################################################################################ # IBM EDMSuite OnDemand # # AIX, HP-UX, Solaris, Windows NT (server/client) # # OS/400 (clients only) # # README # # Release 2.2.1.4 # # December 10, 1999 # # # # 5622-662, 5765-D60, 5765-E20 5639-E12 (C) COPYRIGHT IBM CORPORATION 1999 # # All Rights Reserved # # Licensed Materials - Property of IBM # # US Government Users Restricted Rights - Use, duplication or # # disclosure restricted by GSA ADP Schedule Contract with IBM Corp. # ################################################################################ The information about OnDemand contained in this file was not available at the time the OnDemand publications were printed. Please make note of these items, which supersede the information contained in the OnDemand publications. ================================================================================ 1. Contents 2. Server Change History 2.1 AIX Server 2.1.2.2.1.0 Release/PTF (2.2.1.0) 2.1.2.2.1.1 PTF (2.2.1.1) 2.1.2.2.1.2 PTF (2.2.1.2) 2.1.2.2.1.3 PTF (2.2.1.3) 2.1.2.2.1.4 PTF (2.2.1.4) 2.2 HP-UX Server 2.2.2.2.1.0 Release/PTF (2.2.1.0) 2.2.2.2.1.1 PTF (2.2.1.1) 2.2.2.2.1.2 PTF (2.2.1.2) 2.2.2.2.1.3 PTF (2.2.1.3) 2.2.2.2.1.4 PTF (2.2.1.4) 2.3 Solaris Server 2.3.2.2.1.0 Release (2.2.1.0) 2.3.2.2.1.1 PTF (2.2.1.1) 2.3.2.2.1.2 PTF (2.2.1.2) 2.3.2.2.1.3 PTF (2.2.1.3) 2.3.2.2.1.4 PTF (2.2.1.4) 2.4 Windows NT Server 2.4.2.2.1.0 Release/PTF (2.2.1.0) 2.4.2.2.1.1 PTF (2.2.1.1) 2.4.2.2.1.2 PTF (2.2.1.2) 2.4.2.2.1.3 PTF (2.2.1.3) 2.4.2.2.1.4 PTF (2.2.1.4) 3. Client Change History 3.1 OS/2 Client 3.1.2.2.1.0 Release/PTF (2.2.1.0) 3.1.2.2.1.1 PTF (2.2.1.1) 3.1.2.2.1.2 PTF (2.2.1.2) 3.1.2.2.1.3 PTF (2.2.1.3) 3.2 Windows 3.1 Client 3.2.2.2.1.0 Release/PTF (2.2.1.0) 3.2.2.2.1.1 PTF (2.2.1.1) 3.2.2.2.1.2 PTF (2.2.1.2) 3.2.2.2.1.3 PTF (2.2.1.3) 3.3 Windows 32-bit Client 3.3.2.2.1.0 Release/PTF (2.2.1.0) 3.3.2.2.1.1 PTF (2.2.1.1) 3.3.2.2.1.2 PTF (2.2.1.2) 3.3.2.2.1.3 PTF (2.2.1.3) 3.3.2.2.1.4 PTF (2.2.1.4) 3.4 Windows 32-bit Administrator 3.4.2.2.1.0 Release/PTF (2.2.1.0) 3.4.2.2.1.1 PTF (2.2.1.1) 3.4.2.2.1.2 PTF (2.2.1.2) 3.4.2.2.1.3 PTF (2.2.1.3) 3.4.2.2.1.4 PTF (2.2.1.4) 4. Upgrade Instructions for 2.2.1.0 5. Client/Server Backward Compatibility 6. Documentation Corrections ------------------end of Contents----------------------------------------------- 2. Server Change History -------------------------------------------------------------------------------- 2.1 AIX Server 2.1.2.2.1.0 Release/PTF (2.2.1.0) - Pre-reqs - AIX 4.2.1 or later - DB2 UDB 5.2 or later - ADSM 3.1.0.6 or later 2.1.2.2.1.1 PTF (2.2.1.1) - Added additional generic index file format support (See below for additional details) - Added option to arsview (-x) to not display horizontal/vertical lines - Added option to arsview (-y) to use character space reset algorithm - Added support to arsmaint to ignore the lost+found directory - Fixed a bug which caused arsload to not return a 0 return code - Fixed a bug which caused PDF indexing to fail because of "word search or extraction error" 2.1.2.2.1.2 PTF (2.2.1.2) - Added support to arsdoc which allows the specific of an application group for all functions (-G parameter) - Fixed a bug which caused arsdb to not allow multiple backup devices - Fixed a bug which caused arsdate to fail when displaying values 2.1.2.2.1.3 PTF (2.2.1.3) - Added support to arsload to delete the .ARD file when using the generic index file format - Fixed a bug which caused arsacif to not collect inline resources - Fixed a performance problem with arsacif that was introduced in version 2.2.0.13 - Fixed a bug which caused arsadm user to not delete a user when the user is specified in lower case - Fixed a bug which caused the loading of the generic index file format to not accept CRLF - Fixed a bug which caused arsjesd to not invoke the -x when multiple -d directories are installed 2.1.2.2.1.4 PTF (2.2.1.4) - Added support to arsacif for allowing up to 2 characters when FILEFORMAT=STREAM --------------end of AIX Server------------------------------------------------- 2.2 HP-UX Server 2.2.2.2.1.0 Release/PTF (2.2.1.0) - Pre-reqs - DB2 UDB 5.2 or later - ADSM 3.1.0.6 or later 2.2.2.2.1.1 PTF (2.2.1.1) - Added additional generic index file format support (See below for additional details) - Added option to arsview (-x) to not display horizontal/vertical lines - Added option to arsview (-y) to use character space reset algorithm - Added support to arsmaint to ignore the lost+found directory - Fixed a bug which caused arsload to not return a 0 return code - Fixed a bug which caused PDF indexing to fail because of "word search or extraction error" 2.2.2.2.1.2 PTF (2.2.1.2) - Added support to arsdoc which allows the specific of an application group for all functions (-G parameter) - Fixed a bug which caused arsdb to not allow multiple backup devices - Fixed a bug which caused arsdate to fail when displaying values - Fixed a bug which caused server based text search to fail 2.2.2.2.1.3 PTF (2.2.1.3) - Added support to arsload to delete the .ARD file when using the generic index file format - Fixed a bug which caused arsacif to not collect inline resources - Fixed a bug which caused arsadm user to not delete a user when the user is specified in lower case - Fixed a bug which caused the loading of the generic index file format to not accept CRLF - Fixed a bug which caused arsjesd to not invoke the -x when multiple -d directories are installed 2.2.2.2.1.4 PTF (2.2.1.4) - Added support to arsacif for allowing up to 2 characters when FILEFORMAT=STREAM --------------end of HP-UX Server----------------------------------------------- 2.3 Solaris Server 2.3.2.2.1.0 Release (2.2.1.0) - Pre-reqs - DB2 UDB 5.2 or later - ADSM 3.1.0.6 or later 2.3.2.2.1.1 PTF (2.2.1.1) - Added additional generic index file format support (See below for additional details) - Added option to arsview (-x) to not display horizontal/vertical lines - Added option to arsview (-y) to use character space reset algorithm - Added support to arsmaint to ignore the lost+found directory - Fixed a bug which caused arsload to not return a 0 return code - Fixed a bug which caused PDF indexing to fail because of "word search or extraction error" 2.3.2.2.1.2 PTF (2.2.1.2) - Added support to arsdoc which allows the specific of an application group for all functions (-G parameter) - Fixed a bug which caused arsdb to not allow multiple backup devices - Fixed a bug which caused arsdate to fail when displaying values 2.3.2.2.1.3 PTF (2.2.1.3) - Added support to arsload to delete the .ARD file when using the generic index file format - Fixed a bug which caused arsacif to not collect inline resources - Fixed a bug which caused arsadm user to not delete a user when the user is specified in lower case - Fixed a bug which caused the loading of the generic index file format to not accept CRLF - Fixed a bug which caused arsjesd to not invoke the -x when multiple -d directories are installed 2.3.2.2.1.4 PTF (2.2.1.4) - Added support to arsacif for allowing up to 2 characters when FILEFORMAT=STREAM - Fixed a bug which caused arsacif to drop core --------------end of Solaris Server-------------------------------------------- 2.4 Windows NT Server 2.4.2.2.1.0 Release/PTF (2.2.1.0) - Pre-reqs - DB2 UDB 5.2 or later - ADSM 3.1.0.6 or later - Due to possible installation problems when installing 2.2.1.0 on top of 2.2.0.5 (the initial release of OD/NT), it is recommended that you install 2.2.0.15 on top of 2.2.0.5, and then install 2.2.1.0 2.4.2.2.1.1 PTF (2.2.1.1) - Added additional generic index file format support (See below for additional details) - Added option to arsview (-x) to not display horizontal/vertical lines - Added option to arsview (-y) to use character space reset algorithm - Fixed a bug which caused arsload to not return a 0 return code - Fixed a bug which caused PDF indexing to fail because of "word search or extraction error" - Fixed a bug which caused the configurator to use wrong name for the load data service 2.4.2.2.1.2 PTF (2.2.1.2) - Added support to arsdoc which allows the specific of an application group for all functions (-G parameter) - Fixed a bug which caused arsview to output multiple CR per LF - Fixed a bug which caused arsdb to not allow multiple backup devices - Fixed a bug which caused arsdate to fail when displaying values 2.4.2.2.1.3 PTF (2.2.1.3) - Added support to arsload to delete the .ARD file when using the generic index file format - Added support for Microsoft SQL Server 7.0 - Fixed a bug which caused arsacif to not collect inline resources - Fixed a bug which caused arsacif to GPF during indexing - Fixed a bug which caused arssockd to hang in a high usage environment - Fixed a bug which caused arsadm user to not delete a user when the user is specified in lower case - Fixed a bug which caused the loading of the generic index file format to not accept CRLF - Fixed a bug which caused arsjesd to not invoke the -x when multiple -d directories are installed - Fixed a bug which caused arsview to incorrectly output extra CRLF characters 2.4.2.2.1.4 PTF (2.2.1.4) - Added support to arsacif for allowing up to 2 characters when FILEFORMAT=STREAM - Fixed a bug which caused arsacif to GPF --------------end of Windows NT Server------------------------------------------ ------------------end of Server Change History---------------------------------- 3. Client Change History 3.1 OS/2 Client 3.1.2.2.1.0 Release/PTF (2.2.1.0) 3.1.2.2.1.1 PTF (2.2.1.1) - Added support for viewing line data utilizing iconv - Update AFP Viewer 2.22 - Fixed a bug which caused client to crash when viewing documents in auto cascade/tile mode 3.1.2.2.1.2 PTF (2.2.1.2) - Update AFP Viewer 2.23 3.1.2.2.1.3 PTF (2.2.1.3) - Update AFP Viewer 2.24 --------------end of OS/2 Client------------------------------------------------ 3.2 Windows 3.1 Client 3.2.2.2.1.0 Release/PTF (2.2.1.0) 3.2.2.2.1.1 PTF (2.2.1.1) - Update AFP Viewer 2.48 3.2.2.2.1.2 PTF (2.2.1.2) - Update AFP Viewer 2.49 3.2.2.2.1.3 PTF (2.2.1.3) - Update AFP Viewer 2.50 --------------end of Windows 3.1 Client----------------------------------------- 3.3 Windows 32-bit Client 3.3.2.2.1.0 Release/PTF (2.2.1.0) 3.3.2.2.1.1 PTF (2.2.1.1) - Added support for PDF thumbnails - Update AFP Viewer 1.48 - Fixed a bug which caused the client to not use unified login after a timeout - Fixed a bug which caused the client hang when the view title string was greater than 200 characters - Fixed a bug which caused the client to not properly display PDF documents using the rotation definition 3.3.2.2.1.2 PTF (2.2.1.2) - Update AFP Viewer 1.49 - Updated all national languages to match function available in US/English client - Fixed a bug which caused the client to not properly store print properties under Win95 - Fixed a bug which caused the client to hang if the last character in a line data document is 0x8B - Fixed a bug which caused images to not rotate properly - Fixed a bug which caused copy document pages to file for line data to omit form feed characters 3.3.2.2.1.3 PTF (2.2.1.3) - Added support for including the Monarch integration with the client installation. (Prior to 2.2.1.3., Monarch integration was a separate installation.) - Update AFP Viewer 1.50 - Fixed a bug which caused the client to not work with UNC names - Fixed a bug which caused the client to not retain orientation during local printing 3.3.2.2.1.4 PTF (2.2.1.4) - Added support for the Print and Go To commands as shortcuts - Added support for highlighting the most recent line in Find All - Fixed a bug which caused the client to fail when doing server-based text search against a local server - Fixed a bug which caused the client to print the number of pages instead of the number of copies specified when printing PDF from the hit list --------------end of Windows 32-bit Client-------------------------------------- 3.4 Windows 32-bit Administrator 3.4.2.2.1.0 Release/PTF (2.2.1.0) 3.4.2.2.1.1 PTF (2.2.1.1) - Fixed a bug which caused the graphical indexer to hang when trigger not found on next page - Fixed a bug which caused the graphical indexer to not allow more than 3 characters for the code page 3.4.2.2.1.2 PTF (2.2.1.2) - Added Japanese translation 3.4.2.2.1.3 PTF (2.2.1.3) - Fixed a bug which caused the administrator to GPF when updating a folder 3.4.2.2.1.4 PTF (2.2.1.4) - Added support for allowing up to 2 characters for ACIF when FILEFORMAT=STREAM --------------end of Windows 32-bit Administrator------------------------------- ------------------end of Client Change History---------------------------------- 4. Upgrade Instructions for 2.2.1.0 -------------------------------------------------------------------------------- OnDemand Version 2.2.1.0 adds the following enhancements: NOTE: Please see Section 5, Client/Server Backward Compatibility - Support for loading files up to 4 GB in size. (Previous limitation was 2 GB. PDF Indexer continues to support files up to 2 GB.) - Support for 64-bit integers as a data type. To support the 64-bit integer data type, all clients and service offerings must be at version 2.2.1.0 or later. At this time, the following clients and service offerings do not support the 64-bit integer data type: - All clients older than version 2.2.1.0 - Windows 3.1 Client - CICS Client (Service Offering) - Ad-Hoc Data Distribution (16-bit and 32-bit) (Service Offering) - Production Data Distribution (Service Offering) - Distributed User/Group Management - Distributed Server Print Management - New commands (AIX only): - arsdb replaces ars_db - arsload replaces ars_load - arsdocget replaced by arsdoc - arsquery replaced by arsdoc - arsuser replaced by arsadm - arsgroup replaced by arsadm - New packaging of client and server installation. OS/2, Win16, Win32 clients are provided as separate zip files and are no longer shipped as server-based install images. The enhancements require changes to the OnDemand system tables. Therefore, the following instructions must be completed if you are upgrading to version 2.2.1.0 from a previous version of OnDemand. If you have already installed version 2.2.1.0, you do not have to complete these steps. If you are installing OnDemand for the first time (version 2.2.1.0), you do not need to complete these steps. We recommend that you log all of the following steps. For example, in UNIX, use the script command. - (WinNT) Due to possible installation problems when installing 2.2.1.0 on top of 2.2.0.5 (the initial release of OD/NT), it is recommended that you install 2.2.0.15 on top of 2.2.0.5, and then install 2.2.1.0 1) Make sure that your server has the appropriate pre-reqs. 2) Back up your system (ADSM, DB2, OnDemand). For UNIX installs, the ars.cfg, ars.ini, ars.cache, and ars.dbfs files will be replaced. Make sure that you back up these files so that you can merge the information in them into the new versions of the files installed by 2.2.1.0 (see step 4). NOTE: Many customers have their own way of starting and running OnDemand. If you have custom commands or scripts, make sure that you back them up before you continue. 3) Install OnDemand Version 2.2.1.0 NOTE: You should assume that the entire server directory will be removed. Make sure that you back up any custom files you stored in the server directory before you begin. NOTE: The client subdirectory under the server install directory will be removed and is no longer needed. The clients are now shipped as separate zip files. 4) Reconfigure the OnDemand configuration files (UNIX only): - ars.cfg - ars.ini - ars.cache - ars.dbfs - arsload.cfg (userid/password for arsload command) 5) Reconfigure the OnDemand script/batch files: - ars_adsm (UNIX Only) - arslog - arsprt NOTE: You should also reconfigure any custom commands that may be affected by the OnDemand command name changes. For example, you may need to modify the /etc/inittab to use arsdb instead of ars_db and arsload instead of ars_load. 6) Upgrade the OnDemand system tables: - Create a temporary working directory NOTE: Make sure that you specify a directory with plenty of free space (20 - 100 MB). The upgrade does not export the data tables, only the OnDemand system tables. Space needed can vary depending on the number of Folders/ApplGroups defined to the system and the number of annotations stored in the database. mkdir /tmp/OnDemand.2210 or mkdir c:\tmp\OnDemand.2210 - Change to the directory cd /tmp/OnDemand.2210 or cd c:\tmp\OnDemand.2210 - Export the existing tables NOTE: Because new tables are added in 2.2.1.0, you must specify the -f flag to ignore errors. You should expect errors because the arsprtusr and arsusrgrpid tables are new to 2.2.1.0. arsdb -flxv - Drop the old tables NOTE: Because new tables are added in 2.2.1.0, you must specify the -f flag to ignore errors. You should expect errors because the arsprtusr and arsusrgrpid tables are new to 2.2.1.0. arsdb -dfv - Create the new tables/indexes arsdb -rtv - Import the old table information into the new tables NOTE: Because new tables are added in 2.2.1.0, you must specify the -f flag to ignore errors. You should expect errors because the arsprtusr and arsusrgrpid files do not exist prior to 2.2.1.0. arsdb -filv - Run maintenance on the new tables arsdb -mv - You are done! NOTE: All server-based printer definitions will not be accessible to end-users. You must run the administrative client to grant authority to these printers. ------------------end of Upgrade Instructions for 2.2.1.0----------------------- 5. Client/Server Backward Compatibility -------------------------------------------------------------------------------- OnDemand has always provided backward compatibility between clients and servers. However, please note the following: - Library and object servers must always be at the same version - The administrative client should always be at the same version as the server - Versions of the administrative client older than 2.2.1.0 will not work against a version 2.2.1.0 server - Version 2.2.1.0 or later of the administrative client will not work against a server older than version 2.2.1.0 - When new function has been added to a version of the client and server, you should not use the new function unless both the client and server are at the same version. For example, if you plan to use the new 64-bit integer data type, both the clients and server must be at version 2.2.1.0 or later. ------------------end of Client/Server Backward compatibility------------------- 6. Documentation Corrections -------------------------------------------------------------------------------- The latest versions of the US English OnDemand publications are available in PDF format on the IBM web site: http://www.software.ibm.com/data/ondemand/library.html 6.1 IBM EDMSuite OnDemand: Indexing Reference, S544-5489-05 6.1.1 Page 62. CCTYPE parameter. It is very important to correctly identify the type of carriage control characters in the input file. ACIF may process an input file even though the CCTYPE parameter incorrectly identifies the type of carriage control characters in the input file. However, the output file may be unusable. If you are not certain which type of carriage control characters are in the input file, contact your system administrator. 6.1.2 Page 67. DEFAULT keyword. The default value can be specified either as character data or hexadecimal data. If the input data is anything other than ASCII, the default value must be specified in hexadecimal. If the input data is ASCII, use the following format: DEFAULT='value' 6.1.3 Page 71 - 72. FILEFORMAT parameter. The NEWLINE keyword now supports a two-character line delimiter, which is common in data from DOS, OS/2, and Windows operating systems. The following shows an example of a FILEFORMAT parameter used to process input data containing two-character line delimiters: FILEFORMAT=STREAM,(NEWLINE=X'0D0A') The NEWLINE keyword must be specified if the line delimiter is two characters. If the NEWLINE keyword is not specified, the default line delimiter is X'0A' for ASCII data and X'25' for EBCDIC data. 6.1.4 Page 103 - 104, USERMASK parameter. The USERMASK must be specified in hexadecimal when CPGID=500 and the input data is EBCDIC. For example: USERMASK=1,'*',X'C181C282C383' FIELD3=*,*,15,(OFFSET=(10:24),MASK='*@@@@@@@@@@@@@@',ORDER=BYROW) 6.1.5 Page 203, "Format of the generic indexer file." OnDemand now supports two formats of the generic indexer file. In addition to the format described in the book, the new generic indexer file format can contain the following lines: COMMENT: A comment line. Comment lines can be placed anywhere in the file. CODEPAGE: Identifies the codepage of the index data. You can one and only one codepage. The CODEPAGE: line must appear before the groups. After the CODEPAGE: line, define the groups. Each group represents one document to be indexed. A group consists of the following: GROUP_FIELD_NAME: Identifies a database field name. You can specify 1 - 32 fields. You must specify one for each application group field. GROUP_FIELD_VALUE: The value of the database field. Specify one VALUE for each FIELD. Note: Specify one or more pairs of GROUP_FIELD_NAME and GROUP_FIELD_VALUE lines -- one pair for each index field. After you've defined all of the GROUP_FIELD_NAME and GROUP_FIELD_VALUE pairs, complete the group by coding the following three lines: GROUP_OFFSET: The byte location of the data to be indexed. Specify 0 (zero) for the first byte in the file. GROUP_LENGTH: The length in bytes of the document. Specify 0 (zero) to process the entire input file. GROUP_FILENAME: The full path name of the input file on the server. If one input file contains multiple documents to be to be indexed, specify the input file name on the GROUP_FILENAME: line in the first group and leave the GROUP_FILENAME: lines in all other groups blank. The following example generic indexer file shows how to index three documents with two fields. There is a different input file for each document: COMMENT: COMMENT: Generic Indexer Example 1 COMMENT: Different input file for each document COMMENT: COMMENT: Specify code page of the index data CODEPAGE:819 COMMENT: Document #1 COMMENT: Index field #1 GROUP_FIELD_NAME:rdate GROUP_FIELD_VALUE:07/13/99 COMMENT: Index field #2 GROUP_FIELD_NAME:studentID GROUP_FIELD_VALUE:0012345678 COMMENT: document data starts at beginning of file GROUP_OFFSET:0 COMMENT: document data goes to end of file GROUP_LENGTH:0 GROUP_FILENAME:/arstmp/statement7.out COMMENT: Document #2 COMMENT: Index field #1 GROUP_FIELD_NAME:rdate GROUP_FIELD_VALUE:08/13/99 COMMENT: Index field #2 GROUP_FIELD_NAME:studentID GROUP_FIELD_VALUE:0012345678 GROUP_OFFSET:0 GROUP_LENGTH:0 GROUP_FILENAME:/arstmp/statement8.out COMMENT: Document #3 COMMENT: Index field #1 GROUP_FIELD_NAME:rdate GROUP_FIELD_VALUE:09/13/99 COMMENT: Index field #2 GROUP_FIELD_NAME:studentID GROUP_FIELD_VALUE:0012345678 GROUP_OFFSET:0 GROUP_LENGTH:0 GROUP_FILENAME:/arstmp/statement9.out COMMENT: COMMENT: End Generic Indexer Example 1 The following example generic indexer file shows how to index three documents with two fields. One input file contains all of the documents: COMMENT: COMMENT: Generic Indexer Example 2 COMMENT: One input file contains all documents COMMENT: COMMENT: Specify code page of the index data CODEPAGE:819 COMMENT: Document #1 GROUP_FIELD_NAME:rdate GROUP_FIELD_VALUE:07/13/99 GROUP_FIELD_NAME:studentID GROUP_FIELD_VALUE:0012345678 COMMENT: first document starts at beginning of file (byte 0) GROUP_OFFSET:0 COMMENT: document length 8124 bytes GROUP_LENGTH:8124 GROUP_FILENAME:/arstmp/accounting.student information.loan.out COMMENT: Document #2 GROUP_FIELD_NAME:rdate GROUP_FIELD_VALUE:08/13/99 GROUP_FIELD_NAME:studentID GROUP_FIELD_VALUE:0012345678 COMMENT: second document starts at byte 8124 GROUP_OFFSET:8124 COMMENT: document length 8124 bytes GROUP_LENGTH:8124 COMMENT: use prior GROUP_FILENAME: GROUP_FILENAME: COMMENT: Document #3 GROUP_FIELD_NAME:rdate GROUP_FIELD_VALUE:09/13/99 GROUP_FIELD_NAME:studentID GROUP_FIELD_VALUE:0012345678 COMMENT: third document starts at byte 16248 GROUP_OFFSET:16248 COMMENT: document length 8124 bytes GROUP_LENGTH:8124 COMMENT: use prior GROUP_FILENAME: GROUP_FILENAME: COMMENT: COMMENT: End Generic Indexer Example 2 6.1.6 Page 226. FIELD parameter. Change the second sentence in the first paragraph. You can define up to 32 (thirty two) index fields. 6.1.7 Page 231 - 232. INDEXSTARTBY parameter. The INDEXSTARTBY parameter determines the page number by which the first group (document within the input file) must be found. The first group is identified when *all* of the triggers and fields are found. For example: TRIGGER1 = ul(4.72,1.28), lr(5.36,1.45), *, 'ACCOUNT' TRIGGER2 = ul(6.11,1.43), lr(6.79,1.59),1,'SUMMARY' INDEX1 = 'Account',FIELD1,FIELD2, FIELD1 = ul(6.11,1.29), lr(6.63,1.45),2 FIELD2 = ul(6.69,1.29), lr(7.04,1.45),2 INDEX2 = 'Total',FIELD3 FIELD3 = ul(6.11,1.43), lr(6.79,1.59),2 INDEXSTARTBY = 3 The word ACCOUNT must be found on a page in the location described by TRIGGER1. The word SUMMARY must be found on the page following the page where ACCOUNT was located, in the location described by TRIGGER2. In addition, there must be one or more words found for fields FIELD1, FIELD2, and FIELD3 in the locations described by FIELD1, FIELD2, and FIELD3 which are located on a page that is two pages after the page where TRIGGER1 was found. In the example, the first group in the file must start on either page 1, page 2, or page 3. If TRIGGER1 is found on page 1, TRIGGER2 must be found on page 2, and FIELD1, FIELD2, FIELD3, must be satisfied on page 3. 6.1.8 Page 235. TEMPDIR parameter. On Windows NT servers, the default drive for the temporary directory is C. If you do not specify the TEMPDIR parameter, the default drive and directory is C:\TEMP. 6.1.9 Page 248. -t temp_dir. On Windows NT servers, the default drive for temporary files is C. If you do not specify the -t parameter, the default drive and directory for temporary files is C:\TMP. ----------- end of corrections to S544-5489-05 --------------------------------- 6.2 IBM EDMSuite OnDemand: Installation Guide for UNIX Servers, G544-5598-01 6.2.1 Page 28. Configuration Notes. Add the following item to the list: 5. For Solaris servers, create a link to the DB2 shared libraries. Enter the following commands: cd /lib ln -s /opt/IBMdb2/V5.0/lib/libdb2.so.1 . 6.2.2 Page 31. Installing ADSM. OnDemand requires version 3.1.0.6 or later of the ADSM API. The ADSM API is automatically installed as part of the basic, default ADSM server installation (along with the ADSM administrative client, which is also required by OnDemand). If you install a version of the ADSM server earlier than 3.1.0.6, you must obtain version 3.1.0.6 of the ADSM API and install it before you can use ADSM with OnDemand. 6.2.3 Page 36, item 13, page 37, item 11, and page 39, item7. The correct URL to obtain the latest service update for OnDemand Version 2.2 is: ftp://service.software.ibm.com/software/ondemand/fixes/v221 Follow the links to the latest PTF for your server. 6.2.4 Page 39, item 3. Enter the following command: /usr/sbin/pkgadd -d /cdrom/server/sun/OD2210 6.2.5 Page 114. Preparing storage pool volumes. The example LABEL LIBVOL command is incorrect: o To label storage volumes and check them into a library that is filled with blank storage volumes, use the following command: label libvol archlib0 search=yes overwrite=yes checkin=scratch labelsource=prompt o To label storage volumes and check them into a library by manually inserting each new storage volume when prompted, use the following command: label libvol archlib0 search=bulk overwrite=yes checkin=scratch labelsource=prompt With both forms of the command, you are prompted to label each storage volume. 6.2.6 Page 119. In the example define schedule command, change 'startime' to 'starttime'. 6.2.7 Page 123 - 124. Replace the topic titled "Setting the client node password in DB2" with the following: Setting the client node password in DB2 Before DB2 can use ADSM to maintain log files and backup image files, you must set the client node password in DB2 on the library server. You established the client node password when you registered the client node in ADSM (see Registering the DB2 client node on page 123). Use the dsmapipw command to set the client node password in DB2. The dsmapipw command is installed in the INSTHOME/sqllib/adsm directory of the instance owner. (The OnDemand instance owner is archive, by default.) To set the client node password: 1. On the library server, log in as the root user. 2. Open a terminal window. 3. Set the DSM_CONFIG and DSMI_CONFIG environment variables. These variables contain the full path name of the options file ADSM uses to identify the server that maintains the DB2 files. (You created the options file in Defining client user options on page 120.) For example: export DSM_CONFIG=/usr/lpp/adsm/bin/dsm.opt.db2 export DSMI_CONFIG=/usr/lpp/adsm/bin/dsm.opt.db2 4. Run the dsmapipw command. For example: /home/archive/sqllib/adsm/dsmapipw 5. The dsmapipw command prompts you for the following information: old password, which is the current password for the client node. This is the password you specified when you registered the client node in ADSM (see Registering the DB2 client node on page 123). new password, which is the password that will be stored in DB2. You will be prompted twice for the password. NOTE: Unless you have a good reason for *not* doing so, specify the old password when prompted. 6. Close the terminal window. 7. If DB2 is running, you should stop and restart DB2. Important: If you change the client node password in ADSM, remember to run the dsmapipw command to set the new password in DB2. ---------- end of corrections to G544-5598-01 ---------------------------------- 6.3 IBM EDMSuite OnDemand: Installation Guide for Windows NT Servers, G544-5526-05 6.3.1 Page 7. Server software requirements. Microsoft SQL Server 6.5 is not supported. If you plan to use SQL Server as the database manager, you must install SQL Server 7. 6.3.2 Page 25 - 26. Installing SQL Server. If you plan to use SQL Server as the database manager, you must install SQL Server 7 (SQL Server 6.5 is not supported). See the SQL Server 7 installation information to install SQL Server on the library server. 6.3.3 Page 27. Installing ADSM. OnDemand requires version 3.1.0.6 or later of the ADSM API. The ADSM API is automatically installed as part of the typical ADSM server installation (along with the ADSM administrative client, which is also required by OnDemand). If you install a version of the ADSM server earlier than 3.1.0.6, you must obtain version 3.1.0.6 of the ADSM API and install it before you can use ADSM with OnDemand. 6.3.4 Page 34. Verify Server and Client Options, second paragraph. Add NOPREEMPT to the server options file. This parameter makes sure that mount requests cannot be preempted by other commands (except for the backup database command). 6.3.5 Page 39. Item 3. In the device class statement, change format=2600 to format=2600MB. 6.3.6 Page 43. The VOLRANGE parameter works only with libraries that support a bar code reader (IBM 3995 libraries do not). To label storage volumes and check them into a library by manually inserting each new storage volume when prompted, use the following command: label libvol archlib0 search=bulk overwrite=yes checkin=scratch labelsource=prompt The LABEL LIBVOL command prompts you for the label of each storage volume. 6.3.7 Page 48. In the example define schedule command, change 'startime' to 'starttime'. 6.3.8 Page 49. Change the first sentence from "You must set the COMPRESS parameter to yes." to "You must set the COMPRESSION parameter to yes." 6.3.9 Page 49. Change the "COMPRESS yes" parameter and value pair in the sample dsm.opt.db2 file to COMPRESSION yes. 6.3.10 Page 51 - 52. Replace the topic titled "Setting the client node password in DB2" with the following: Setting the client node password in DB2 Before DB2 can use ADSM to maintain log files and backup image files, you must set the client node password in DB2 on the library server. You established the client node password when you registered the client node in ADSM (see Registering the DB2 client node on page 51). Use the dsmapipw command to set the client node password in DB2. The dsmapipw command is installed in the \sqllib\adsm directory. To set the client node password: 1. On the library server, log on with the OnDemand system administrator account. 2. Start a command prompt. 3. Set the DSM_CONFIG and DSMI_CONFIG environment variables. These variables contain the full path name of the options file ADSM uses to identify the server that maintains the DB2 files. (You created the options file in Configuring client options on page 48.) For example: set DSM_CONFIG=d:\adsmv3\baclient\dsm.opt.db2 set DSMI_CONFIG=d:\adsmv3\baclient\dsm.opt.db2 4. Run the dsmapipw command. For example: \sqllib\adsm\dsmapipw 5. The dsmapipw command prompts you for the following information: old password, which is the current password for the client node. This is the password you specified when you registered the client node in ADSM (see Registering the DB2 client node on page 51). new password, which is the password that will be stored in DB2. You will be prompted twice for the password. NOTE: Unless you have a good reason for *not* doing so, specify the old password when prompted. 6. Exit the command prompt. 7. If DB2 is running, you should stop and restart DB2. Important: If you change the client node password in ADSM, remember to run the dsmapipw command to set the new password in DB2. 6.3.11 Page 56 - 57. Important things to remember. Make the following changes to the list of items in the third paragraph: 1. Change "Size of the database (SQL Server)" to "Size of Database" 2. Change "Location of primary log files (DB2)" to "Location of log files" 3. Change "Number of primary log files (DB2)" to "Number of log files" 4. Change "First file system named (DBFS and Cache)" to "First cache file system named" 6.3.12 Page 60. Add the following item to SQL Server: o Number of Log Files 6.3.13 Page 63. Configuring scheduled tasks, OnDemand database backup. If you specify a variable block size you must also specify a buffer size in the backup command that is less than or equal to the maximum block size limit for the backup device being used. For maximum performance, you should set the buffer size to the maximum block size for the backup device. See the DB2 documentation for details. See also DB2 technical note JMER-3PFMYE and PMR 00711,370. 6.3.14 Page 111. PDF indexer. The Acrobat software is distributed on the OnDemand product CD-ROM. Use the setup.exe program to install the software. The setup program is located in the \adobe\win\acrobat3 directory. When prompted for the product serial number, use the serial number provided by IBM (listed in the README.TXT file found in the root directory of the product CD-ROM). You may install the Acrobat Distiller software on any OnDemand library or object server for which you hold a license. 6.3.15 Page 112. Item 5 (Configuring Acrobat distilling software). To configure the Distiller to delete input files from the IN directory rather than moving them to the OUT directory after successfully distilling a file, open the Watched Folders dialog box and select "After Distilling, the PostScript file should be: Deleted". ---------- end of corrections to G544-5526-05 ---------------------------------- 6.4 IBM EDMSuite OnDemand: User's Guide, SC26-9810-00 6.4.1 Page 3. Change the second paragraph, that begins with "To view PDF documents..." to "For Windows clients, to view PDF documents..." 6.4.2 Page 4. Installing Adobe software. OS/2 software. Change the first sentence in the second paragraph to "You need the Acrobat Reader for OS/2 Warp to view the OnDemand publications provided in PDF format on the CD-ROM." The OS/2 client does not support viewing PDF documents retrieved from OnDemand. 6.4.3 Change items 2 and 3 at the top of page 12 to: 2. Select Run from the File menu. 3. In the Command Line field, type x:\client\windows\win16\setup (where x is the letter of your CD-ROM drive or the network drive). Then press Enter. 6.4.4 Page 12, To install on a user's PC. Change items 2 and 3 to: 2. Select Run from the File menu. 3. In the Command Line field, type x:\client\windows\win16\setup (where x is the letter of your CD-ROM drive or the network drive). Then press Enter. 6.4.5 Chapter 4, Installing for Windows 95, Windows 98, and Windows NT, pages 16 - 17 describes how to install the client from Autorun. However, the client CD-ROM does not contain the Autorun files. Therefore, you cannot install the client from Autorun. 6.4.6 The following information can be used to obtain and install the latest version of OnDemand client software from IBM service on the World Wide Web. Note: these instructions assume you use Netscape Navigator to browse the Web. The steps for other browsers may be different. 1. Point your browser to: ftp://service.software.ibm.com/software/ondemand/fixes/v221 Follow the links to the latest PTF for your client. 2. Click the ZIP file for the client you want to install or upgrade. For example: odwin32.zip 3. Select Save it to disk. 4. Click OK. 5. Choose the Save as location. 6. Click Save. After the download is complete, you can expand the ZIP file or store the ZIP file in a shared location, depending on how you plan to distribute the client software: o A user can expand the ZIP file to a temporary directory and then run the Setup program to install or update the client on a PC. After expanding the ZIP file, follow the instructions in the OnDemand User's Guide, SC26-9810, to install the client. See "Install on a user's PC" for details. After installing the client, the user can delete the temporary directory and the ZIP file. o An administrator can store the ZIP file in a shared location so that other users can access it. o An administrator can expand the ZIP file to a shared location so that other users can run the Setup program from the shared location. o If your organization shares a copy of the OnDemand client software from a network server, an administrator can expand the ZIP file and run the Setup program to install or update the client on the server. o If your organization distributes user-defined files with the OnDemand client, an administrator must expand the ZIP file to a shared location and then copy the user-defined files to that location. Users can then run the Setup program from the shared location to install or update the client software and the user-defined files on their PCs. ---------- end of corrections to SC26-9810-00 ---------------------------------- 6.5 IBM EDMSuite OnDemand: Administrator's Reference, S544-5293-08 6.5.1 Page 80. Searching the library. The VOLRANGE parameter works only with libraries that support a bar code reader (IBM 3995 libraries do not). To label all of the volumes in a library that has been filled with blank storage volumes, use the following command: label libvol odlib0 search=yes overwrite=yes checkin=scratch labelsource=prompt The LABEL LIBVOL command prompts you for the label of each storage volume. 6.5.2 Page 110. Administering SQL Server. SQL Server 6.5 is no longer supported. Use SQL Server 7 instead. 6.5.3 Page 139 - 152. arsadm command reference. The following changes have been made to the arsadm command: 1) The User types "Create Users" and "Create Groups" are now supported in the arsadm user function. Two new values for the -t flag have been added 'c' and 'e'. In addition the functional of the flag has changed. The following is a list of -t flag values: a = User - Create Appl Grps b = User - Create Appl Grps, Folders c = User - Create Users d = Appl Grp/Folder Administrator e = User - Create Groups f = User - Create Folders g = User Administrator s = System Administrator u = User The value 'b' represents a combination of the values 'a' and 'f'. This value was already available and will continue to be supported. Rather than creating single letter values for every combination of the four user types, the values 'a', 'c', 'e', 'f' can now be specified in any combination or order. The remaining values of 'b', 'd', 'g', 's', 'u' can only be specified by themselves. For example here are some valid combinations for the -t flag: -t acef "User - Create Users, Groups, Appl Groups, Folders" -t ce "User - Create Users, Groups" -t af "User - Create Appl Groups, Folders" -t cef "User - Create Users, Groups, Folders" Some examples of invalid combinations are: -t cs "Can't specify User - Create Users w/System Admin" -t eg "Can't specify User - Create Groups w/User Admin" 2) The Inactivity Time Out value can now be specified for users when adding or updating a user. The new flag is -s. Valid values for the -s flag are: -1 = Never Time Out 0 = Use System Inactivity Time Out Value >0 = Number of Minutes ( 1 - 1440 minutes ) 3) Inactive userids can now be deleted with the arsadm user function, using two new flags: -e expire_date and -f date_format. The two new flags are valid only when deleting inactive userids. You can use one of following three formats to delete inactive userids: A) Delete one userid: arsadm user -u userid -p passwd -h host -a d -i USER1 -v For example: arsadm user -u wagner -p "" -h everest -a d -i DEBBIE -v B) Delete all userids inactive since a specified date: arsadm user -u userid -p passwd -h host -a d -e expire_date -v For example: arsadm user -u wagner -p "" -h everest -a d -e 11/12/1998 -v C) Delete all userids inactive since a specified date using a date format: arsadm user -u userid -p passwd -h host -a d -e expire_date -f date_format -v For example: arsadm user -u wagner -p "" -h everest -a d -e "November 12, 1998" -f "%B %e, %Y" -v When deleting userids with the date method, arsadm uses the following process to determine if a userid should be deleted: 1) A query is performed for all of the userids that can be maintained by the logged on user. 2) Each userid is examined. If the user's password has not been updated between the specified date and the current system date, the userid is deleted. For example, assume that passwords expire every 180 days. User A logs on to the server on 3/15/1999; User B logs on to the server on 10/20/1998. Both users changed their passwords when they logged on, and have not logged on to the server since that time. You run the arsadm user command on 11/12/1999 to delete inactive userids. You specify an expire date of 11/12/1998. When the arsadm command runs, User A is not deleted because the password was updated within the date range (11/12/1998 - 11/12/1999). However, User B is deleted because the password has not been updated since 10/20/1998. The following example shows how to configure the system to automatically remove inactive userids on a regular schedule. For example, you could use the process to remove users that have not logged on for at least 180 days. 1) On the server, set Minimum Password Age to Expires In 180 Days. 2) Create a shell script that uses arsdate to determine the date for 180 days prior to today's date: arsdate -g -d -180 -f "%m/%d/%Y" 3) Using the date returned by arsdate, run the arsadm user command to delete inactive userids: arsadm user -u wagner -p "" -h everest -a d -e expire_date -v 4) Run the shell script every day from inittab. When the shell script runs, userids that have not logged on to the server in at least 180 days will be deleted. Notes 1) The -f flag is optional. If you do not specify the -f flag, the default date format is "%m/%d/%Y". 2) Userids should *NOT* be deleted with the date method if you do not use password expiration on the server (that is, Minimum Password Age is set to Password Never Expires). 3) When you add a userid to the server, the user must change their password the first time that they log on. If you add a new userid and you run the arsadm user delete function with the -e flag before the user logs on for the first time, the userid will be deleted. 6.5.4 Page 169. arsdate command reference. Purpose The arsdate command can be used to display the database value for a given date and time string or display the date and time string for a given database value. Values whose data types are Date, Time, Date/Time, or Date/Time (TZ) are represented in an internal form that is transparent to the casual user of OnDemand. Casual users enter these values the same way they appear in a report. However, to search the database with an SQL string, a user must enter the internal form of the value. The arsdate command shows you the internal value of a date or time string. Parameters -a Use to display database values and date strings for Date fields. For example, to display the database value for the date 9/1/99, enter: arsdate -a 9/1/99 The arsdate command returns: 9/1/99 -> 10836 To display the date string for the database value 10836 in the default format, enter: arsdate -a 10836 The arsdate command returns: 10836 -> 9/1/99 -i Use to display database values and date and time strings for Date/Time fields. The time part of a Date/Time field is not adjusted for the local time zone. You typically use the -i parameter to find out one of two things: o Given a date and time printed on a report, what value did OnDemand store in a Date/Time database field? You can use the result to search a Date/Time field with an SQL string. o Given a value stored in a Date/Time database field, what would be the date and time printed on a report? -t Use to display database values and time strings for Time fields. The time is not adjusted for the local time zone. For example, to display the database value for the time 04:00:00, enter: arsdate -t 04:00:00 The arsdate command returns: 04:00:00 -> 4800 To display the time string for the database value 4800, enter: arsdate -t 4800 The arsdate command returns: 4800 -> 04:00:00 -z Use to display database values and date and time strings for Date/Time (TZ) fields. The time part of a Date/Time (TZ) field is adjusted for the local time zone. If you run the arsdate command with the -z parameter on systems in different time zones and you specify the same date and time value, the result will be different. For example, suppose you need to determine the value stored in a Date/Time (TZ) field for "09/01/99 04:00:00". The command: arsdate -z "09/01/99 04:00:00" on a system in the Eastern time zone will return: 09/01/99 04:00:00 -> 936187200 If you run the same command on a system in the Mountain time zone, the result will be: 09/01/99 04:00:00 -> 936180000 A typical use of the -z parameter is to determine a database value with which to search the OnDemand system log. You can use the result to search the Date/Time (TZ) field of the system log with an SQL string. For example, suppose a user in New York logs on to a server in Denver. To retrieve the log on message with an SQL string, you must specify the date and time part of the query using the local time of the client that is running the query. If the user logged on to the server at 4 a.m. Eastern time, a query run in Denver must specify 2 a.m. to retrieve the message. Examples 1. The following example shows how to determine the database value for the specified date and time string. The data type of the database field is Date/Time. arsdate -i "09/01/99 04:00:00" 09/01/99 04:00:00 -> 936158400 In this example you could use the database value to search a Date/Time field in the database with an SQL string. For example: arsdoc get -i "WHERE somedate=936158400" ... 2. The following example shows how to determine the date and time string for the specified database value. The data type of the database field is Date/Time. The result is shown using the default display format. arsdate -i 936158400 936158400 -> 09/01/99 04:00:00 3. The following example shows how to determine the database value for the specified date and time string. The data type of the database field is Date/Time (TZ). The arsdate command adjusts the time part of the result for the local time zone. arsdate -z "09/01/99 04:00:00" 09/01/99 04:00:00 -> 936180000 If you were to run the same command on a system in the Eastern time zone, the result would be: arsdate -z "09/01/99 04:00:00" 09/01/99 04:00:00 -> 936187200 Notes 1. When displaying the date string for a given internal value, by default, the arsdate command displays the date string using the mm/dd/yy format. If you want the arsdate command to display the date string using a different format, you must specify the date format with the -f parameter. For example: arsdate 10907 10907 -> 11/11/99 arsdate -f "%m/%d/%Y" 10907 10907 -> 11/11/1999 2. When displaying the internal value for a given date string, by default, the arsdate command expects you to specify the date string using the mm/dd/yy format. If you want to specify the date string using a different format, you must specify the date format with the -f parameter. For example: arsdate 11/11/99 11/11/99 -> 10907 arsdate 11/11/1999 11/11/1999 -> -1 (Error) arsdate -f "%m/%d/%Y" 11/11/1999 11/11/1999 -> 10907 6.5.5 Page 181. arsdoc command reference. All of the arsdoc command functions now support the -G parameter to specify the name of the application group. o ADD. When the database query is run to retrieve the document that contains the data used in the add, the search is limited to the specified application group, even if the folder named with the -f parameter contains more than one application group. This ensures that only documents in the specified application group can be used for the add. o UPDATE. When the database query is run to determine the document(s) to update, the search is limited to the specified application group, even if the folder named with the -f parameter contains more than one application group. This guarantees that only documents in the specified application group can be updated. o DELETE. The -G parameter has been added as an optional parameter. If specified, the database query that is run to determine the document(s) to delete is limited to the specified application group. The addition of the -G parameter allows you to delete documents from a specific application group in folders that contain more than one application group. If you do not specify the -G parameter, the query runs against all of the application groups in the folder. o GET. The -G parameter has been added as an optional parameter. If specified, the database query that is run to retrieve documents is limited to the specified application group, even if the folder named with the -f parameter contains more than one application group. If you do not specify the -G parameter, the query runs against all of the application groups in the folder. o QUERY. The -G parameter has been added as an optional parameter. If specified, the database query is run against the specified application group, even if the folder named with the -f parameter contains more than one application group. If you do not specify the -G parameter, the query runs against all of the application groups in the folder. o PRINT. The -G parameter has been added as an optional parameter. If specified, the database query that is run to determine the document(s) to print is limited to the specified application group, even if the folder named with the -f parameter contains more than one application group. If you do not specify the -G parameter, the query runs against all of the application groups in the folder. NOTES: 1. For the ADD and UPDATE command functions, you can use either the -g or -G parameter to specify the name of the application group. 2. You can use the -G parameter with the -i parameter to query folders that contain more than one application group. For example, a folder contains three application groups. You want to query only one of the application groups. Use the -G parameter to specify the application group that you want to query. Use the -i parameter to specify the application group's database field. You can also use the -G and -i parameters when the application groups have different database field names. The following example shows how to search a folder and three application groups that have different database field names: arsdoc get -f "Student Information" -G loans -i "where number LIKE '123456' and loan_date = 10593" arsdoc get -f "Student Information" -G grades -i "where number LIKE '123456' and grade_date = 10593" arsdoc get -f "Student Information" -G transcripts -i "where number LIKE '123456' and trans_date = 10593" 3. You can use the -G parameter with the -q parameter to query folders that contain more than one application group. When you specify the -G parameter and you specify a public named query with the -q parameter, the arsdoc command queries the application group named with the -G parameter instead of the application group specified in the named query. (If you do not specify the -G parameter, the query runs against the application group specified in the named query. If the named query does not identify an application group, the query runs against all of the application groups in the folder.) 6.5.6 Page 223. arsview command reference. The following changes have been made to the arsview command: 1. The arsview command now supports an orientation of 180 degrees. The -o parameter now accepts values of 0, 90, 180, and 270. 2. The arsview command now accepts the -x parameter. If you specify -x, the output will not contain horizontal or vertical lines that were present in the AFP document. 3. The arsview command now accepts the -y parameter. Specify the -y parameter to use the character space reset algorithm to place text in the output. ----------- end of corrections to S544-5293-08 --------------------------------- 6.6 IBM EDMSuite OnDemand: Introduction and Planning Guide, G544-5281-04 6.6.1 Page 25. Windows NT server requirements. Microsoft SQL Server 6.5 is no longer supported. Microsoft SQL Server 7 is now supported. 6.6.2 Page 26. Optical and tape storage. OnDemand requires version 3.1.0.6 or later of the ADSM API. The ADSM API is automatically installed as part of the basic, default ADSM server installation (along with the ADSM administrative client, which is also required by OnDemand). If you install a version of the ADSM server earlier than 3.1.0.6, you must obtain version 3.1.0.6 of the ADSM API and install it before you can use ADSM with OnDemand. 6.6.3 Page 58. OnDemand supports the following database and folder field types not listed in Table 3. FIELD TYPE FIELD SIZE DESCRIPTION ---------- ---------- ----------- Big Int Eight (8) bytes Contains whole numbers between -922337036854775807 and 922337036854775807. Big integer fields hold a 64-bit integer representation of a number or character string in the form of an integer constant. Annotation 1-254 Use to search annotations to a Text Search document for the specified string. A match occurs and an item is added to the document list if one or more annotations contain the text specified in the search field. You can define one annotation text search field per folder. Annotation N/A Use to search annotations to a Color Search document by specifying a color. A match occurs and an item is added to the document list if the color of the text in one or more annotations is the same as the color specified in the search field. You can define one annotation color search field per folder. 6.6.4 Page 71. Database transaction log storage. Change the fourth sentence in the first paragraph, that begins with "If you are using SQL Server" to: If you are using SQL Server, when a log file fills, SQL Server closes the full file and opens a new log file, provided you have configured the transaction log to use multiple log files. The OnDemand database uses the AUTOGROW feature to reduce the potential of running out of transaction log space. The log files are truncated after a successful backup of the transaction log and can be reused. ---------- end of corrections to G544-5281-04 ---------------------------------- 6.7 IBM EDMSuite OnDemand: Getting Started with the Administrative Client, S544-5463-08 6.7.1 Page 126. Printer Type. o Change the third sentence to: "A server printer type of Printer with Information lets users enter information used by PSF to generate a cover sheet that is printed with the document." o Add the following: "NOTE: To use a server printer type of Printer with Information, an administrator must configure the PSF print queue associated with the printer by activating the Header Page Exit and specifying the name of the Header Page Program. The header page program merges the information the user enters with a user-defined overlay to produce the cover sheet. (IBM provides a sample header page program, source code, and overlay.) See the PSF documentation for details about configuring PSF print queues, creating a header page exit program, and creating overlays. 6.7.2 Page 159. File Extension. The file extension can now be from three to twenty characters in length. ---------- end of corrections to S544-5463-08 ---------------------------------- ----------------- end of Documentation Corrections -----------------------------