Troubleshooting Broken PiPEs

  1. My command script doesn't give the expected molecular view. Use Preferences to designate yourself as an expert. After you run the offending script, use the option (below the message box) Edit script to view the script that is being executed. Simply examining the script closely may reveal the problem. Of course, you can actually edit it and try the changes. Desired changes would then have to be edited into the script in the PiPE Block of the *_pipe.pdb file.

  2. When you click the link to start a PiPE chapter, a new window opens for PE, two frames appear, the right frame turns black with MDL in the lower right, and nothing else happens. PE can stall during startup if a javascript error occurs (execution of an invalid javascript statement). This kind of problem is easier to diagnose in Netscape (not Internet Explorer). Type "javascript:" (be sure to include the colon at the end) in the location slot at the top of the browser, and press Enter. A new window, the "Javascript Console" should open. Press the Clear button to clear all the messages in the js console, many of which may have accumulated earlier and be irrelevant. Close the stalled PiPE window and attempt to re-start it. Now examine the reports in the js console. If no errors are reported, a javascript error is not the problem. In Netscape 7 and Mozilla one harmless "unhandled exception" error occurs normally during PE startup and can be ignored. (In Firefox, three occur.) If another kind of js error is reported, hopefully it will help you to find and correct the invalid javascript statement.

    Example: If you fail to quote the title, a fatal js error occurs stalling PE.

  3. I moved the folder containing my PiPEs. It worked fine before I moved it but now it doesn't start. In order for PiPEs to work, there must be a copy of Protein Explorer on the same computer or server, and it must be "side by side" with the directories my_molecules/pipes/, which must begin with exactly the directory names my_molecules/pipes/. For example, these two must be side by side in the same directory: Then, make sure that the file is present and correctly defines the path from itself to PE in the variable pe_path. In the example above, the correct path would be "../../protein_explorer_2.73_alpha/". You can have multiple versions of PE side by side; pe_path specifies which version your PiPEs will employ.

    Finally, make sure the paths to the files pe_path.js and pstart.js in pipes/template are correctly specified in the index.htm file in your PiPE. For example, if your PiPE is in

    then index.htm must contain near the top The number of steps back ("../") to template must be correct.

    See also publishing your PiPEs on a server.

  4. When I save a PiPE PDB file from PE (Chime) onto a Macintosh, the filename is truncated. When I load the saved file into PE, it fails with an error message that the filename should end in "_pipe.pdb". Netscape 4.8 (running in Classic) under Mac OSX cannot handle filenames longer than 31 characters. Thus, a file named "1ucy_consurf_1111617002_pipe.pdb" (32 characters long) will be saved as "1ucy_consurf_1111617002_pipe.pd" (31 characters long) with the final "b" truncated. If you rename the file, adding the final "b", attempting to load it into Empty PE with the Browse button will produce a strangely corrupted filename. The solution is to use shorter filenames (less than or equal to 31 characters).

  5. More troubleshooting hints may be added here in a future revision. Send your suggestions to emartz@microbio.umass.edu.