home *** CD-ROM | disk | FTP | other *** search
-
- PowerBASIC Patch Documentation
- ==============================
-
- 6/15/90 --- PowerBASIC patch3 data file: PB061590.PAT
-
- This is the third in a series of patches which should be applied to all
- copies of PowerBASIC version #2.00b. The three patches are:
-
- PATCH1.ZIP --> PB033090.PAT
- PATCH2.ZIP --> PB050790.PAT
- PATCH3.ZIP --> PB061590.PAT
-
- This patch data file is to be used in conjunction with PBPATCH.EXE, to add
- enhancements, or to correct error conditions, in PowerBASIC. The executable
- program (PBPATCH.EXE) is distributed with the first patch data file (PATCH1),
- so it is not duplicated here. Instructions for usage of PBPATCH.EXE are
- also included in PATCH1.ZIP. If you do not have the first patch data file,
- it is available from Spectra Tech Support or may be downloaded (PATCH1.ZIP)
- from the PowerBASIC forum on CompuServe or GENIE.
-
- Be certain to install all three patches to your working copy of PowerBASIC.
- If there is any question, do it again. You can't install the patches
- too many times.
-
- **************************************************************************
-
- This patch data file is designed to modify only Rev. #2.00b of PowerBASIC,
- both PB.EXE and PBC.EXE.
-
- It provides the following enhancements and/or corrects the following
- error conditions:
-
- 1. In certain cases, the CHAIN statement may execute incorrectly, generate
- a "String/Array Memory Corrupt" error (#242), or even cause a system "hang".
- This is particularly noticeable when the calling program contains numerous
- string variables which were not declared as COMMON. A "work-around" which
- usually allowed correct operation in the past was to specify $LIB ALL ON,
- which forced the entire run-time library to be included. This should no
- longer be necessary.
-
- 2. When numeric error checking is disabled, an error may be introduced
- when the result of a floating point operation is assigned to an integer
- variable.
-
- 3. PowerBASIC may generate incorrect executable code, when the INP()
- function is referenced in certain complex numeric expressions.