I find myself a lot away from the computer and with little bits of time to spend on research and ideas.
After many years of designing with GH, I can many times work on it without looking at the Rhino viewport, hence the idea.
As GH files are mostly (if I’m not wrong) XML code, would it be feasible to create a standalone/mobile editor for the graphs that can save GH files that could later be opened in GH+Rhino?
We could fall in many errors, but I think the wonderful error proofing system GH is stable enough for it to not be a concern and regarding posible f€&k ups regarding impossible definitions… We have all already been there once or twice inside GH, haven’t we?
1 Like
Grasshopper files can be saved as xml, but to say they are xml is not correct in my opinion. Furthermore the precise layout of the xml is not some strict standard, each object and component individually decides how to (de)serialise itself. That means each component must be actually loaded into memory before it can be included into a file. For components to be loaded, other assemblies will need to be loaded. If this includes RhinoCommon, then it will also require Rhino itself.
The Grasshopper UI itself also requires RhinoCommon in some places, so a significant amount of functionality would have to be re-implemented in order to disassociate GH from Rhino.
Possible solutions to these problems include:
- A stand-alone notation which represents a gh network. Some time and effort has already been spend on a human-centric textual format that can represent most of the attributes of a gh file, but it was never implemented. This textual format could have a visual representation as well, if someone is willing to make a UI for it.
- Somehow access an actual instance of Rhino+GH in the cloooooud.
- Pen and paper (my personal favourite).
1 Like
Your first solution is precisely what I was trying to express.
Yess. I say only yes. If that would exist I would immediately make a code generator for Enterprise Architect, my favorite tool besides Rhino/GrassHopper, to generate gh networks from std UML class & action diagrams. That would be one way to work “unplugged” from Rhino/GH.
Wow, that would be something.
// Rolf
Sure, but the question is to what degree should it be similar? Should the stand-alone editor know about the topology of each component (what inputs and outputs it has)? If not, do you really think you remember them all? If yes, what about plug-in components, should they be included? What about icons and tooltips?
How many people do you think could use and would enjoy using a program that’s the GH equivalent of a blindfolded chess simulator?
WHY, WHY, WHY?
Name (full name and short name), inputs names and types. “A capsule is a class”, or an instance of.
Any component with a name and ports, plugins or not. The network just doesn’t instantiate if not all named classes (I mean, components) are installed.
One can start by writing the netwrok-def file using Notepad. text format 1.0 will do for ever (any class diagram notation does the trick). Same basic idea as if replacing Visual Studio with Notepad. Now we don’t have to, but most people get the idea.
Dump a “xml-library” of all installed components. Only then a GH Editor (wiyth "intellisense) will make more sense than Notepad. Or any code generator (which I have in mind).
It’s called MDD. More and more people understand that idea. Doing it from a text editor is actually a good idea. It opens GrassHopper up to the rest of the IT-world. Now it’s in the closet.
Text, text, text. Software “communicate” via text. At least in the development stage.
Sorry for highjacking the thread, but I’ve asked before about making GH able to dump and import GH definitions from pure text.
One reason is Version Control. Another is Code Generation. The most basic stuff in software development.
Security is not the problem. Block the function in “sealed” components. I don’t want to steal others work, but I do appreciate access to my own work. At least while working with it. 
As anyone does when coding Visual Studio or whatever.
Do not make it to look like a problem that which has not been a problem for decades for every other tool and “eco-system” apart from GrassHopper.
GrassHopper has potential. That’s why. Have a nice weekend.
// Rolf
Edit: Although my personal experience from Model Driven Design (and executable models defined in plain text, Notepad if you wish) was not in the field of aircraft and jet fighters (Swedish “Gripen” in the linked text below) it is actually used also in the software control systems of jets fighters. I think one can say that MDD is useful for many fields of application between Business Systems and Jet Fighters. So automating also the design process of those beasts, including CAD, isn’t beyond in any way.
From the link:
"…model-based development methodologies, the development of Gripen (the Swedish jet fighter) Es computer platform and software architecture…
Your future challenge
To successfully develop a world-leading fighter like Gripen E while being cost-effective, we place high demands on the development methodology we develop. One way we have chosen is that a large part of all Gripen E software is today generated from models.
Google translation (original only in Swedish):
Models. Text. Plain text. Version Control. Security is no problem.
Edit: +link
1 Like
As similar as possible. It should help a lot to use it “blindfolded”.
Of course it should know about the topology. If not, the later use of the definitioms in the real GH would be impossible.
As similar as possible, already said. Food4rhino provides an extensive database of maaany of the existing add-ons for GH so it should be possible not only to have every existing component but also new ones created in the very same editor.with new topologies like when we build C++ nodes or similar.
I don’t remember them all, of course, but if they are all preloaded, I can call them.
I’m not talking about installing all the functional components but just a graph editor that lets me place and link components, and that when I save the file and open it in grssshopper, this file is what GH engine needs to properly and fully work.
How many people did you thought would use GH when you embarked in this adventure? I discovered and got hooked on 2009 (helped us win our first architectural competition). Give your faithful users some credit 
1 Like
Actually, back up. You want this to run on a mobile phone or some other device that doesn’t have windows and thus cannot have Rhino? Or do you want this as a sort of freeware stand-alone that will run on any windows without the need for a Rhino license?
1 Like
First one. I had the idea thinking how cool and productive it would be to be able to “sketch” on the iPad or phone while commuting to work or during coffee break away from the desktop, send the sketch to the desktop and open and test it in the proper GH. The foreseeable red components would be fixable then… And continue the normal workflow.
1 Like
Well, ok. Becoming an app developer wasn’t really on my wishlist, but I guess the first step is to create a more easily writeable/readable/parseable textual notation for a GH network, which is something that was on the long term list anyway. Once this exists, making a standalone UI for it is something that anyone with sufficient skill can do (and my skill when it comes to apps is zilch, this is the phone I use, and I didn’t even add a memory card for the camera).
3 Likes