| Zombie0915 |
Hello Everybody, I am looking for advice for a project that I am working on, and need a clue from somebody who is familiar with server side programming(specifically CGI) and web development.
My project is a web based application that can be used to control a specific device. The device has a command line type interface which is used to control it, it is connected to a gateway which issues the commands based on input that the user provides to some web forms. This is all done with CGI scripts running on the gateway, it generates the forms, takes the input, forms the commands, opens a connection and sends the commands to the device to be executed. Then the device sends back output and the CGI scripts parse it and give a nice graphical display that represents the information to the user.
This application also has a window that acts much like a terminal emulator allowing the user to type out raw commands and send them directly to the device without using the menus. This part of the application is where I need a clue. Currently the terminal page interacts directly with the socket to send the commands, when the window is closed the user is logged off of the command session. I need to separate the frontend and backend of this web page terminal. I want the user to ba able to open that window, issue some commands, then be able to close it WITHOUT ending the command session, then I want them to be able to re-open that window at a later time and be able to pick up where they left off. So it is like a web based terminal that can detach and reattach much like the way GNU Screen can detach and reattach to a session of command line programs in *nix.
The problem is that I have no idea how to implement this in a web based application and need an idea of the basic components required to realize this goal. I really have no idea how to make a web app that spawns a background process that isnt prone to abuse. I am also not quite sure that I want to put another layer between the user and the server that this interface is meant to control, this would mean that the user would have to go through the web page, the web page through this backend, then the backend would be communicating with the actual server and it just sounds kinda inneficient. Has anyone done anything similar to this before? It seems like something that one of you might know about, or at least know somebody who knows, or even knows of a better place where I can ask this. I Know it is kinda useless trying to get info like this from the COR but if you have something you think might be helpful to me just post it :).
thanks |
|
|
