Re:How did software get so reliable without proof? (1996) [pdf]
by nikofeyn on Monday September 24, 2018 @02:13AM
(#18055418)
was 1996 really that different than today? because reliable is the last word i would use to describe software today.
[ Parent
| on HN ]
Re:How did software get so reliable without proof? (1996) [pdf]
by passwd on Monday September 24, 2018 @04:01AM
(#18055818)
It surely is different. We at least managed to contain some of the issues in normal operation. Back in the day, crashing whole OS in BSOD/panic style was more common, drivers unstable, things written in not-so-safe languages more popular.
So ultimately it depends what do you mean by reliable - by itself, functionality might not be so much better, but yes, sometimes you don't mind as much as you would do back then.
Re:How did software get so reliable without proof? (1996) [pdf]
by ashleyn on Monday September 24, 2018 @06:10AM
(#18056396)
The internet wasn't as big and software frequently shipped as a boxed product on a disk, cartridge, or CD-ROM. You had one shot to get it right without major bugs because fixing them required another run of physical media.
Re:How did software get so reliable without proof? (1996) [pdf]
by 2sk21 on Monday September 24, 2018 @08:18AM
(#18057031)
Even in the late 90s, source code revision control was a dark art: either you used extremely expensive proprietary systems or else you used CVS which was very low level and completely lacked integration with any bug tracking system. Merging in a big feature in CVS was a nerve-racking experience.
Re:How did software get so reliable without proof? (1996) [pdf]
by patrickg_zill on Monday September 24, 2018 @06:39AM
(#18056523)
There was more Pascal and more assembly language.
Some believe that Macintosh suffered a lot of instability from the switch to C from Pascal.
I don't have any source code to point to however, only anecdote.