[LMB] OT: Prime News

Marc Wilson marc.wilson at gmx.co.uk
Wed Nov 15 14:28:29 GMT 2017

On Wed, 15 Nov 2017 08:35:41 -0500, Walter Bushell <proto at panix.com>

>> On Nov 14, 2017, at 10:37 PM, Marc Wilson <marc.wilson at gmx.co.uk> wrote:
>> Someone, can't currently find the reference, claimed that using short
>> dates, given the vast cost of early memory, actually saved money overall
>> despite the price of Y2K fixes.  
>> And some of the problem was the old: "Well, this won't be around in xx
>> years..." fallacy.  I've been using 8-digit dates in all new systems
>> since the late 80s, and people said I was wasteful.  When I worked on
>> the Y2K project, none of the systems needing fixing were mine, I was
>> pleased to say (and occasionally, insufferably smug, I must admit).
>The root of the problem was getting all the information on one 80 column 
>card. Having more the information on more than one card caused operational
>problems.  The 80 column card carried way into the age of mag tape and disk 
>drives even.

I remember 80-column cards, I've even punched them.  But the system I
was working on could cope with 132 columns even back in 1980 when I
first encountered it (now does 998).

Graduate school is where you learn to call a spade a leveraged
tactile-feedback geomass delivery system. - Martha Koester

More information about the Lois-Bujold mailing list