Talk:Unisys 2200 Series system architecture
 | This article is rated Start-class on Wikipedia's content assessment scale. It is of interest to the following WikiProjects: | ||||||||||
| |||||||||||
ones complement
[edit]This whole article doesn't mention that the 2200 series machines, like the 1100 series, use ones complement arithmetic. Seems to me that would be worth noting. Gah4 (talk) 21:57, 5 October 2015 (UTC)
36 bit word requirement
[edit]The claim that the DoD required a 36 bit word, rather than a 32 bit word, sounds very dubious to me. 32 bits were as good as unknown at the time, and how do you fit 6 bit characters into them? I'm flagging this with a "citation needed". Consider removing the claim. Groogle (talk) 05:00, 15 December 2015 (UTC)
- I suspect that the comparison to 32 bits isn't true, but the requirement for (at least) 36 might be. Gah4 (talk) 14:29, 31 March 2019 (UTC)
- I believe the correct explanation is given in the 36-bit computing article. It was not driven by DoD, but rather a general desire to process positive and negative 10-digit integers, and it was not unique to Univac. This is in agreement with my own recollection, having worked on OS2200 for 13 years in the 1980s. I propose to remove the DoD claim from the article and replace it with a reference to 36-bit computing. --MinnMike (talk) 04:37, 21 June 2021 (UTC)