We recently finished adding the ability to define more than one Dispo Call URL in a campaign or in-group. This new feature exposed an issue that had been more rare before, but with this new feature it became more prominent. The issue was that the agent screen AJAX call that sends the Agent's selected disposition and completes the logging of the lead was also the function that sent the Dispo Call URL requests and waited for a response to those requests. For companies that used slow-to-respond web pages for their Dispo Call URLs this resulted in agents sometimes going on pause several seconds after their calls dispositioned or other agent screen and logging issues. The first fix for this was done a week ago when that process was made to be "blocking", meaning that the next steps wouldn't automatically happen until the Disposition functions had finished. That created a problem of agents having to sit and wait for all of the Dispo Call URL requests to finish before they could move on to their next call. I tried several different methods of fixing this, but the solution I settled on was to let the Disposition process create the Dispo Call URLs to send out, but not actually send them, then once the disposition process was complete, a separate non-blocking AJAX request is sent from the agent screen to run in the background sending the URLs(using their url_log_id) as logged in the vicidial_url_log table. This solution allows the agent to move on as soon as the disposition process is finished(not including dispo call url sending). The logging also now stores the timestamp of when the response was received.
The other URL change is related to the AST_send_URL.pl script, which handles No-Agent Call URLs and a few others. This process had not been set up to log failures with any amount of useful logging, and it also did not work with HTTPS web addresses. These two issues are fixed, and now all logging from wget for each request is stored in the vicidial_url_log table along with the response.
While we have tested these changes pretty extensively, there is always the possibility that an issue may come up. If you notice one, please post here or on the Issue Tracker.