This table lists the most recent releases of Texpress 6.2 for each platform.
Select the Number in the Version column for Release Notes.
6.2.012 - 2 July 2002
- Fixed problem with SAVESET and COPYALL commands that could leave the last record of a matching set loaded in the Insertion Form, while not viewing the last record. If an edit is performed the wrong record is modified.
- Fixed problem where records inserted into the temporary file and then retrieved would be deleted when saved. The problem only occurs if freelist=no is turned on.
- The "doexec" function used to start another process was modified so that child process groups are set correctly.
- A problem (mainly on Linux) where a script running a Texpress program may fail without any output has been corrected. The terminal characteristics where not being saved correctly.
6.2.011 - 28 June 2001
- Fixed problem where latitude and longitude query values with a lower case
directional value (s, n, or e, w respectively) returned incorrect query results.
6.2.010 - 14 March 2001
- Fixed a bug in texql/texserver where a multi field "Key Type" item used as an indexed linked key would not correctly find matching key values on a database search.
- Removed some redundant preliminary double nested grid code from the Attribute display area of texdesign.
6.2.009 - 6 March 2001
- Automatic form order command added to texdesign. Will automatically order items in a left to right, top to bottom order based on the form position of the item prompt or first field (if there is not a prompt).
6.2.008 - 5 February 2001
- Fixed a bug in server-based global replace where replacement of an empty string in a column using regular expressions could cause incorrect text to be inserted into the column.
- Fixed a bug in server-based reporting where suppressing the output for a column (using a leading '-' character) could cause the server to fail.
- The texreindexprep program was changed. A -w option was added to request that the process waits for the texreindex program to finish before exiting.
- Updated terminal description for NetTerm terminal emulators.
- The texserver.sh shell script has been removed from the release.
6.2.007 - 5 September 2000
- The auditing process was modified as follows:
- Addition of seconds to the texaudit output
- New option "auditfields" which allows one or more fields to be audited instead of the key item, for appropriate operations
- New option "auditlevel" which can be used to reduce audit overhead:
0 - all operations (default)
1 - major operations (auditon, login, etc.)
2 - data changes
3 - key changes
4 - queries
5 - display
6 - miscellaneous database maintenance
7 - trace
- Added a timefmt() function to the interpreter. The function works in a similar way to the datefmt() function.
- Fixed bug in texexport where temporary files were not removed correctly.
- Fixed bug where a reindex after adding a key to a database failed because the temporary key file did not exist.
- Added texviewer to permissions list in texbldperms script.
6.2.006 - 27 July 2000
- Fixed bug where incorrect status could be set on a record during an edit.
- Fixed bug in processing of characters from standard input after an alarm. Previously programs would exit with "Unexpected end of input". This was particularly a problem in texmonitor.
- Improved granularity of auditing. Each operation is now audited to the nearest second, rather than to the nearest minute.
- Changed timeout when attempting to lock a resource from 30 seconds to 2 minutes.
- Added more information to the "Dump record information" command in texforms.
- Added a -v flag to texkfdump to print more information about each entry in the key file. Also added a -k command to allow the state of individual keys in the key file to be checked.
- Fixed bug in option parsing code where texenv -v would dump core if run without a database name.
6.2.005 - 21 July 2000
- Added support for multi-language exporting. The data for each column is now split by language delimiter and the appropriate language value included in the XML file. If the column does not have enough language entries to match the requested language ID, the first language is used.
- Fixed a bug in exporting where XML data for groups was occasionally not dumped.
- Adjusted terminal code so that LINES and COLS environment settings take precedence over the automatically determined values.
6.2.004 - 19 Jun 2000
- Added a new interpreter function ymdhmscmp() to compare two time stamps. The function takes a character flag which specifies the "granularity" of the comparison and two time stamps, each in the form of six arguments - y, m, d, H, M, S - making 13 arguments in all.
The "flag" is a single character representing the "granularity" you want in the result: acceptable values are:
y - granularity is in Years
M - granularity is in Months
d - granularity is in Days
h - granularity is in Hours
m - granularity is in Minutes
s - granularity is in Seconds
Note the uppercase 'M' for months and lowercase 'm' for minutes. This is consistent with the date and time formatting conventions.
The other twelve are the two date/times. All values except the seconds are integers. The seconds are floats.
The function compares the two times and returns a string with the differences between the times. The result string starts with a leading '+' or '-' which indicates the direction of the difference and then a colon-separated list of numbers which indicate the difference in
everything from "granularity" down. For example:
|Arguments ||Results ||Explanation
|"y", 1999, 1, 1, 0, 0, 0, 2000, 1, 1, 0, 0, 0
||+ 1 year, no months, no days etc.
|"M", 1999, 1, 1, 0, 0, 0, 2000, 1, 1, 0, 0, 0
||+ 12 months, no days etc.
|"d", 1999, 1, 1, 0, 0, 0, 2000, 1, 1, 0, 0, 0
||+ 365 days etc.
|"M", 2000, 1, 1, 0, 0, 0, 2000, 3, 1, 0, 0, 0
||+ 2 months
|"d", 2000, 1, 1, 0, 0, 0, 2000, 3, 1, 0, 0, 0
||+ 60 days (it knows about leap years etc.)