System Settings menu goes off screen when vertical resolution is limited
The system settings menu would appear but be “off screen” at the bottom on screens with limited resolution. Now, if the browser reports a low resolution at login, the system settings menu is moved up to the top to enable the maximum number of rows to be displayed. Note - this will not fix the problem on all devices / screens - it is best to work as an administrator with a screen of a “reasonable” resolution. Please also note: using the “zoom” factor of the browser helps you access menu items on a small screen.
In the chat connector, network errors could cause messages to be lost. This problem has been fixed, when the network connection is re-established within a reasonable time limit.
Phone Numbers could disappear after being set in the MiniClient when Free-Seating is configured.
Strange behaviour could ensue when trying to enter a phone number in the MiniClient if Free-Seating was configured: After selecting the country code, it could disappear. Similar bahavior would happen when editing Users > Communication and trying to add a new telephone number: after selecting country code, the code would move to the phoneNumber field. This problem has been fixed.
Dialler REST API - allow update of contact with appointment date
The dialler REST API now allows updating or importing contacts with an appointment date and optionally a follow up user. This is achieved by passing the following two parameters with the DiallerContact JSON data structure:
The Users UID is matched in the system. If no match is found (or this field is not specified), then the appointment will be made for no specific user. Otherwise the appointment is made for the specific (found) user.
The validation in the IVR object Connect Variable has been corrected. The object would work correctly if variables with $ were used, however in the GUI the object would be coloured red with a validation error.
The hazelcast cluster error handling has been enhanced when some nodes crashed. This situation could occur when a network outage (for longer than a few seconds) or a VM crash on several VMs occurred. The hazelcast instance on the newly re-started server could cyclically log the following message: Exception: java.util.concurrent.ExecutionException: java.lang.IllegalArgumentException: sequence:2 is too large. The current tailSequence is:-1. If this occurred, the only remedy was to restart the server again. The error handling has been improved, to retry the cluster initialisation in this case.
Silent Monitoring could fail in some cases with multiple telephony servers
Silent Monitoring could fail in some cases with multiple telephony servers. Now, the destination host is read from the actual call from the database, which fixes this problem.
The automatic session reload behaviour has been changed in the IVR editor. This could cause problems editing the application on some (usually slow) internet connections.
The KIP AutologoffAll has been corrected in ACD Agent Report 2. It now counts only logouts of the agent according to the “total” counter (LogoutActionCode = 6). Previously it would also count LogoutActionCodes 7, 8 and 9 too (7 = Auto logout (daemon daily autologout), 8 = Auto logout (busy calls exceeded), 9 = Auto logout (no answer calls exceeded)).
PBX Connector Names - new option for TAPI Connector to not overwrite names
The TAPI PBX connector previously overwrote all reported PBX extension names. However, some TAPI connectors report non useful names to the system (e.g. Extension 300 or SIP 300 for example). A new feature has been added which prevents the names being overwritten. Extensions are created once when first seen, but the name will no longer be modified.
This allows for changes to be made using the portal, in particular for a more useful name to be entered.
This feature is activated in the TAPI connector configuration, by setting the following flag in the configuration file jtelTAPIMonitorService.exe.config:
VoiceMail - play by phone could fail in installations with multiple 8-Servers
In installations with multiple telephone servers, the feature 'voice mail - play by phone' could fail, if an additional server with 0 telephone lines was being used purely as an application (daemon) server. This problem has been fixed.
Supervisor 500 Exception with ACD Group Numbers and Dialler Campaigns
If the ACD groups associated with dialler sub-campaigns were configured with a group number, and more than one different group was associated with the campaign by having two master campaigns with different ACD groups configured, then the supervisor would fail with an http 500 error. This problem has been fixed.
All Prompt uploads of prompt via the "new" popup window fail.
All uploads of prompt via the "new" prompt popup was failed due to an error in Javascript (wrong id). This was detected when trying to create a new prompt in ACD Group rules. A side window failed. A further problem was when changing a language, the Input file name will would disappear. This was fixed for all prompt uploads via popup (Reseller Master Data > Call Recording...)the popup.
Dialler Campaign does not disapear from supervisor view (wallboard and dialler) after deleting.
Data Cache mechanism for dialler compaigns was simply storing all records forever, even deleted ones. So in supervisor view (wallboard and dialler), deleted dialler compaigns was not disapearing. This is fixed and another cache mechanism is used for dialler compaigns so it updates even the deleted ones.
Portal Optimisations - Caching of Texts and User Settings
The portal web application now caches texts, parameters and user settings. This speeds up the login process in particular (up to 5 times faster), but all areas of the portal profit from this speed optimisation.
Outbound Statistics Report - Total Calls Completed
Total Calls Completed was giving wrong statistics with a dialler, and the same statistics as Total Calls in normal outbound. This was fixed and now: - Outbound Completed Call means that the call was connected and CONNRES equal to 1 (Successful outbound call to destination).
The queue distribution algorithm has been optimised. This particularly affects queues with > 120 waiting callers, the speed of the calculation can be reduced by a factor of 2 to 10 in some cases.
