Re: [Cobolforgcc-devel] Re: gnubol: which list, and cobol.el
Status: Pre-Alpha
Brought to you by:
timjosling
From: Tim J. <te...@me...> - 2000-08-13 20:03:47
|
Abbrevs ******* I didn't realize the abbrevs weren't implemented. Maybe just put a comment in about this. I would like the common statement starting verbs like GO TO (just kidding), MOVE, PERFORM, COMPUTE, ADD/SUBTRACT/MULTIPLY/DIVIDE, IF and the respective END-Xs, maybe also something for data division like $n -> 0n xxx PIC X(1) USAGE DISPLAY, EXIT PROCEDURE, DISPLAY, STOP RUN. Two Projects ************* Yes I do want to maintain COBOL For GCC and GNU COBOL2C as two separate projects. This is dues to the different approaches taken by the two projects: - Use GCC back end vs generating C code. - Extreme programming (testing) used in COBOL4GCC - Stricter approach to standards on COBOL4GCC - Using Source Forge - Source forge offers a really good infrastructure for running an open source project GNU COBOL2C has been inactive for some time. But it is not for me to say that the project is over. It has achieved a lot especially in terms of getting the parsing issues resolved, but nothing has been happening for a while. My hope is that it will work like this. People will join COBOL4GCC and contribute code, tests and documentation. We will have a viable COBOL subset soon, to allow COBOL programmers to contribute to the project. As momentum grows, more and more developers and users join in. Soon we have it all done. We leverage the experiences of the earlier free COBOL projects (cobcy, cobol2c) and share code with Tiny COBOL (this is possible especially in the complex runtime such as sort/merge - I'd love to write that but I need to concentrate in the parsing and GCC interface). In 17 months COBOL4GCC has accumulated 39,000 lines of code (6,000 lines of Ted Seward's code are not checked in yet), far more than any other free COBOL project. I know that LOC is an imperfect measure, but if you use the same language and have similar levels of documentation and use similarly skilled programmers it is a fair comparison. While functionally it is a baby, the 39,000 lines represent all the hard problems solved. So we can move forward at a rapid rate from this sound base. I would urge anyone wishing to contribute to register themselves to www.sourceforge.net and let me know your ID, then I can put you in the project and we can track tasks and bugs etc, and give you CVS write access in due course. Tim Josling Matthew Vanecek wrote: > > Duly noted. abbrevs were never implemented--I guess the original > developer never got to it. I will, soon. The thing to decide is, > what's important enough to include as an abbrev? Consider opinions > solicited... > Are you wanting to maintain gnu-cobol and cobolforgcc as two seperate > projects? How's that gonna work? > Matthew Vanecek |