eddie Posted May 28, 2016 Share Posted May 28, 2016 The company I was working for had systems that had some programs/modules dating from the late 1960's in some instances, user exits to Fortran and Assembler subroutines etc and occasionally it was impossible to locate source code, yet the object code (the executables) would exist. Other problems would manifest themselves regarding version control, so you could compile some code, compare the object code to a previous version and find that the size differed - so something was different. The worst case would be when you then tested both versions and got different results. That was before we had even changed a line of code. Luckily that was a rarity and for the most part things were where they were supposed to be. After I went back to work in August 2000, I worked in Ireland as a consultant converting financial systems from Punts to Euros, so it really was just employing exactly the same techniques as the Y2K conversion. I was disappointed that the UK decided not to go into the Euro. I would have made an absolute fortune, but probably finished off my liver for good. Link to comment Share on other sites More sharing options...
davenportram Posted May 28, 2016 Share Posted May 28, 2016 did Eddie create the problem or was he the one who fixed it Link to comment Share on other sites More sharing options...
Sith Happens Posted May 28, 2016 Share Posted May 28, 2016 3 minutes ago, eddie said: The company I was working for had systems that had some programs/modules dating from the late 1960's in some instances, user exits to Fortran and Assembler subroutines etc and occasionally it was impossible to locate source code, yet the object code (the executables) would exist. Other problems would manifest themselves regarding version control, so you could compile some code, compare the object code to a previous version and find that the size differed - so something was different. The worst case would be when you then tested both versions and got different results. That was before we had even changed a line of code. Luckily that was a rarity and for the most part things were where they were supposed to be. After I went back to work in August 2000, I worked in Ireland as a consultant converting financial systems from Punts to Euros, so it really was just employing exactly the same techniques as the Y2K conversion. I was disappointed that the UK decided not to go into the Euro. I would have made an absolute fortune, but probably finished off my liver for good. I dont know how true it is, but i did read that a lot of 'quick fix' approaches were taken with regards to dates, so code was changed so if a date exceeded yy30 for example it was assumed it meant 1930, wheras if the date was yy02 its assumed it means 2002, and as a result some really old systems that are still running will start to fall over when the quick fix dates are reached. I guess if its true the concern would be that whomever implemented the quick fix is long gone. Link to comment Share on other sites More sharing options...
King Kevin Posted May 28, 2016 Share Posted May 28, 2016 Wages clerk at the City Ground. Link to comment Share on other sites More sharing options...
eddie Posted May 28, 2016 Share Posted May 28, 2016 22 minutes ago, davenportram said: unsung hero. on a serious point is it the case that a lot of these large systems such as the Nuclear arsenal computers were built on bespoke hardware, with bespoke code that has one purpose, and as that purpose hasn't changed why update it? Physical system degradation, especially with regard to magnetic storage devices. Things break, R/W heads have a tendency to score disk surfaces (thnk CDs/DVDs which may 'jump') especially when dust is present, and in the worst possible case, disks may actually shatter. Nothing is designed to last for ever. Link to comment Share on other sites More sharing options...
eddie Posted May 28, 2016 Share Posted May 28, 2016 6 minutes ago, Paul71 said: I dont know how true it is, but i did read that a lot of 'quick fix' approaches were taken with regards to dates, so code was changed so if a date exceeded yy30 for example it was assumed it meant 1930, wheras if the date was yy02 its assumed it means 2002, and as a result some really old systems that are still running will start to fall over when the quick fix dates are reached. I guess if its true the concern would be that whomever implemented the quick fix is long gone. Yes, that approach was taken in many instances, but not really with respect to dates which had to interface with the operating system. Gradually the old legacy systems are being replaced - whch brings me back to where I am working now and what I am doing (looking at old code and telling people what it does) so they can re-engineer and re-platform the functionality. Link to comment Share on other sites More sharing options...
davenportram Posted May 28, 2016 Share Posted May 28, 2016 4 minutes ago, eddie said: Physical system degradation, especially with regard to magnetic storage devices. Things break, R/W heads have a tendency to score disk surfaces (thnk CDs/DVDs which may 'jump') especially when dust is present, and in the worst possible case, disks may actually shatter. Nothing is designed to last for ever. yep, but still why upgrade unless you really have to. Id guess the upgrade cost would be greater than the repair costs or it would happen. Link to comment Share on other sites More sharing options...
eddie Posted May 28, 2016 Share Posted May 28, 2016 2 minutes ago, davenportram said: yep, but still why upgrade unless you really have to. Id guess the upgrade cost would be greater than the repair costs or it would happen. I tend to agree - after all, these systems must have been evaluated/amended for Y2K. I'm not sure whether repairs are feasible though in some instances - nobody manufactures 8" floppies any more, and IBM can't have many 3330-I disk drives left. Link to comment Share on other sites More sharing options...
Sith Happens Posted May 28, 2016 Share Posted May 28, 2016 2 minutes ago, eddie said: I tend to agree - after all, these systems must have been evaluated/amended for Y2K. I'm not sure whether repairs are feasible though in some instances - nobody manufactures 8" floppies any more, and IBM can't have many 3330-I disk drives left. 8" floppies - snigger Link to comment Share on other sites More sharing options...
davenportram Posted May 28, 2016 Share Posted May 28, 2016 13 minutes ago, eddie said: I tend to agree - after all, these systems must have been evaluated/amended for Y2K. I'm not sure whether repairs are feasible though in some instances - nobody manufactures 8" floppies any more, and IBM can't have many 3330-I disk drives left. id hope they have created a way of replacing the old floppies with modern media but who knows. Something like this but for floppy drives. Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.