Overview
Request 1189585 new
- Update to 3.2
* improved dialect handling including changed defaults to better match the selected dialect (see NEWS if you compile with any `-std` to know more about the implications), a complete new dialect GCOS and support for more COBOL statements, intrinsic functions and syntax from both "old" and new dialects
* highly improved run-times for several statements, along with less memory usage, especially if runtime checks are enabled
* fileio changes to support LINE-SEQUENTIAL per COBOL2023 and runtime options to change the way files are handled, see NEWS and runtime.cfg
* improvements for source-level debugging via GDB and coredump support
* output of context for diagnostics
* improvements for reproducible builds
- Created by frantisek.simorda
- In state new
-
Package maintainers:
MihailJP and
jayssj11
Request History
frantisek.simorda created request
- Update to 3.2
* improved dialect handling including changed defaults to better match the selected dialect (see NEWS if you compile with any `-std` to know more about the implications), a complete new dialect GCOS and support for more COBOL statements, intrinsic functions and syntax from both "old" and new dialects
* highly improved run-times for several statements, along with less memory usage, especially if runtime checks are enabled
* fileio changes to support LINE-SEQUENTIAL per COBOL2023 and runtime options to change the way files are handled, see NEWS and runtime.cfg
* improvements for source-level debugging via GDB and coredump support
* output of context for diagnostics
* improvements for reproducible builds
I tried that, please note that the testsuite fails on Aarch64:
Failed tests: GnuCOBOL 3.2 test suite: GnuCOBOL Tests test groups:
NUM: FILE-NAME:LINE TEST-GROUP-NAME KEYWORDS
843: run_misc.at:14097 runtime check: write to internal storage (1) runmisc call bounds exceptions
I have reported the issue to bug-gnucobol@gnu.org
OK, should I revoke this request?
I have filed https://sourceforge.net/p/gnucobol/bugs/981/ about the failure
Apparently, the failure is expected (ugh). We either need to ignore the testresults on some platforms or rather remove the testcase from the testsuite. But that means regenerating autotools stuff which I have not got around to yet (and may not for a while yet).
@MihailJP, @jayssj11: review reminder