I think that oc is near to act as a stable production compiler.
All messages I recently received from list, however, are relative to:
numbers > 18 digits. (do they exist?);
paths whith spaces inside (parhaps 11th commandment?);
rarely used IMB cobol parameters;
and so on.
On the contrary, same essentials features: alternate keys, split-keys and
keyboard function-keys have not been implemented yet.
I belive that cobol still exists because IT HANDLES DATA like no other
That is why I belive that keys handling is THE PRIORITY. User interfaces
comes soon after.
Everyone has their own priorities I guess. Open COBOL currently compiles 100% of our 30+Million lines of COBOL application, we have no
screen sections, all the UI code is written in C with Curses which is called from COBOL.
The only problem keeping us from using OC in anger is that OC does not yet support nested sub programs with global and local variables. To compile any one of the 4000+ application programs that we have, it needs a manual exercise to strip the nested sub programs out, change to pass the globals through the linkage section and compile each sub program up as a seperate cobol program. It all works if this is done but it's a huge effort.
So from my point of view, nested sub programs are more important than file IO or screen handling.