About Multiple Applications

  • Article

About Multiple Applications

Speech Application Deployment Service (SADS) is the application deployment service for Microsoft Speech Server (MSS). MSS uses SADS to manage multiple speech applications in a single environment.

In a multiple application environment, Speech Engine Services (SES) uses a dynamic manifest file (SESManifest.aspx), and Telephony Application Services (TAS) uses a dynamic Start page (TASStartPage.aspx). SADS generates both SESManifest.aspx and TASStartPage.aspx by using the individual application's manifest file. Each application manifest file reads information from an XML file, Applist.xml. The Applist.xml file is generated by the MSS Microsoft Management Console (MMC) snap-in. For more information, see Configuring SES and TAS for Telephony Applications.

If you are running MSS in a server farm, configure each SES computer to point to SESManifest.aspx, and configure each TAS computer to point to TASStartPage.aspx. For more information, see Configuring Speech Engine Services, and Configuring Telephony Application Services.

When the value of the PreloadedResourceManifest property of the SES class changes, or when SES starts, SES reads the value of PreloadedResourceManifest and retrieves the file specified in the property setting. Then SES parses the content as XML. SES checks the XML to make sure it is syntactically correct, but it does not validate the XML file against a schema. SES ignores those sections of the XML file that it does not understand.

SES can point to only one manifest file, so it is important that you keep the list of preloaded resources contained in the manifest file in sync with the application. If you change the setting of the PreloadedResourceManifest property, either by using the MSS MMC snap-in or by using Windows Management Instrumentation (WMI) scripting, SES will check the validity of the currently loaded manifest file. When SES identifies a change to the manifest file, SES loads the file and begins to implement the changes by loading and unloading resources as needed.

Use Microsoft Speech Application SDK to create the manifest as you develop the application. Then deploy the manifest to the Web server along with the other application files in the application's virtual folder.