Closed freakout42 closed 10 years ago
Please retest with current 3.2.0 and update your report.
[ If an issue is present in 3.2.0, it's best to report it here. If the issue happens only with a fork (like 3.4.0), always use the fork's own Issue page. ]
Retested on Win7 with 3.2.0dev (mingw 4.9.0) and the problem is present (so the problem is not specific to the fork, nor to Linux)
Also confirming that 3.0.0 (mingw 4.5.2) works correctly on the same system.
After reverting to this version of /src/rtl/gtstd/gtstd.c
, the problem is gone: https://github.com/harbour/core/blob/1b4008723b01b67444e9baf50de4c90e6655f941/harbour/src/rtl/gtstd/gtstd.c
It means that this diff causes the regression: https://github.com/harbour/core/commit/93d3a46d843daaf6d08149d83fa5203ff910c484#diff-70 (giant commit, may load slowly)
a small program which just writes to stdout (with -gtstd not screen oriented - console app) produces different output with development version
OS: uname -a: Linux bongo 2.6.18-371.6.1.el5.028stab112.3PAE #1 SMP
this program
with development version output has a double line for the empno variable and a lot of junk spaces - 3.0.0 version is clean
first the garbled development version: harbour-3.4.0dev-2 /opt/harbour/bin/hbmk2 test_dbs_clp_hrb_cmd.prg -gtstd Harbour 3.4.0dev () (2014-07-08 11:21) produces output
in hex
here the clean output from release version: harbour-3.0.0-2 Harbour 3.0.0 (Rev. 16951) produces output
in hex