I am pretty sure is being truncated because it's got a lot of copy files and I simply don't believe they're all totally fine!
I actually think they are - but to be sure you can increase the amount of maximum error before stopping (the default is 128 which is not reached here) by something like -fmax-errors=9999.
Other than "redefinition of seamless-msg" (section/paragraph name is the same as a variable), which is tracked under [feature-requests:#260] and will possibly stay there for a bit longer - if no one provides a patch. If possible I suggest to rename whatever of those two is less used also for acu - this will provide less confusion and IDE and grep support in any case...
I think most of the other errors should be fixed with the next acu-compat commit, soon. Note: the NOTHING message above should be fixed by using -std=acu-strict or by removing the list from the reserved words -std=acu -fnot-reserved=nothing (the NEWS file has the explanation on the difference of the std's).
Other than that: thank you for the list, that helped to find some issues. I'll work on those before the 3.2 release and will drop a note here allowing you to check more programs :-)
The goal is to at least get the parsing to work (so no errors in -std-srict), but to be honest about the "outcome": COBOL programs that do rely on the ACU graphical extensions and its (pseudo-) thread support won't run correct under GnuCOBOL.
For some environments this means that batch processing will work fine and only some frontent parts need to be adjusted/replaced - for some this will mean they can only parse for possibly additional helpful messages and to possibly prepare migrations but otherwise use only "external" services of libcob like the ODBC/OCI interface.
Note: the windowing support itself is actually very likely to be added, but the graphical controls are not; the threading is also something that is likely to be added, but quite unlikely do be done soon.
I actually think they are - but to be sure you can increase the amount of maximum error before stopping (the default is 128 which is not reached here) by something like
-fmax-errors=9999.Other than "redefinition of seamless-msg" (section/paragraph name is the same as a variable), which is tracked under [feature-requests:#260] and will possibly stay there for a bit longer - if no one provides a patch. If possible I suggest to rename whatever of those two is less used also for acu - this will provide less confusion and IDE and grep support in any case...
I think most of the other errors should be fixed with the next acu-compat commit, soon. Note: the
NOTHINGmessage above should be fixed by using-std=acu-strictor by removing the list from the reserved words-std=acu -fnot-reserved=nothing(the NEWS file has the explanation on the difference of the std's).Other than that: thank you for the list, that helped to find some issues. I'll work on those before the 3.2 release and will drop a note here allowing you to check more programs :-)
The goal is to at least get the parsing to work (so no errors in
-std-srict), but to be honest about the "outcome": COBOL programs that do rely on the ACU graphical extensions and its (pseudo-) thread support won't run correct under GnuCOBOL.For some environments this means that batch processing will work fine and only some frontent parts need to be adjusted/replaced - for some this will mean they can only parse for possibly additional helpful messages and to possibly prepare migrations but otherwise use only "external" services of libcob like the ODBC/OCI interface.
Note: the windowing support itself is actually very likely to be added, but the graphical controls are not; the threading is also something that is likely to be added, but quite unlikely do be done soon.
Related
Wish List: #260
Last edit: Simon Sobisch 2022-03-28