This table lists the most recent releases of Texpress 8.1 for each platform.
Select the Number in the Version column for Release Notes.

Top

Release Notes

8.1.020 - 23 April 2009

• Added Solaris Opteron support (32 bit support for amd64 based machines).
• Fixed issue where the second matching record is displayed first when the last record is a false match.

8.1.019 - 10 February 2009

• Fixed problem with re-indexing of accented UTF-8 characters.

8.1.018 - 17 October 2008

• Improved regular expression handling for "top-bit" characters.
• Fixed problem in global replace where Ignore was treated as Cancel and aborted the replace.
• Fixed problem where cancelling a record insertion or edit could result in a duplicate key value being assigned.
• Fixed problem with generation of temporary file names which could result in a failure to create a temporary file.

8.1.017 - 10 June 2008

• Improvements to generate better index configurations for data sets with a large variance in the number of retrieval terms per record.
• Improved searching for the characters œ, Œ, æ, Æ, ß and related two character searching equivalents "ae", "AE", "oe", "OE" and "ss" respectively.
• Improved texserver API call that runs a server program such that passing of an interrupt signal (SIGURG) from the client will translate to a server side program interrupt signal (SIGINT).
• Fixed a texload program unique number auto incrementation error introduced in version 8.1.015.

8.1.016 - 02 April 2008

• Fixed issue where generating a report for a single record on a multilingual system where a specific language is specified (except for the last language) would result in the current record showing only the language selected, rather than all languages, after the report is complete.
• Fixed issue with producing a Summary with one record on a multilingual system where a specific language is selected would result in the record displaying with that language after switching back to view all languages.

8.1.015 - 11 March 2008

• Change to unique key pending/commit functionality Unique values are now pending only if the system has allocated the next key or the user has entered a fully specified value Partial values are only marked as pending when the record is saved.
• Fixed problem where linked data values (via a link key) may not be updated correctly when the record is saved.
• Parsing of multi-part unique keys into multiple columns has changed so that components are now split on field separators only (rather than any punctuation).
• Fixed problem where exporting records with empty link fields would result in missing XML tags for the empty link.
• Fixed memory leak introduced if a table cannot be opened successfully.
• Fixed issue where XML audit records generated incorrect column names for double nested columns made up of fake tables (LatLatitude0_nesttab generated rather than LatLatitude_nesttab).

8.1.014 - 19 December 2007

• Removed © symbol mapping to 'c' character as it was also marked as a punctuation character. The inconsistent mapping caused problems when searching for the symbol.
• XML audit records are no longer generated for updates that do not change any values.
• Fixed problem where duplicate IRNs may occur when an insertion is commenced while on the first row and then cancelled.

8.1.013 - 14 December 2007

• Corrected error which could occur when merging rows on two cursors where the "selection" has to be split to slot in a non-selected row.

8.1.012 - 4 December 2007

• Fixed problem in texlicstatus where  the number of licences in use may not relect the actual usage.
• Added new option xmlauditinfo that allows an <info> tag with the setting value as data to be included in XML audit records.
• Added special parsing code to handle ODBC format dates regardless of dateorder option setting.
• Fixed problem on SFU on Windows 2003 where Device Busy error may occur when perfoming queries.

8.1.011 - 2 November 2007

• Added program name to XML auditing information.
• Changed texcopy so that the user file is copied if the person copying the table is the DBA.
• Modified texcolwrite() so that a NULL format argument implies the default format for unique columns rather than clearing the format.
• Fixed problem where changing the key value while inserting a record to a key that is already assigned and then cancelling the insertion would result in the already assigned key being unassigned.
• Fixed issue where trying to assign an empty key value would result in a previous key value being unassigned. 

8.1.010 - 7 September 2007

• XML auditing format changed from texforms based representation (items, fields) to texql based format (table, tuple, atom).
• Simplified xmlaudit options by removing xmlaudititems, xmlauditmode, xmlauditcomputed, xmlauditquantity, and xmlauditmsgpath.
• Fixed issue with searching for very large numbers (> 2^31) within text fields which may result in incorrect and slow matching.
• Fixed issue where the cancellation of an edit in texforms, texql or texserver would not result in the original data being re-fetched. The old version is now re-loaded correctly.

8.1.009a - 4 January 2008

• Fixed problem where duplicate IRNs may occur when an insertion is commenced while on the first row and then cancelled. The fixed was back ported from 8.1.014 to 8.1.009.

8.1.009 - 12 July 2007

• The error message generated when a lock on a record is denied now displays the name of the user who holds the lock.
• French translations for texserver based errors add to lang/fr/dflt/messages.
• New option xmlauditmsgpath added. The option outputs the complete xml of the new record to the assigned path.
• The interpreter has been changed to force a linked key update if a value assigned to linked key column.
• New option filepartitionmax added to set the maximum number of file partitions allowed. The default value is 10.
• New option lang added that specifies the language to use for messages and bindings. The setting may be changed "on the fly" to allow dynamic loading of error messages.
• NULL values for format based data types (latitude, longitude, date, time) now produce empty strings rather than the formats punctuation (generally a series of spaces).
• A new debug program, texrddump, has been added. The program is used to check the number of records in a segment against the value stored in the distribution file and report any discrepeancies.
• Corrected a potential deadlock issue when both texserver and texload are inserting or updating records at the same time.
• Added an optimisation to improve performance when exact matching records for queries that contain a large number of OR planes.
• Fixed problem where stem and phonetic searches were not indexing terms for inclusion in search.
• Corrected an issue where sorting and summarising of values that contain extended characters did not summarise correctly.
• Fixed sorting where Case Sensitive is selected and extended character folding disabled, that resulted in the sort not being case sensitive.
• Fixed problem with comparison of large numbers (> 2^31 - 1) in text fields.

Top

8.1.008 - 15 Janaury 2007

• Increased maximum length of a field to 131071.

8.1.007 - 29 December 2006

• Corrected rule in query optimiser where column = NULL generated NULL rather than UNKNOWN.
• Corrected unique value assignment problem caused by a validation expression with an insert() statement changing the unique value. In some circumstances this could lead to records with duplicate IRNs.
• Fix to texforms  where in some circumstances following a record edit subsequent inserts could end up with the same unique value.
• Added code to texforms to cleanup pending unique values upon Insert/Edit cancel.

8.1.006 - 13 November 2006

• Added identification value to auditing information. The identifier is a unique value used to link up the display of query results with the query performed. Use texaudit -i to include identification information.
• Fixed a problem with calling TexRowNew() on a cursor that projects columns from a base table. The problem resulted in the incorrect assignment of values to given fields.
• Changed the way the '=' character is interpreted in front of searches where punctuation is involved. If the '=' operator is used Texpress no longer folds the punctuation to its base character. This allows exact matching on punctuation to be performed.
• Changed XML auditing to correct issue with fields that are both user keyed and computed being audited as "computed" rather than not "computed".
• Fixed texenv for XML auditing based options so that it returns the correct settings for the supplied environment.
• Fixed internal buffer overflow when using texforms reporting facility with very long fields joined together.
• Added support for new fifo() interpreter call used to provide fifo interaction between server and external processes.
• Fixed rounding error in latitude/longitude calculations for seconds values.
• Improved efficiency of link table lookups by keeping data and key handles open between lookups.
• Improved texload loading where loading is exclusive to a table and the schema contains unique column(s).
• Added in support for describe grants to texql so that user access can be described.
• Fixed texql.pm so that defunct processes are cleaned up correctly.
• Added support for retrieveal of columns by position within TexAPI. The column name "COLUMN X" where  is the column number may be used.
• Fixed TexExport() so that only one row is printed for columns that have Table As Text enabled.
• Fixed problem with global replaces in server of words, where the second match in a string would result in an incorrect update.

Top 

8.1.005 - 12 May 2006

• Corrected potential unique value assignment issue with simultaneous access by multiple users. A New record operation followed by Cancel record without saving operation could result in a number being skipped.
• Fixed problem with clearing of precision values for latitude/longitude centroid calculations when a high precision is changed to a lower precision.
• Corrected integer/float parsing issue for numbers starting with a leading + or -. Problem manifested as negative number queries that were (incorrectly) unable to use the index and hence resulted in a somewhat slower null query with exact matching.
• Fixed problem in expression interpreter sort call where sorting may not work on Intel based machines that use binaries complied with the GNU gcc compiler.

8.1.004 - 19 April 2006

• Updated column format setting for incrementation fields to ignore case based formats (i.e. uc, lc, etc).
• Fixed problem with unique value checking where an insert on one cursor with an edit on another cursor before saving the insertion may result in duplicate unique values being allowed.
• Fixed problem where clearing a nested table field and saving the record without making any other changes may result in an error indicating a value is not unique, when it is unique.

Top 

8.1.003 - 7 April 2006

• Added -a option to texexport. The option exports the whole table in data file order (provided you can view every record). It is extremely quick.
• Fixed problem introduced in 8.1.002 where texenv -v may cause a segment violation.
• Auto incrementing facility changed so that formats with an optional last incrementing component will not generate the second number unless the separator is supplied.
• Extended TexReplace() API call to provide support for removing blank rows.
• Fixed query optimiser problem for wildcard queries on key items.
• Altered security check on key queries to be performed when records displayed rather than when query is performed.
• Fixed problem introduced in 8.1.02 with TexRefresh() where you may have to refresh a record twice before the new version is retrieved.
• Fixed problem with performing global edits in texforms on unique fields, which may have resulted in duplicate values.

8.1.002 - 16 March 2006

• Fixed problem with date validation where dates that used a string month representation would fail validation for correct dates.
• Improved speed of getfldcont() interpreter call when a bad field name is supplied.
• Corrected percentage generated for TexReplace() call where selected rows were used for update.
• Fixed problem with very slow searching for tuple data where pattern searches are used.
• Fixed searching on multi-part keys where multi-part key is used for a reference link.

8.1.001 - 22 February 2006

• Record level security added to to complete suite of Texpress programs (texforms, texserver, texql, texload, texreindex, texexport). Access control is available to restrict access for viewing, editing and deleting records. A new file in the database directory (security) details on a per user basis access levels available for that user.
• Updated TexRowStatus() call to provide feed back indicating if a record can be edited or deleted.
• Texpress server (texserver, texql) changed to allow false matches in matching set. False matches are removed before being viewed. Now provides same functionality as texforms.
• Fixed rounding error in lat/long code which could result in a minutes or seconds value of 60.
• Improved performance of getfldcont() and related interpreter calls.
• Improved perfomance for TexSummary() call on large tables.
• Changed TexReplace() call so that if the first occurrence flag is turned on, only the first match in a table is updated, rather than the first match in each row.
• The abort function now works correctly for TexReplace() calls.
• Fixed problem where large file transfers may be interrupted at random intervals.
• Improved query performance when a large number of OR conditions are specified.
• Support added for HPUX 9000.
• Support for AIX 5.2 with gcc compiler.

Top