General Technical Discussions for Service

Get Involved. Join the Conversation.

Comments

  • harrison

    Thanks, anurag.

    Is it possible to prevent logging to console?

     

    Cheers, 
    Greg

     

  • Anurag Mittal

    This error comes when we use same username and password for initConnectAPI()  and also for logging to console, as we can not create 2 session with same username & password. 

  • harrison

    Thanks for the response, Seeking Solution.

    Yes, I used a login and password of an account made explicitly for initConnectAPI(), but when used in testing, I get a session conflict error. ( I can't reproduce the error exactly without risking the console crashing for logged in users)

     

    Greg

  • Seeking Solution

    In manual it says its deprecated functionality for getAdminURL on Nov 2012 connectPHP. Did you tried by passing staff login and password details in initConnectAPI function which is required to access Admin URL. 

  • Richard Morgan

    Mark:

    That was absolutely the right answer.

    Defeated by a comment, how embarrassing.  My reading comprehension is at an all-time low as I totally missed how important that comment is.  I even read Ben Johns' excellent tutorial where that is clearly spelled out.

    Thanks so much.

  • Mark Rhoads

    The Version: tag is the version of the Connect Common Object Model to use for this handler -- it should specify the same version as the Connect-for-PHP namespace that you are using.

    Currently, "Version: 1.0" is causing CPM to supply the script a version 1.0 object, but the script is binding to version 1.2 of Connect-for-PHP via "use \RightNow\Connect\v1_2 as RNCPHP".

    The object model is different between the two and is likely causing much of the problems that you are seeing.

    If you want to stick with Version 1.2, then change the Version: tag to be "Version: 1.2".

     

     

  • Anuj Behl

    You can request for a Pilot from Oracle. Go here and check right sidebar for Pilot link.

  • Bishnu Paudel

    Thanks a lot Mark!

    Prefixing class name with \ did the trick. The apply function need to include different classes dynamically based on the values of fields of the triggering row ($obj) .

    Thanks,

     

     

  • Mark Rhoads

    To avoid name collisions between different handlers due to functions, global variables, classes, etc that a CPM object event handler might define, CPM object event handler scripts are put into their own namespace.  In PHP, namespaces are in file scope, so the CPM script is in one namespace and the globals/functions/classes in the included file may be in a different namespace.  If no namespace is defined in the included file, then the globals/classes/functions in the included file will be in the "root" namespace.

    So, prefix the class/function names that are from the included file with the namespace of the included file when used in the CPM script.
    If they are in the "root" namespace, prefix them with \

     

    Also, move the "require_once()" to the top of the file instead of in the apply method ... there's no reason to execute require_once() upon each apply().

  • Davide Antonietti

    Hello Charles,

    I was looking for a similar functionality.

    I thought a system attribute was my right tool; at the moment I didn't succeded in having AccountView Control available in workspace editing.

    My goal is to have this Custom object to link it in additional tabs to workspaces of

    Opportunities, Incidents (your needs),Organization and Contacts.

    After having this object attached to standard transactions I can manage accessibility by building relevant reports.(my needs)

    Can you please share some screenshot about your findings?

    I really appreciate it.

    Thank you.

    regards,

    Davide

  • Mark Rhoads

    Yes.

     

  • Justin Anderson

    Hi Marcin:

    When you launch the Tap application on your phone or ipad, click on launch screen to add an account. There are 2 fields Account Name and Server URL.  

    • For Account Name you can give any reference name you want.  Also note that you can have multiple accounts setup on the device in case you want to connect to multiple instances (such as production vs. development instances).
    • For the Server URL, you need to enter a URL that looks like this:
      https://domain_name_of_your_site_including_dotcom/cgi-bin/name_of_your_site.cfg/php/tap/v1

      For example, if my site is sample.custhelp.com, the url would be:   
      https://sample.custhelp.com/cgi-bin/sample.cfg/php/tap/v1

     

    For further information, here are a few items to look at

    • When you first login to the app, there are some welcome screens that explain the overall layout of the app for first time users.
       
    • See this article on this site for initial setup of the mobile agent app.
      Answer 5493: Enable and Customize Mobile Agent App  https://cx.rightnow.com/app/answers/detail/a_id/5493/kw/mobile%20agent
       
    • End User Guide for Mobile Agent App https://cx.rightnow.com/ci/fattach/get/801333/1382998060/redirect/1/filename/Mobile%20Agent%20End%20User%20Guide%20Aug%202013.pdf

     

     

  • Suresh Thirukoti
    We also got the similar issue ..........only solution is to contact your Oracle POC who will release the lock in their DB...
    
    ~Suresh
    
  • Justin

    First off, thanks again @mark.  I was able to create a 10 or 12 test cases in my CPM, and I was able to get them all working.

    However, once I moved on to my next CPM, I ran into another issue that I am hoping someone can shed some light on.

    I need to see if there is a new attachment in this CPM.  My plan was to get the list of file attachments from both the current and previous object and compare the FileNames and their respective UpdateTime fields.  Unfortunately UpdatedTime is a read-only field on the FileAttachmentCommon object, so I cannot set dummy UpdateTimes in the TestHarness and then execute different test cases.  Again, this would only be an issue in the test harness, but I'm a big fan of thoroughly testing my code...

    Any ideas on how I could get around that?

     

     

  • Mark Rhoads

    One good practice is to get as much testing automated via the test harness as is possible ... it is much easier to test here than to set up and read log files, etc.