Oct-21-2016, 04:58 AM
i have a plan to develop a program to dig deep into the Mandelbrot and Julia spaces and display the result in color. an intended feature is to do the heavy calculation in a separate process written in C. yes, i know i can join Python and C and do this in on process. i want to separate it so i can do the heavy calculation in the cloud. so my idea is to design a protocol for this based on JSON (easier in C than XML). an additional program in Python would receive requests, divide up the work space, and pass on the requests to other heavy calculation servers/instances. the result would be an array of data which would be 8-bit or 16-bit numbers giving the calculation counts at each requested point. the final client would take the numbers, lookup a color table, and display the result as a deep view in the Mandelbrot or Julia space.
i have zero experience designing any GUI program. i need to think how best to build the array of colorized numbers to be displayed. i also want to divide up the space into a 4x4 set of blocks and separate them by rows and columns of one pixel width in my client and use hotkeys to select the next deeper space to view. i may decide to also use mouse clicks for that.
suggestions on what i should read for this?
i do have experience writing programs to do image manipulation (in C). later i will write a "client" to take a high precision point and make a smooth logarithmic zoom down to that point and output a video. the first client will ultimately have an option to save an HD image (for use as a background, etc.). I know what i need to do for the heavy calculation server and the space splitting proxy. it is the GUI client that i need to think about. i have ideas for the protocol, too.
i have zero experience designing any GUI program. i need to think how best to build the array of colorized numbers to be displayed. i also want to divide up the space into a 4x4 set of blocks and separate them by rows and columns of one pixel width in my client and use hotkeys to select the next deeper space to view. i may decide to also use mouse clicks for that.
suggestions on what i should read for this?
i do have experience writing programs to do image manipulation (in C). later i will write a "client" to take a high precision point and make a smooth logarithmic zoom down to that point and output a video. the first client will ultimately have an option to save an HD image (for use as a background, etc.). I know what i need to do for the heavy calculation server and the space splitting proxy. it is the GUI client that i need to think about. i have ideas for the protocol, too.