Blender taking off?

And yet in the linked quote he also professes how he instantly loved Modo the first time he used it. Presumably he had the same lack of motivation to learn Modo so why the different response?

I think xrg’s observations are spot on. If people with lots of experience and the ability to compare your application to other similar applications keep telling you that your application is hard to use, maybe it actually is. That doesn’t mean you should turn it in to a clone of something else, but it might be worth taking those notes on board and looking at what other applications are doing right.

I agree 100%, being different is fine, if it adds value for the user. Being different to show how clever of a developer you are, or because it was slightly easier to code that way, or because that’s the way it has always been done; is likely bad design.

It’s also inverted compared to the real-world hand drawn animation and drafting techniques where those other applications drew their inspiration. In the real-world you have a background layer that physical cells or vellum layers are placed on top of to add components to the drawing.

For me it’s all of Blender’s little inconsistencies like this that are the most frustrating, doubly so because they are really hard to get addressed. The way that Blender’s issue tracker is managed, I would assume that any bugs filed about this would be closed as “Not a bug” because it is, in fact, working as designed. Even if that design is not what the user expects and is arguably a bad choice. What other option do users have for raising these issues? 40 page long threads a la Colored Wireframes?

I agree that is a design flaw. The same happens in the node editor -

Actually, this one is a consistency; In blender all calculations are done from the top, and that includes Modifiers, Layers and Nodes.

But it’s a case where actually breaking consistency where image layers are involved (or layering images in the Node Editor) would be beneficial.

When you instantly love a software, you are more inclined to put in effort to learn it. You therefor learn it quicker. It will “seem” easier. If you’re half-heartedly trying to learn something, you’ll learn slower. It will seem harder. That’s what i meant by there being a perceptual component.

I’m not saying blender doesn’t have quirks or problems, I am arguing against this myth that blender is fundamentally wrong and that all these other softwares have it “right”. As if Blender is in its own weird corner and all the other softwares are hanging out together. The truth is they all have quirks, they all do things that are weird and they all develop in their own paradigms. Just like Blender.
Some of these softwares have worse UIs and are harder to learn (ZBrush,Houdini) yet still manage to be top industry tools. So there’s obviously something else going than continuously dredging up this old argument that blender’s UI is to blame.

The thing is, first impressions mean a lot to people (not to everyone, but most expect to feel the power of the software within the first few minutes).

It’s possible someone could open Blender, select the ‘Maya’ preset, left-click the cube, and find out he moved the 3D cursor instead. This is just one reason why it would be paramount to upgrade the control presets to include a full automatic change of the preferences rather than just changing the hotkeys (so that they would be encouraged to explore more when the left-click selection paradigm actually works).

It may be consistent with those parts but in addition to being inconsistent with what most artist would probably expect it is also inconsistent with the VSE, which uses the bottom up convention.

I would argue that modifiers, at least, are also a bit inconsistent. I always need to stop and think about what order they should be in when switching between interacting with modifiers via the Python API and the main UI. In the Python API modifiers are essentially presented as an abstract stack with push and pop operators. In my mind a stack is a bottom up data type so the Python API seems right to me while the top down representation in the UI always seems backwards.

Personally, I would prefer that all of these areas in Blender consistently follow the bottom up convention, with a base object/image at the bottom of the stack and modifications stacked on top of each other. I think a pretty strong argument could be made that this is the ‘expected’ way of doing things. Other than being the way Blender has always done it what advantage does the top down convention offer?

I don’t know, i have trouble with this argument. Why stop at the mouse button? The spacebar doesn’t do the same thing in Maya (hotbox) as it does in 3DSmax (selection lock) as in Softimage (selection tool). Yet I’ve never seen any outcry over this from any of those users. Even though the 3 platforms were developed by the same company. You learn a new software, you expect to find differences. If the first difference you find seriously discourages you from learning the software, your heart was never really into it.

It discourages to explore further because it doesn`t make sense when 99% other software uses left click to select. It requires a concious effort to use it at first and leaves a negative first impression. When a basic thing like object select does not work like in every software ever, it might leave an impression that the rest of Blender is just as awkward.

And a lot of people never get over it, me included, we just switch buttons in preferences (which also requires effort). For people who are exploring Blender as a potentional alternative or supplement to established software, this can be quite a significant turn off. For people who intend to use it as the main package for various reasons that of course is not an obstacle but it`s still annoying for the most part.

You really count doing a single-time change in the preferences as “effort” ? common… That’s a bit of a stretch.
And for requiring conscious effort at first which over time becomes second nature ; that describes all learning ever.

A new user does not know where preferences are or where to change this option, an effort however small is still an effort. And after several months of actively using right click as select I still could not get used to it because all the other software I use has left click select so I changed it to left click.

Some standards should not be broken. Its not just 3D app standard, its literally standard for most software out there, OS included.

That makes sense. I’ll try to push for this.

I guess it was set this way due to how we read text, or how code is executed.

Couple of things since last night. I’ll have to mash them together :wink:

Original Meaning of “Blender is for Blender Users”

It is possible it did mean as you posit, Campbell. Only one person truly knows what they meant by that line. Unfortunately, as you seem to acknowledge by your phrasing, we can only throw possibilities of it’s intent around because the person who actually coined the phrase refuses to clarify it for us noisy forum users… leaving it defined by how it is used here & elsewhere on the Internet (which is far less nuanced & balanced than your guess as to it’s desired meaning).

It’s not that important, all things considered, compared to the general attitude one gets exposed to when suggesting changes aimed at increasing adoption &/or speed of learning. There is a distinct air of “Acceot the Blender way or GTFO” to most discussions about such things. Even a few of the posts in this thread are moving toward that position. :frowning:

zBrush Example

I’m sorry folks, but this is an incredibly silly one to bring up. zBrush brought something to the table that nothing else did. People got over the quirks in user interface and odd way of doing things because, in the end, they got something out of that effort they could get nowhere else.

Blender doesn’t hold such a position in the industry. It hasn’t got the best/fastest modelling, rendering, animation, texturing, sculpting, or anything else for that matter. I’m not saying it’s the worst package in the field, because it isn’t, but it isn’t the best at anything either, so comparing it to a piece of software that revolutionised digital sculpting and provided a level of detail in the final result you could not get anywhere else is comparing Hera’s golden apples with somewhat old & nasty looking oranges.

There is already better modelling, sculpting, texture painting, rendering, etc software out there. They also happen to share enough interface standards that they are easy (or easier than Blender, at least) for someone familiar in one to quickly pick up on the other. zBrush gets away with flaunting such standards with professionals because it brings them something the other packages cannot. Blender would get the same flexibility showed it’s interface quirks were it wowing them with something they couldn’t get elsewhere too.

Honest question - in an environment where time = money, what feature does Blender bring to the table that justifies the time needed to get used to it’s contradictory (to other apps in the pipeline) user interface?

Left / Right Mouse Argument

Not going to get into the argument, because it’s been thrashed out heavily in the past and the decision to go with the left-mouse select default has already been made according to the page where developers were discussing it. What I do find interesting is that the instant the subject of standards come up, everyone immediately asks that no-one discuss that particular example. It feels a little like discussing how individual developers affect the development of Blender but being asked not to discuss Ton at all in that context. laugh

When people see the term ‘Maya Preset’, they expect it will change aspects of interaction to make things somewhat more like Maya, otherwise the name might be little more than a misnomer at best and misleading at worst (it changes a few keys, that’s it). I don’t think a Maya user in any sense would want that preset to turn Blender into a Maya clone, but at the least have a few basics performed like changing the selection key from right-click to left-click.

I wouldn’t worry about it if the code started allowing those presets to go further like auto-changing preferences, the default will still be Blender.

Also, saying things like “your heart was not into learning Blender” tends to be derided by commercial software users as they already accuse the community here of finding excuses to ignore their needs completely, even if they’re basic and can be easily changed like left-click select in, again, the Maya preset.

@BTolputt - Ton could clarify Blender is for Blender users, but in his talk at last BConf - I think he explains it here, expect you saw it already, but for reference.

re: zBrush Example, Ok, so Blender (arguably) doesn’t (yet) have any killer feature (though some would disagree here), … lets work on that :slight_smile:
Its not Blender’s fault there are very few open-source 3d tools in existence.
Others are free to write their own tools which can follow so-called industry-standards nicely.

re: being asked not to discuss - This derails too many threads, if you want to discuss divisive UI topics which were already discussed many times - Open a new thread or comment on existing one. Discussing is fine, but continually de-railing isn’t.

some random thoughts on this:

  • in general its interface does shine with tablet input, virtually no other application can cope with the haptic feedback. (Tabs have been slightly a step in the wrong direction, although they were needed…not arguing that)

  • the whole uv->texture (and partly) paint workflow is a bit rough and still needs some tools, but once you get used to it its pretty fast

  • there is also a lot of henn/egg situations going on: For example, in other software you usually work with a more or less fixed or task based layout, while in blender you usually work with a workflow based layout (adapts to the current workflow). Now the problem is, as long as you don’t get used to the concept and didn’t know how the interface is changed (splitting/merging windows switch to different layouts / screens) you get a lot of frustration. As soon as you get used to it, you actually want such flexibility in other software (again imho).

What could help would be to have some basic navigation introduction videos, like maya has when you start it the first time.

  • and I know I’m lobbying now, but the general concept of using 2point polys (or edges) with a subdivision surfaces modifier as a curve based modeling replacement. (curve based modeling in other applications is usually less flexible than this method.) This method is pretty unique to blender (only application I know wich allows this too is houdini, not sure about max)

  • animation is not bad either.

  • very bad: multible object / surfacing properties editing. time killer

I did, indeed, watch that one Campbell. Remember the claim about him being “absolutely powerless” and how you don’t even listen to him? Same conference. :wink:

Thing is, even the meaning outlined in that speech isn’t as fair & nuanced as you posit. That speech (and accompanying text) pretends that the world is black & white. You either are a Blender user (& aren’t mentioning how Blender’s interface could be altered to better meet defacto standards set & used elsewhere by professionals) or your someone that doesn’t use Blender or attend the conference (& so must be commenting from ignorance). The world (& Blender userbase) isn’t that simple though.

Blender users are more than just the people that agree with Ton’s interface decisions and can fly across the planet to attend a conference. They are everyone from the guy that create this tutorial, through the guy who helped pay for Rigify development, to the guy that runs a successful Blender-specific tutorial website. They too get brushed off with that same line when making comments about how Blender could be better if it adopted one or two standards professionals expect.

Anyhow, I didn’t want to be discussing Ton’s line specifically. I was more commenting about the attitude exhibited toward anyone making suggestions about how Blender could be changed by people who know what their talking about professionally, but aren’t names in the Blender community*.

My point wasn’t to rag on Blender. zBrush can ignore defacto standards so long as it is useful because there isn’t an easier tool that does what it can do. Blender isn’t in that position. When it gets into that position, I reckon one of two things is going to happen - professionals will use it a lot more often because they must have access to that killer feature… or someone forks Blender, applies some of those defacto standards, and professionals will use that distribution. Until then, professionals will (for the most part) use what is easier for them to learn because, when time = money, one needs to balance time learning a counter-intuitive interface with the time one could otherwise be making a living.

Um, not exactly sure what this is referencing. Whether a user interface follows defacto standards or strikes out on it’s own isn’t divided by commercial vs open-source. Plenty of successful open-source projects following defacto interface standards and the examples provided in this very thread of software that succeed without following them are proprietary :slight_smile:

I think you misunderstand me again. I don’t want or need to discuss it. I just find it amusing that whenever the subject of adhering to user standards comes up, the first thing people try to prevent discussion of is the best exemplar of such in Blender. I found it amusing & thought I’d share. I wasn’t trying to re-discuss the issue, mostly because that particular battle has been fought & won already. Apologies if it appeared otherwise :slight_smile:

[SUB]* I honestly don’t believe that the UI Debate would have had any legs whatsoever were it brought up by someone other than Andrew Price. Whilst he didn’t have much pull with the Blender Foundation itself - he did have quite a bit of support amongst the very people that feel the interface weirdness strongest - newer Blender users seeking instruction to understand the application.[/SUB]

@BTolputt - Hrm… fair point, but its probably over simplified for purposes of discussion, of course there are all kinds of users, blender users dont just use blender… etc.
Andrew Price has since become apart of the UI team, so I wouldn’t say he got the brush-off, we value his opinion.

Yep, licensing vs defacto standards are mostly unrelated, just asserting that blender isn’t necessarily for everyone, users who prefer other apps better use those - or invest in bringing Blender up to speed if they have the resources.

I agree, now his comments on the matter are not being ignored. Before the forum melted down during the UI Debate, the very same arguments he was making were brushed off, and I bet you the first round at pub of your choice I can find posts where that “Blender is for Blender Users” line was applied to his commentary. :wink:

Of course, as the saying goes, the proof is in the pudding. It is not unheard of for people to be given positions that they find are ineffectual in driving change. Look at what happens in politics :wink: I look forward to seeing actual implementation of decisions made by said team in the future. The mouse button one especially. :wink: :stuck_out_tongue:

I’d be interested in what kind of investment you’re actually speaking about. Is there a price tag on certain user interface principles, for example. Or is there a certain number of hours one has to be developing a feature in order to get BF approval? If so, what are these numbers? How does one invest for such things outside the donation channels, given the frequently raised point that “donation is not investment”? And so on.

The investment line can seem a little trite to those who donate, pay for features, pay for open films we’re not interested in actually watching, etc. We all know there are certain areas where the final decision has next to nothing to do with investment and almost everything to do with who gets to approve code going into the official distribution. User interface decisions (i.e. that raised by the zBrush comparison) are, as recent experience has ably demonstrated, very much one of those areas.

@BTolputt, Me and Thomas talked with Andrew before he published his new UI proposal and said I was sure it wasn’t going to be accepted (and indicated we didnt have the dev team big enough to make the changes in a reasonable timeframe) - & didn’t follow the reaction from the community after. (iirc my response wasn’t totally negative - he did show some nice ideas too).

Re: other UI stuff, Look forward to pie-menus, they’re pretty close to being ready :slight_smile:

Re: Investment, not referring to UI stuff much, this better be handled by design team with some authority to make the changes - But there are many examples of companies who hire devs to work on some feature they need. Though often these are weeks of dev time, so not something single users would typically go in for.

In my specific case:
I used SoftImage in my first employment till 2000 (which at that time broke many industry standards, undo=u unique but cool ui) then had to switch to Maya in my second.
Since I don’t believe in software piracy, all my previous work was then unaccessible for me. Also since the Maya license we bought was tied to one machine we soon had to do 3D on the slowest computer in the agency because time vs. money didn’t justify permanent updates for a company only using 3D occasionally.

Plus I couldn’t do any private work at home.

So with the first Alpha of blender 2.5 I gave blender a try for a private project and found the switch much easier than from SI to Maya. I soon started using it in the office as well.
After that I suddenly had the possibility to turn our office into a render farm when necessary which would have cost a fortune with mental ray licenses.

And yes I actually like blender’s ui and find it much less contradictory than for example vektor editing in various Adobe applications (Photoshop, Illustrator, After Effects, Flash) that are full of contradictions.

Maya currently costs 1.794 Euros per Year which is much cheaper than a few years ago but together with the savings in render licenses would still justify a week or two of training for an artist in many (not all!) use cases.

Second part of your question :wink:
Yeah, what feature could that be? - Many would like it to be ease of use probably (judging from all those annoying UI threads) but it will be hard to beat Sketchup in that respect.
If the discussion here would find an answer, a use case where blender could really shine, then we’re back on topic of blender taking off.

Maybe to make it a perfect fit for unreal game developers could be an achievable short term goal.
Since I don’t develop games, I wouldn’t be too exited about that though…