gnucobol-users
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [open-cobol-list] Gary Cutler's GNU Cobol 2.0 Programmer's Guide


From: Vincent Coen
Subject: Re: [open-cobol-list] Gary Cutler's GNU Cobol 2.0 Programmer's Guide
Date: Thu, 28 Nov 2013 11:32:01 +0000
User-agent: KMail/4.10.5 (Linux/3.8.13.4-server-1.mga3; KDE/4.10.5; x86_64; ; )

On Thursday 28 Nov 2013 03:27:59 Brian Tiffin wrote:

> A GNU Cobol merge will start soon. I was thinking a 1.1 release to

> GNU could soak before all the new 2.0 features, 2.1 now with RW, were

> unveiled. There won't be much soak time. I'll get on the meta data

> for the GNU Cobol 1.1 soon, (ChangeLogs etc) at the same time as the

> 2.0 packaging. After synching up Simon, and getting the signature

> keys all in place for GNU and get new primary links up. Then back on

> the forge, 2.1 will get merged into trunk as soon as we get the GNU

> Savannah link in place. Umm, I think that's how it should go now. 2

> has too many cool things to hold back. *Simon has been kind enough to

> manage the keys to the source repositories, and I'd really like to

> allow him veto powers on any of these planned manoeuvres.*

 

On the v1.1 branch a change was made I think to fileio.c etc that changed the way sequential (line) processing is done in that writing print records out each line had the equivalent a blank line. This happened for me on all print reports in code that has been around one way or another since 1967 in a wide range of compilers from IBM 1401, 7094, 360, 370, ICL 1501,1900s, 29 & 3900 series systems as well as Mfand compilers starting with CIS Cobol in the late 70's including workbench v3 & v4, the only change made was to alter organisation from sequential to line sequential in the late 70's/early 80's when 'line' was implemented into the standard.

 

Since then printing has always worked correctly on kit from line printers, matrix, laserjets and currently inkjets but OC CE just does not play correctly when writing out records with advancing before/after is used.

 

Please ensure that such changes does NOT go into the v2 releases.

 

It might help to work out why, if ALL changes to the compiler go through a vetting process e.g., peer review before inclusion which is standard practice throughout the industry. Perhaps a new forum area on SF to cover that would help with details of all changes purposed to each module etc,

 

Vincent

 


reply via email to

[Prev in Thread] Current Thread [Next in Thread]