How to build a processing sketch loader ?

Programming Questions

owaunandr..

How to build a processing sketch loader ?

in Programming Questions • 3 years ago

I would like to build an electronic sketch viewer of sorts. I can open alternate windows with any type renderer but outside of copying and pasting the code and then hitting the run button I want to develop a more seamless transition method.

It would be like an electronic version of the defunct builtwithprocessing.org.

Think of what an electronic version of openprocessing.org would look like.

There is a site where you can type the openprocessing.org id and you are returned with the contents.

http://emoc.org/opcode/

I believe it would be possible, where you just select the sketches from a folder and the script builds a series of sketches and you just pick from the selected.

However, I want to know if it is possible to do this on the fly ?

I have tons of awesome sketches I would like to share.

I think it is possible.

Replies(4)

noncom

Re: How to build a processing sketch loader ?

3 years ago

I guess that the most simple way to do this is to have a copy of a program exported from processing as an applet and it's source code presented separately. All else is html or php or java too...

I doubt that it is possible to embed a processing sketch into performance on the fly.. you know, it requires compilation and stuff...

What else I can think of: install processing on the server side and create a script that would automatically parse a selected sketch with processing (using maybe some command line parameters) and creating a page with the applet. Basicaly it is the same thing that I wrote in the first paragraph except that if you have an overwhelming number of sketches, this way helps you to automate the process.

owaunandr..

Re: How to build a processing sketch loader ?

3 years ago

I agree with you. Having the application bring up a web page with the exported sketch, however, I foresaw frame rate issues with that but I like the idea.

I want to be able to even change the values on the fly if I needed to like hit a pause button change a variable and then hit resume.

However, I will try your idea and see how the frame rate really is. Maybe with optimization you could have a DHTML processing sketch viewer. Even with a .txt file, the code lines still have to be compiled, so I agree with you.

noncom

Re: How to build a processing sketch loader ?

3 years ago

I want to be able to even change the values on the fly if I needed to like hit a pause button change a variable and then hit resume.

Well, actually it is impossible. A java program must first be compiled into some sort of a feed file for the JVM (I am not quite a professional with this part, just the general idea). And it could only be run after the compilation. Changing variables on the fly would require to recompile the source. So it leads to creating a compiler or, better say, an interpreter. And if you are going to create a java interpreter, you need a clear mind, much knowledge and tonns of spare time.

However, I can think of a somewhat.. "solution" for this problem. It is about including a 'variables' class into your programs and making all the program variables available for monitoring and changing by it's methods. So instead of 'x' you would now always write 'variables.x'. Or you can try arrays instead of a class.

real power comes to this idea when you redefine all the equations in the program through classess thus making it possible to modify them and even make new variables... expanded further, this concept would allow you also to define as class objects not only variables, equations and stuff, but also functions, conditioning and other. This way your program's morphing capabilities could rise to a very high level...

But really it is quite tricky and unnecessary to implement. Also I am not sure about the speed.. I would rather just install Processing on the server and re-compile the programs on-the-fly. It would only take maximum of 10 seconds to re-compile the applet and present user a page with it.

However, I will try your idea and see how the frame rate really is. Maybe with optimization you could have a DHTML processing sketch viewer. Even with a .txt file, the code lines still have to be compiled, so I agree with you.

Then you should really write a translator from java to javascript since javascript is a completely different thing from java.. and it is much more primitive, so it may be impossible.

owaunandr..

Re: How to build a processing sketch loader ?

3 years ago

noncom you made some good points. Javascript I find is moree suited for online use and more like processingjs.

I really like your idea of the class variables and the ability to create new variables. I am a fan of live coding yet i have never been able to come up with a construct that is effective enough even for the simplest for loop.

e.g.

If I wanted to change

for (int x = 0 ;x<10;x++)

into

for (int y=4;y<45;y++)

during the execution of the program, the best i could do is create a for loop like this

for ( int _x = xstart; _x < xmax; _x += xby)

That way I can change the variables xstart,xmax and xby but _x is constant variable.
I believe I wold have to research it more.

It looks the only method is like what you said; creating an application that calls the program and that takes about 10 seconds. I just wanted a method so that I can choose the folder, the system reads the .pde as a simple text file, or a file I created that can be parsed into code. The difficulty is that like you said, it has to be compiled, and furthermore I am not capable of taking text and intuitively executing the code.

I would have to create pre-defined constructs that when the program reads
setup the program interprets and executes void setup()

I like your idea alot. I have emailed the person who run the opcode site, emoc. I will see if he has any ideas.

Top Reply