Example setup
The following steps will walk you through a basic setup of nFoldMan to monitor and control two folding@home processes on a single computer. (Network monitoring is covered in another part of this help file).
For this example we'll assume
- You are connected to the Internet.
- You have already downloaded the Stanford folding@home client (called "fah4" when this help was written).
You have created two directories -- "fah1" and "fah2" -- and you have a copy of the "fah4" program in each.
- You have not run the unix command-line client manually, so no configuration of the client has been done.
- You have found a unique user name for yourself - "myname" - by checking at the Stanford site.
- You have decided to join team 1971 - "Team Mac OS X"
Identify and configure the first client
Launch the nFoldMan program. A new document window will open. Click on the "Set Path" button to tell the program where the client is:
In the file-open dialog that results, navigate to the "fah4" (or equivalent) client in the "fah1" directory:
Click OK. You will see the path name is now filled in on the settings sheet. Since you have not configured the Folding@Home client yet, we will do that now. Click on the "Config Client" button:
![]()
A sheet will open to configure the client settings. Fill in your name, team number, and the number "1" for Machine ID:
![]()
Click OK to close the configuration window, and click OK again to close the settings window. Now you will notice two things:
An additional file -- "client.cfg" -- has appeared in the directory where the Stanford folding@home client is located. This configuration file will be read by the client - you should not touch it.
A document status window now appears. Eventually this will contain all the status information about the process you just defined. It doesn't contain any information yet because the process isn't running yet. (If you have taken a shortcut and pointed the document to an already-running client process, you will see information in this window.)
Test the first client
Click on the "Untitled" window to be sure it is at the front, then select "Start Process" from the "Process" menu:
![]()
Now open the Console utility from the Applications/Utilities menu, and you should see a lot of console lines going by as the Stanford client sets itself up, downloads its first work unit, and begins the computation:
After a few minutes (possibly ten or so) the status display will automatically refresh itself to show the details of the assigned work:
![]()
If you don't feel like waiting, you can periodically run the Refresh command from the Process menu:
![]()
The document will update with detailed information as soon as the started folding process is ready to produce it.
Note that the "Estimated time remaining" bar won't be filled in for 30 minutes or so. The client program must run for some time before enough information is available for nFoldMan to estimate how long the assigned work will take.
Now if you look in the directory where you downloaded the Stanford client, you will see that several more files have appeared, downloaded by the running client process:
![]()
You should never touch any of the files in this directory. They are managed by the downloaded client and by nFoldMan.
Finally, save the nFoldMan document you have created somewhere -- perhaps on your desktop.
This document is all the settings you need to run this client process, and you will generally interact with the process in the future by double-clicking on this file, or by invoking nFoldMan directly.
Identify and configure the second client
Now you'll repeat this process to set up the second client process. Repeat all of the above steps for the second client, except:
- Point to the fah4 program in the second directory you created.
- On the "configuration screen", put "2" in the Machine-ID field, instead of "1":
Tidy up the windows
After you start the second process and save the document file describing it, you can move the two document windows into a pleasing alignment somewhere on your screen:
![]()
Note that the process status windows also have a compact form. You click the little green "zoom" button to toggle between the detailed and compact forms:
![]()
Finally, put the two documents you just saved somewhere convenient:
Automate startup
Let's suppose you want nFoldMan to start automatically when you log in, and to launch the computation processes as soon as it starts. There are three steps:
- Tell your computer that nFoldMan should be launched automatically on startup.
- Tell nFoldMan that your two documents "fah1" and "fah2" should be opened automatically when the program launches.
- Tell each document "fah1" and "fah2" that it should automatically start the folding computation when the document opens.
Here's how we do these steps:
Congratulations, you are up and running.
- Tell your computer that nFoldMan should be launched automatically on startup.
Use your System Preferences / Accounts / Startup Items control panel to arrange that the nFoldMan application launches automatically when you log on.
- Tell nFoldMan that your two documents "fah1" and "fah2" should be opened automatically when the program launches.
Run the nFoldMan application, and select Preferences from the nFoldMan menu:
Ignore the Network tab for now. In the AutoOpen tab, click Add, and select one of the documents you saved a few moments ago. Then click Add again and select the second document. You'll end up with something like this:
![]()
Now these two documents will be opened automatically any time you directly launch nFoldMan. (Not if you start nFoldMan by double-clicking on a different document.)
- Tell each document "fah1" and "fah2" that it should automatically start the folding computation when the document opens.
This is actually the default state when you create a new document, so you don't need to do anything now. You can turn it off and make other changes through the Control tab of the Edit Process Settings menu item in nFoldMan's Process menu:
In this options sheet you'll find a variety of controls to fine-tune how you can control the folding process.