GUI opinions sought

Hi All,

since the last attempt at unchaotifying the toolbar panels was somewhat of a failure, I figured I'd ask your input ahead of time.

Instead of hiding a lot of components from the toolbars, I've now created separators that visually group related components. It requires more screen estate since separators take up an additional 10 pixels each, and not all icon columns are now fully used:

It is fairly trivial to draw the icons the old-fashioned way, so I can add an option for this feature if a lot of you are appalled and disgusted by this new scheme.

--
David Rutten
david@mcneel.com
Poprad, Slovakia

Replies to This Discussion

Permalink Reply by Vicente Soler on December 30, 2009 at 3:36am

That is better, but it's just a band-aid. the current scheme is unsustainable if new components keep appearing all time. I think the main problem for new users is the number of components. If you keep procrastinating about it many new users are going to get turned off from Grasshopper.

Permalink Reply by Pep Tornabell on December 30, 2009 at 3:43am

Hi All

I think its a good way,even more thinking of new user components (weaverbird, rabbit, etc..). But at the same time I have to agree with Vicente, more components more mess... but I think its something necessary, more comoponents, more flexibility and capabilities.

Permalink Reply by Damien Alomar on December 30, 2009 at 7:48am

I think that's an interesting observation, but no solution to the proposed problem. As I see it, its 6 of one, half dozen of the other. Yes there are a ton of icons, but in reality that's because there's a ton of things that can be done with GH. If the desire to do 1000 different things is there, then its hard to not have that be expressed as 1000 different components. I think this is part of the learning process of GH.

On the other hand, I believe that there are opportunities for component consolidation. For instance, there's a curve frame component that takes a curve parameter, and a curve frame component that takes an integer for multiple curve frames. There doesn't need to be two different components for this, but their are. The trade off is that if you want to do multiple curve frames with the single curve frame component you have to add a few more components to get it done. So its the question of the extra real estate of an icon versus a few more components on the canvas and a few extra steps to get the same thing done. For beginners, I think having the extra icon is more useful, and although I expect advanced users could deal with it either way, having a definition that's less crowded and cluttered would be considered a positive.

The other opportunity for component consolidation (because I doubt any components will just be tossed) would be to make certain operations, such as creating a circle, articulate itself as one component which can change its format as opposed to individual components of the several operations that can create that object. So keeping with the circle example, the standard circle component, circle 3 pt, cen nrml rad, and possibly the InCircle component would all be one uber circle component, and you'd have to select the type or method of circle creation within the component, probably by right-clicking. I'm not sure this is a good thing for new users either as it hides functionality just as much, if not more, than having an icon for each operation. Yes, it would save screen space, but at what cost.

So as I see it, its a push as to whether saving the screen space of the icons is actually worth it. There are opportunities for saving space, but at a cost that may not be worth price being paid. A lot may hinge on the changes made to the interface.

Permalink Reply by Vicente Soler on December 30, 2009 at 11:09am

"I think that's an interesting observation, but no solution to the proposed problem."
I'm not sure if this is directed to me. If it is:
I have proposed a solution several times but i didn't want to repeat myself again. I think David knows it. It's basically is to borrow heavily from Cinema's 4D Expresso. Rhino could also benefit a lot if they just decided to plagiarize how C4D's GUI works.

Permalink Reply by Damien Alomar on December 30, 2009 at 3:02pm

Done a few searches, but not getting enough of idea of the interface approach of Expresso. Any links to a good video showing how it works, or better yet one of the threads where you've gone through this before?

Permalink Reply by Vicente Soler on December 31, 2009 at 7:33am

"or better yet one of the threads where you've gone through this before?"

I found this thread:
http://www.grasshopper3d.com/forum/topics/wish-inverse-list-item
It's actually a reply I made to you. I probably had already explained it in another thread (since i mention I'm repeating myself), but can't find that.

Permalink Reply by Danny Boyes on December 30, 2009 at 4:28am

I like this new grouping feature.

Is there room for a user customized quick access tool bar where components can be easily added/removed by a right click option? This way the user can collect their most frequently used components together without having to switch tabs. (Similar to the Word 2007 QATB)

Permalink Reply by David Rutten on December 30, 2009 at 3:12pm

Hi Danny,

I've been keeping well clear of user customized UIs. It's very difficult for a third party library of components to insert itself into a custom workspace. I've never thought about having only a custom panel with popular components. It's a pretty good idea probably. However, before I start on that I'd like to see whether or not the Markov widget is lifting its weight. It takes a few days for it to 'get used' to the specific usage pattern of any given individual, so it will be a week or two before any meaningful feedback will appear on this.

--
David Rutten
david@mcneel.com
Poprad, Slovakia

Permalink Reply by Danny Boyes on December 30, 2009 at 4:04pm

I meant to ask how the Markov thing was going ... will it be in the next release for beta testing?

Permalink Reply by David Rutten on December 30, 2009 at 4:40pm

Yes it will. It's recording up to three steps and provides up to 10 (but that's customizable) suggestions for the next step. For example, my behaviour over the past few days has resulted in the following pattern based on a single Point Parameter:

--
David Rutten
david@mcneel.com
Poprad, Slovakia

Permalink Reply by Danny Boyes on December 31, 2009 at 1:14am

Very Nice.

Does the Markov widget reside as a static entity? ie a floating toolbar where ever you place it.

Have you thought about a radial menu accessed from the mouse wheel button? Like that found in the game Crysis to access suit controls. You click and then move in the direction of the component you wish to select, or would you be stepping on patents?

Permalink Reply by David Rutten on December 31, 2009 at 3:59am

At the moment it's fixed in the lower left corner of the canvas. I meant to make it draggable, but it proved to be too much work for the next release (hopefully still in 2009).

I know about those radial menus, and I like them a lot. I have a number of things I'd like to implement this way, but I don't know yet if we'd be violating patents, and whether or not the firms who own those patents are likely to pursue violations (most companies generally don't).

I'm not sure how well it would work for a Markov feature, since you'd have to click somewhere to find out whether or not there actually is a useful suggestion in the list. The whole point of a Markov chain is to eliminate this additional click.

--
David Rutten
david@mcneel.com
Poprad, Slovakia