Output image preview
How about nodebox provide a output image preview feature? I mean make the node which outputs geometry drawing a preview of the geometry. It's good for debugging.
And I hope the rendered node which is selected do not change after I create a new node until I indicate a ndoe as rendered node manually.
Comments are currently closed for this discussion. You can start a new one.
|?||Show this help|
|ESC||Blurs the current field|
|r||Focus the comment reply box|
|^ + ↩||Submit the comment|
You can use
Command ⌘ instead of
Control ^ on Mac
Support Staff 1 Posted by john on 05 Jan, 2021 12:24 AM
At first I thought you were talking about the output of Nodebox itself (when you export your current image as a PNG, SVG, or PDF). But the canvas area already gives you a preview of that.
Reading again, I think what you are suggesting is that, instead of the little rectangles with icons Nodebox currently uses to represent each node, it should instead show a preview of what would be displayed on the canvas if that particular node was rendered. Is that right?
If so, that's an interesting idea. I can immediately see problems. The output is sometimes quite complex and would become unreadable if squished into the size of a node. Some nodes produce long lists or tables with hundreds of thousands of rows, or appear very different if data mode instead of viewer mode. You could make the nodes bigger to handle some of these variations, but that would cause sprawling networks and more scrolling, making it harder and more time consuming to scan, navigate, and understand networks.
If I understand your second point you are also saying you would not want all nodes to constantly update automatically as new nodes are added or changes made upstream, but instead want manual control over when to update the representation of each node. I can see why you might want that level of control, but adding that feature would further complicate the UI, add more work for the user, and might cause confusion.
None of this is likely to happen anytime soon in any case. For better or worse, Nodebox is likely to stay as it is now for some time to come.
But this is an issue I've devoted some thought to. Just as you are using Nodebox to represent circuits, Nodebox can be used to visualize itself. In fact, I created a Nodebox network that could draw itself years ago. My drawing stuck close to the way Nodebox represents nodes, but there's no reason it had to.
Nodebox is extremely good at visualizing complex information, and, in the right hands, can illuminate hidden aspects of systems. If you turn that power on the problem of debugging and understanding a Nodebox network the possibilities are unlimited.
This is a project that has been in my own hopper for years now, and I hope to turn to it again sometime soon. There are MANY things to try. You could use color to represent performance bottlenecks, distinguish value links from list links, represent ALL the nodes, including those currently hidden inside subnetworks, change the shapes used for certain classes of nodes to make their function more obvious at a glance - the list goes on and on.
My plan is to offer a general purpose node that can convert any .ndbx file into a table along with other sample code to illustrate some of these ideas and release that to the Nodebox community. It would be great if other people could pitch in and come up with news ways of representing a Nodebox network.
To that list of possibilities I will now add your idea to somehow represent previews of what each node produces. As I said above it would be hard to do well, but might be worth a try.
Thanks for an interesting suggestion.
2 Posted by ahuman on 05 Jan, 2021 04:24 AM
About the output previewing, you get it. But about the rendered node, I mean I hope the rendered node do not change into be another one after I add a new node.
For example, suppose that I have already finished adding some nodes to draw a graph. Then I want to add a number node to make the graph parameterized. After I add a number node, the reneder node would be changed into the number node. I don't want it to be changed.
Support Staff 3 Posted by john on 05 Jan, 2021 06:37 AM
That makes sense. More generally, for any parametrized graphical output you might want to choose previews based on a particular or typical set of parameters.
As I said, I don't see this as feasible for the working UI for all the reasons I mentioned above. But if you wanted to make a custom representation of a particular (relatively small) Nodebox network, and were willing to take the time to choose the best preview for every node in that network, it could make for a beautiful and instructional visualization.
I will keep this idea in mind when I get around to experimenting with possible Nodebox network visualizations.
ahuman closed this discussion on 05 Jan, 2021 07:27 AM.