don't have the time to get behind what can be changed where I get frustrated very fast.

[kAinStein]

From the other hand, I can understand that the "hard core" users, after many years of study and work, harshly reject a too radical innovation of the procedures.

Sorry, but that's simply not true. You should get one thing: Using Blender does not bring workflow - it makes work flow. So I'm not against a change until it really integrates well into the existing.

First: Blender is not tool oriented - it is action oriented. That's quite a difference! Tool oriented -> traffic light, action oriented -> traffic circle. So many demanded changes can't (or better shouldn't) be made the way they habe been requested. Take for example Azrael's suggestions in the other thread: Most of the icons don't make sense in an action oriented environment - they just take place on the screen that can be used somehow else. It also does not integrate well (If you would make it the way Python space handlers can bring their own UI with them with icons and stuff then it would integrate much better). But if you take a look at Matt's trial with the tool panels it does not only give you the option to have the most needed stuff at hand (I would suggest that you could choose between "text", "icons + text" and "icons only"), it can be extended to record macros, to make script starters and probably a lot more. In addition you could give along the arrangement via a Python script without having to load a whole UI default file. Having a new window type full with icons so it might look integrated is quite dumb and does show that the problem is not understood!

In addition there are a lot of requests that take wrong premisses: For example the intuition point. Many things are not intuitive though people claim it to be. The big difference: They are used to it - but it's not intuitive by definition.

Then some demands are not ergonomic: Though Maya has the QWER short cuts they are not very ergonomic from my point of view because they are too close together (And I've got small hands... Makes it all too stiff...). Some people say that it is good that they are so close together because it speeds up the workflow but if you look at your daily work you will notice that you move your hand often from the keyboard - so there is no real advantage at all. I think the Blender hotkeys are better arranged and in addition they are easier to memorize being close to intuitivity.

There are more points that criticize such demands. Just read older threads. I don't want to write it all again. And be sure to read what people say that criticize such. As I mentioned: There are quite a lot of inconsistencies that weren't there before 2002! So a UI change has to be done. But in a rational way. And "we must have XYZ because ABC does have it that way" is simply not a good reason. This has also been said every time!

Also: It doesn't matter which package you intend to migrate to, you will always need to take your time and learn it! And don't think learning Maya is easier than learning Blender - some things might look weird at the first glimpse but the learning process is exactly the same!

And one last thing: The point that the different usage doesn't fit with other programs - well, I've never used Blender hotkeys with some other program or vice versa. When I use Blender then I use Blender - and if I'm using something else then I'm not using Blender. I know the shortcuts of several programs and don't mix them. And I guess you don't make swimming movements while riding bike - or do you?

EDIT: A big boost would be if the Blender widgets would behave more like other UI widgets. Something I've requested a long, long time ago...

[Antlab]

Hi Kainstein, thanks for the long and detailed answer. I will touch only a few key points.

There are more points that criticize such demands. Just read older threads. I don't want to write it all again. And be sure to read what people say that criticize such. As I mentioned: There are quite a lot of inconsistencies that weren't there before 2002! So a UI change has to be done. But in a rational way. And "we must have XYZ because ABC does have it that way" is simply not a good reason. This has also been said every time!

Yes. I always read a lot of previous posts before entering in a new forum. So I discovered what I wrote to Sabrina, that the discussions on Blender interface usually finish in a dead end because, IMO, the needs of "reformist" and "conservative" are quite different. Being very curious, after reading here some references to the work of Jef Raskin on Human Interfaces, I did a little research about him and his book. I also visited the site that presents his works, http://jef.raskincenter.org/home/index.html.
One reviewer on Amazon criticizes the book, saying that Raskin seems ignoring the relationship bewteen aestethics and ergonomics, for example concerning the use of colors and fonts.
I think that this could be said also for the Blender UI.
I would like to underline that I understand the point of view expressed here by some expert users and developers, regarding their ability with the program and the speed of their workflow. Anyway I remain convinced that a different interface, probably less based on Raskin ideas, and more pleasant to the eye, could ease the approach by many users, including myself.

Ciao

Antonino

[lukep]

One reviewer on Amazon criticizes the book, saying that Raskin seems ignoring the relationship bewteen aestethics and ergonomics, for example concerning the use of colors and fonts.
I think that this could be said also for the Blender UI.

I would like to underline that I understand the point of view expressed here by some expert users and developers, regarding their ability with the program and the speed of their workflow. Anyway I remain convinced that a different interface, probably less based on Raskin ideas, and more pleasant to the eye, could ease the approach by many users, including me.

Ciao

Antonino

Hmm, that is a very strange critic of Raskin work. aestethics and ergonomics are fundamentaly tied in his work. they were even the base of the macintosh design ideas. I would even say that we need to stick more to Raskin ideas than less.

Now, the issue seems to be that most softwares are now tool oriented interfaces, while blender is action oriented (partly, you cannot be 100% of one kind). the difference is fundamental and cannot be fuzzed upon.
But, that is what makes blender workflow so fine, the UI is the same whether you use the shorcuts or buttons.

In blender, the general rule is that you select something then act upon, not select a tool then select something to use it. The only exceptions are the modal tools like sculpt, but the first rule is still true at entry point.

Tool oriented interfaces are more pretty because of the nice icons, but they are a lot less efficient in productivity. Main advantage of action oriented interfaces is their flexiblity. Switching from one model to the other would be dropping all blender strengths so is not even thinkable.

Now, as i said before, blender lacks a lot of visual cues and feedbacks, which is due to old event system design, and that is what should corrected. There is ample work to do.

If we do it well, we should have both systems benefits, user friendly interface and efficiency. but dont expect nor request copycat interface of C4D/Maya/3DS.

And by the way, which one ?

[Antlab]

Hi Lukep.

Hmm, that is a very strange critic of Raskin work. aestethics and ergonomics are fundamentaly tied in his work. they were even the base of the macintosh design ideas. I would even say that we need to stick more to Raskin ideas than less.

Actually, after reading a bit more on Raskin's work and projects, I better understand the reviewer's opinion. From the Wikipedia I read:
"The machine he envisioned was very different from the Macintosh that was eventually released and had much more in common with PDAs than modern GUI-based machines."
And, going to his site and looking at the demos for Archy, his last project, I was a bit shocked, because the concept of G(raphical)UI is completely absent.

Tool oriented interfaces are more pretty because of the nice icons, but they are a lot less efficient in productivity. Main advantage of action oriented interfaces is their flexiblity. Switching from one model to the other would be dropping all blender strengths so is not even thinkable.

Sincerely I don't see the strict relationship between graphical interface and tool/action oriented paradigms. I understand what you say about the working model of Blender, but this is not directly related to how the different commands and options are showed on the screen. IMO a simple and clean use of tabs, boxes, icons, expandable menus could maintain the same working model, but making the workflow easier for a lot of people (and less intimidating for new users, and more "standard" for professionals like Sabrina's colleagues).

Now, as i said before, blender lacks a lot of visual cues and feedbacks, which is due to old event system design, and that is what should corrected. There is ample work to do.
If we do it well, we should have both systems benefits, user friendly interface and efficiency. but dont expect nor request copycat interface of C4D/Maya/3DS.

Perfectly, we agree. And I don't expect a clone of some existing interface, also because I much appreciate the originality of Blender (ok, maybe we could take inspiration from the UI of Bryce ).

Ciao

Antonino

[LetterRip]

In Tons Siggraph report during the meeting - he mentioned that about a dozen people told him they were waiting on adopting Blender till the interface/customization refactor. It is definitely a priority for him - so please - stop the flamewars already

LetterRip

[kAinStein]

Actually, after reading a bit more on Raskin's work and projects, I better understand the reviewer's opinion. From the Wikipedia I read:
"The machine he envisioned was very different from the Macintosh that was eventually released and had much more in common with PDAs than modern GUI-based machines."
And, going to his site and looking at the demos for Archy, his last project, I was a bit shocked, because the concept of G(raphical)UI is completely absent.

Well, Raskin's work was an inspiration for the Blender UI and surely there are some issues with Raskin's vision. This is not a religious topic where "The Humane Interface" is something like a holy book. But there are a lot of things that simply are right. Archy is ok, but I wouldn't use it I think. But it demonstrates that content and editing can be arranged like someone needs it to be - and hides functionality that a user doesn't really need to be aware of without taking the option to use it. Additionally Archy was not only meant for the avarage user but also for disabled people! And you should see that aspect, too (Trust me, a blind person doesn't need a mouse cursor and lots of colorful icons!).

Sincerely I don't see the strict relationship between graphical interface and tool/action oriented paradigms.

The relationship is not strict, but it's there. If you transfer the virtual buttons for actions to real life buttons on your keyboard and restrict visual feedback to what is happening instead of showing that a button is selected then you don't need virtual buttons that take space from your working area. This means less clutter, better visibility on your work and more space for your workspace. As a human in the first place you are used to work with your hands - and it not only feels more natural to hit a button with your hand - it's also a lot of faster! And if you are practiced your body memorizes where to hit a key and you don't have to think about it anymore. That is much more natural.

IMO a simple and clean use of tabs, boxes, icons, expandable menus could maintain the same working model, but making the workflow easier for a lot of people (and less intimidating for new users, and more "standard" for professionals like Sabrina's colleagues).

I strongly disagree. Especially if you are using tabbed toolbars you lose a lot of time switching between tools. So you should restrict tabbing only to procedures that you actually don't need constantly. This is simply not a good way. Further switching between tools via graphical buttons isn't the best way either - especially for making work flow (I intentionally left there a space!). I think the analogy with traffics lights and traffic circles (first does have a strict division between two procedures - meaning you've got to first select a new tool, do something, change tool, etc. - also like stop and go - and the latter does have a natural division - all happens when it should happen - but then instantly).

The only point that I can agree with is that Blender could be made better accessible to people that have never used it before. The first step to do would be better documentation. Because documentation is the key to every single piece of software on earth - whichever the program might be! And it's a very cheap excuse for a user to blame the software instead of working on the own deficits! The second would be to make the widgets easier to identify their function - meaning making them look and behave more like widgets people are used to. The third thing to do would probably be to place commonly used functions in the user's view because some people are used to that. That's ok! I again would recommend Matt's way instead of plastering half of the screen with tabs and meaningless icons like other applications do - which doesn't mean more productivity but actually less (more productivity does automatically mean more room for creativity). This can be on by default but should be optional - like the most. Then an application should be adaptive. Meaning that a user can configure his workplace like he needs it to be. Blender is already (and has ever been) highly adaptive (though there is a lot of more potential do so - with the possibility to change key settings after the event system rewrite there is more - and with a system like Matt has tested there would even be a lot of more: Just imagine to start scripted tools and functions via a click on the screen instead of going down in a menu structure.) and the navigation through the different screens is extremly efficient and easy (nearly identic with some of Raskin's proposals). I don't know if I'm right but I think to have read something that indicates that Sabrina might have missed that point.

Perfectly, we agree. And I don't expect a clone of some existing interface, also because I much appreciate the originality of Blender (ok, maybe we could take inspiration from the UI of Bryce ).

To this originality of Blender belong things like pressing a hotkey that pops up a menu and hitting a NUMKEY to select the entry directly or the way screens can be changed and organized. And there is much more! You should probably take notice that a beginner doesn't know the true power behind an interface that he might dislike on the first sight. And a better accessibility does not justify to lose such a power in any case. Also the past has shown that changes made because some people that weren't into the Blender interface caused quite some problems with the UI consistency. So I only can warn again to be cautious what to change in which way because it can break some of the advantages the Blender UI offers.

Also: Eyecandy is good - but functionality and a natural, ergonomic and effective access to it is better - at least in the long run.

[Antlab]

In Tons Siggraph report during the meeting - he mentioned that about a dozen people told him they were waiting on adopting Blender till the interface/customization refactor. It is definitely a priority for him - so please - stop the flamewars already

LetterRip

But if this is one of the few threads on Blender interface that is remaining civil ..
Jokes aside, I am happy to read that Blender developers have fully realised how much important is for many people a substantial revamp of the interface, in particular after years of many improvements in other areas of the program.

Ciao

Antonino

[GDP_Sabrina]

The first step to do would be better documentation.

IMHO, the documentation of Blender is extraordinary ! All those tutorials, video tutorials, wikis and all those things demonstrate the community and the personal engagement of so many addicted users. But personally, I don't think that more documentation or pdf files with hotkeys will be the key to new (beginner or occasional) users.

Though there are so many wonderful tutorials, I have never made it to create a fluid water or wind example, because I simply didn't find my way through the shortcut jungle. I would like to _see_ all possible features and functionality grouped in toolbar comboboxes and play around with them. With every new release (like the upcoming 2.45) I see so many interesting features, but after being excited for a couple of seconds, I always remember: It will be hard way again to find out, how these goals will be achievd using the blender interface. The problem for me is, that I don't understand the blender UI philosophy (if there is any, I don't understand it). For me, it is a bunch of shortcuts, which you have to combine as shortcut patterns to achieve useful results.

Right now, I am really willing to learn from tutorials, but it is no fun crawling around in videos to find out hotkey combinations for creating effects, where other applications just offer simple to understand dropdowns with story-telling icons.

Regards
Sabrina

[kAinStein]

In Tons Siggraph report during the meeting - he mentioned that about a dozen people told him they were waiting on adopting Blender till the interface/customization refactor. It is definitely a priority for him - so please - stop the flamewars already

LetterRip

I also would wait to migrate from one software package to some other after a bigger change has been done. That would minimize the cost for the educational part of such a migration.

Do already some rough drafts exist how things can be changed in a way that nothing gets broken?

Not to speak about the documentation problem (How long ago the subsurf button has become a modifier? Take a look into News & Chat how many people are constantly asking about it. It counts to avoid some cheap shots that have been done in the past - that's probably better for further development - take for example the way Python scripts are organized: Though things have become surely better, it's still far from perfect because there is nothing that avoids scripts to register directly in the UI and act as true plugins with a "transparent" execution from the user's point of view - so a user wouldn't notice the difference between native functionality and a Python plugin. And there's a lot of more.).

So instead of having this discussion we could already discuss those - which probably would make more sense.

[kAinStein]

But if this is one of the few threads on Blender interface that is remaining civil ..

Oh this is so true! The discussion has become calm and is going in a direction with more understanding I guess...

[kAinStein]

The first step to do would be better documentation.

IMHO, the documentation of Blender is extraordinary ! All those tutorials, video tutorials, wikis and all those things demonstrate the community and the personal engagement of so many addicted users. But personally, I don't think that more documentation or pdf files with hotkeys will be the key to new (beginner or occasional) users.

...

Right now, I am really willing to learn from tutorials, but it is no fun crawling around in videos to find out hotkey combinations for creating effects, where other applications just offer simple to understand dropdowns with story-telling icons.

Regards
Sabrina

Yeah, I know. That's why I recommended to take close look at really basic stuff. I know that this is no fun! Of course making some real cool stuff would be a lot of more! I've got full understanding for your situation. But you can save a lot of time if you follow the recommendation.

There are not that many keys for a humble beginning: You need to have some knowledge about the modes and then R, S, G for basic transformations (or use the widgets), E for extrude, W for specials. The rest can be done with the menus (actually nearly everything can be made using the menus - but I would not recommend that). The rest comes by itself and with time. I guess you will get it rather fast. I'm pretty sure you will! But start with basic stuff...

EDIT: Oh! With "better documentation" I did not mean the quantity (or even quality - there's a lot of really good docu out there). But most of the "official" documentation has become quite outdated.