Please enable JavaScript to view this site.

A-Shell Development History

883.4.1

(ATE) Various improvements to the FTP routine. The FTPDLX.LOG file is now considerably more detailed than before. You can also copy entire directories in a single command by setting the filename to *. For example,

! copy all files in /vm/miame/dsk0/011222 to c:\temp....

? chr(27);chr(1);chr(127);"0"; &

   "/vm/miame/dsk0/001222;chr(0);"*";chr(0); &

   "c:\temp";chr(0);

 

! copy all files in c:\temp to /vm/miame/dsk0/001222

? chr(27);chr(1);chr(127);"1"; &

   "/vm/miame/dsk0/001222;chr(0);chr(0); &

   "c:\temp\*";chr(0);

 

883.4.2

Various minor fixes and enhancments to the INFLD GUI version:TYPE S (password) was not working properly. Display colors DFCLR and DBCLR were not being selected for display. TYPE |R now disables the normal redisplay of the field on exit (useful when you are doing your own redisplay).

883.4.3

TRACE=GUI (or SET TRACE GUI ON in SET.LIT 1.2(137) or higher) now logs all TAB(-10,x) commands to the ashlog.log file.

883.3.1

(AIX) Fix a corrupted link library with which causing various segementation faults, particularly in processing MMAPLIST and printing.

883.3.2

Passing -1 for the background color argument in TAB(-10,20) and related control-creation operations now works just as it does for the foreground color, i.e. the -1 gets converted to the current background color. Previously -1 for background color was being intepreted as the "default" background color for that type of control.

883.3.3

Fix a problem with the ESC key exit from the GUI version of INFLD acting as 2 escapes.

883.3.4

Clicking in the middle of an INFLD (GUI version) field now clears the destructive carriage return flag (just as if you had just editing keystrokes to accomplish the same thing.)

883.2.1

(ATE) FTP routine now creates a log file, FTPDLX.LOG, in the dsk0\001004 subdirectory under the ATE installation directory, which might be useful in debugging transfer problems.

883.2.2

(ATE) Selecting Disconnect from the file menu now works in situations where it didn't work before (like when you were within a local GUI control.)

883.2.3

(ATE) Fix various bugs in the remote INFLD GUI control, including the Control-C problem mentioned previously, plus a problem with ESC acting strange, and fields being returned with a trailing or leading CR (causing all kinds of havoc in field validation routines.)

(PSHELL) AUXLOC printing will now wait until the job has a foreground context. (Previously such printing would just spill out into the bit bucket.)

883.0.1

(ATE) INFLD on a server with ATE on the client now executes locally on the client when TYPE |G specified. There are some issues yet to be worked out, so don't expect to put this into production. (But do send us bug reports.) Requires edit 883 or higher on both server and client.

Three known problems are:

The cursor will be turned off, so you may need to turn it back on manually when you exit the program.
Control-C generally doesn't work, unless you use TYPE V. (To overcome this, we need to implement a means for ATE and the server to synchronize their SET CTRLC/NOCTRLC settings.
The control is noticeably slower on some platforms to bring up than the standard INFLD. It does not appear to be much of an issue with Linux.

883.0.2

HLPIDX syntax (see item 2 under edit 881 below) has been changed to use chr(126) as the leadin instead of chr(127). Warning: you must correct any existing programs using this technique or they won't work properly under this and later releases. The choice of chr(127) was a bad one because it conflicts with the remote/ATE implementation of INFLD (which works through the TAB(-10,x) interface and thus uses chr(127) as a terminator. Hopefully few programs will be affected.

883.0.3

(ATE) A problem with intermittent function key operation due to timing issues (mainly seen under AIX) has been resolved by forcing all the bytes of a function key sequence to travel in the same packet.

883.0.4

TAB(-10,20) now supports an abbreviated syntax, useful for those who want to convert a lot of PRINT statements to use proportional fonts but who still want to be able compile their programs under AMOS. (The TAB(-10,20) won't work unless a special terminal driver is created for AMOS, but at least it won't interfere with compilation, so as long as you have the appropriate IF THEN logic to avoid executing them, you can have one source for all platforms. The abbreviated syntax is:

TAB(-10,20);STRING;chr(127); or

TAB(-10,20);STRING;",";JUSTIFY'CODE;chr(127);

 

In the first case, the statement acts just like TPRINT, i.e. it displays STRING using the proportional GUI font, left justified, using the current color scheme.

The second version allows you to specify justification codes, either using the MBF'LEFTJUST and MBF'RIGHTJUST codes (from ashell.bsi), or using the following simplified codes:

"TL" - Text, left justified

"TC" - Text, center justified

"TR" - Text, right justified

"DL" - Data (MBF'SUNKEN), left justified

"DC" - Data, center justified

"DR" - Data, right justified

 

Note that if you use the MBF'xxx codes, you can actually specify others besides justification, such as MBF'STATIC and MBF'SUNKEN, but your capabilities are limited by the lack of any other parameters.

883.04a

PCKLST.SBR FLAGS argument now supports a new bit, 1024, which causes it to time out in 200 seconds. (Currently this only works in the non-GUI version of PCKLST but will be added to the GUI version later.)

883.4

(ATE) The installation was not creating a DSK0:[1,7], which was leading to a cryptic error during ATE printing about "ATE#.PQI not found".

883.4

(ATE) The GDI/Passthrough printer setting was not being saved.