Please enable JavaScript to view this site.

A-Shell Development History

979.9.1

MX_GETVER now returns the individual parts of the A-Shell version.

979.9.2

Further cleanup of the calendar control messaging problems.`

979.9.3

Further cleanup of the up/down arrow handling in multi-line INFLD controls.

979.9.4

Cleanup several problems related to scaling of fonts and controls.

979.8.1

Vista themes are now supported by A-Shell for most control types. Note that the separate manifest file is no longer needed.

979.8.2

Type ||f was added as a code for controlling the INFLD fixed pitch font.

979.8.3

When INFLD is in multi-line mode (|M or ||M) and up/down exits are enabled (3 and 5), and is not using Windows keyboard conventions (i.e. not specifying |K), then the up and down arrows will cause INFLD to exit if used when the cursor is already on the first or last line of the control. Note that the last line is defined as the line containing text.

979.7.1

Implement a new GDI print commnd TEXTRIGHT for right-justified output.

979.7.2

The maximum number of ERSATZ definitions has been increased from 400 to 1000.

979.6.3

XTEXT Fix: When re-entering an existing tree using XTROP_RESELECT (op 4), if the XTF_NOREDRAW flag is NOT set, then XTREE will now respond to the XTR'EXPANDLEVEL parameter. This goes back to the original behavior; it was inadvertently changed a few versions back.

979.6.2

XTEXT: Fix a problem in which reloading an existing tree with XTROP_REPLACE (op 1) would sometimes wipe out the popup-menu and combo box definitions associated with the tree.

979.6.3

When re-entering a non-editable, single-select tree, the selection bar was often reverting back to the first row.

979.5.1

Fix problem with unexpected ANSI ESC sequences causing ATE to lock up. Also, fix similar problem, mostly seen under SCO, where a bad IAC (FF) sequence would cause ATE to start spitting out garbage.

979.5.2

MX_ASHFILE now returns the name of the A-Shell executable.

979.5.3

SUBMIT.LIT 3.1(145) now uses the actual A-Shell executable name rather than defaulting to "ashell" under UNIX.

979.5.4

Adjust ISAM'PUSH command so that it uses the current record number according to the record number variable in the open statement.

979.5.5

The bottom border of groupboxes has been raised up 2 pixels to give vertically adjacent groupboxes just a little bit more space. (This may be rescinded if the effect isn't appreciated.)

979.4.1

Fix TAB(-10,AG_WALLPAPER), which wasn't working properly. The usual symptom was that restoring previously saved wallpaper wasn't working.

979.3.1

Further fix to AUI_CONTROL fix in 979.2, which was causing large-scale unwanted clearing of controls within dialogs.

979.3.2

ATE/UNIX side was inappropriately defaulting the aux data argument to "QQQQQQQQQQ..." which wasn't hurting anything but was causing some confusion in the ATE trace of the debug window.

979.2.1

Rework INFLD fix #2 in 979.1 below, as it was causing GUI fields to expand inappropriately on redisplay.

979.2.2

COMPIL version 1.1.348 solves a problem compiling FOR/NEXT loops in /X mode that are embedded in the midst of MAP statements.

979.2.3

Close a loophole in XTREE that would lead to a protection fault if the SftTree DLL control was not found. Note that in this case, as well as several other system errors, XTREE will set EXITCODE to the MMO_SYS value, which happens to be 207 (this goes back to INMEMO and PCKLST). So your app should detect such an error and do something appropriate.

979.2.4

The CTLOP_CLR flag in AUI_CONTROL was not taking into account child/parent offsets when clearing a rectangular area based on the coordinates (rather than the more common case where a control ID or name is specified in order to clear all of the children of the control). For example, a clear request with the CTEXT="*" and coordinates of 1,1,5,40 would have (improperly) removed controls that were in that coordinate range but which were children of a control which was not in that range.

979.1.1

Fix INFLD problem where the row value would be unexpectedly doubled (related to changes in the V parameter format discussed below). Problem mainly seemed to occur with ATE in conjunction with GUI and use of the original legacy V format (F,6,1). Problem only affected ATE (Windows side only.)

979.1.2

Fix INFLD problem with the final display of expanding fields (like date, $, phone), being chopped off at the original XMAX length. Problem affected both UNIX and Windows.

979.0.1

This version marks the beginning of the cutover from the development version 4.9 to the stable version 5.0. As of today, this should be considered only a "release candidate". After a week or two of field testing and patches, it will become the official stable release and we will begin a new development on 5.1. Those to whom the concept of a stable release is important should take advantage of this period to test this version thoroughly and report any bugs so that they can be resolved when the stable release is rolled out.

979.0.2

DIR.LIT 3.0(139) fixes a bug where SET LONGDIR would only work for the first set of files processed. (For example, DIR A,B would display the A files in LONGDIR format and the B files in the old short format.)

979.0.3

ATSYNC.LIT 1.0(105) fixes a minor annoyance where the command line history buffer would contain garbage after executing ATSYNC (meaning that you'd have to hit Control-R several times to repeat the same ATSYNC command line over again).

979.0.4

Several minor updates to ABOUT.LIT were made in the latest version 1.0(107).