From: SMTP%"everhart@star.zko.dec.com" 24-AUG-1996 15:09:30.76 To: EVERHART CC: Subj: Re: Converting VMS Fortran source code/DCL to Windows NT - comp.os.vms #153520 Date: Sat, 24 Aug 1996 11:40:17 -0400 Message-Id: <96082411401691@star.zko.dec.com> From: everhart@star.zko.dec.com (Glenn C. Everhart 603 881 1497) To: everhart@gce.com Subject: Re: Converting VMS Fortran source code/DCL to Windows NT - comp.os.vms #153520 X-VMS-To: smtp%"everhart@gce.com" In article <765304345wnr@nugate.demon.co.uk>, Dave Pickles writes: In article: <4vij3t$4mg@sjx-ixn4.ix.netcom.com> acrphl@ix.netcom.com(ACR Computer Dept.) writes: > > Does anyone have any information about where to look for a estimate of > the effort involved in a VMS to Windows NT conversion/rewrite. > Something like a rule of thumb such as # of lines of code vs. > programmer days or case studies or personal experieence would be > most helpful. > Are you by any chance going to use Microsoft Fortran Powerstation 4.0 to do the job? Then beware. The product is buggy as hell, there are pages of VMS-compatibility features in the manual that didn't make it into the code, and other bits that MS tech support admit don't work. Most of our problems were to do with I/O processing. For example, there is no equivalent to the VMS CARRIAGECONTROL=NONE qualifier to the OPEN statement, but there is a FORMAT parameter which inhibits newlines. Fine, except that if you have a WRITE statement that overflows its FORMAT, then from that point on you will always get newlines whether you asked for them or not! If your code is mainly mathematical you might not have too many problems. Anything with fancy I/O I would be tempted to recode in C. -- Dave Pickles