On Tue, 21 Aug 2018 10:40:56 -0400, Phil Smith III wrote: > >>There is one. It's called Unicode. IBM should open its eyes to the 21st >>Century. > >Now that's funny! Sure, we'll just transform all our data from EBCDIC. > The slower you peel the adhesive tape, the longer you endure the pain.
On Tue, 21 Aug 2018 17:16:53 +0000, Seymour J Metz wrote: >> This whole "tagging" business is a pathetic attempt to cover up two >> original blunders. > >But not the ones that you cite. The original blunders were small byte sizes >for both ASCII and EBCDIC, leading to a plethora of incompatible extensions of >both. The answer is Unicode, though perhaps not RFC 8369 >(https://sandbox.ietf.org/doc/rfc8369/ <g, r & d>), but getting there will be >a bear. > When I see something like that, I always check the date, if only to see how recent it is. 2? 3? 4 and counting? 1) Small byte sizes To accommodate expensive storage economically. 2)) EBCDIC To accommodate 7-track tapes economically. https://web.archive.org/web/20180513204153/http://www.bobbemer.com/P-BIT.HTM 3) Overloading committed code points ("varying code positions" -- http://www.unicode.org/reports/tr16/#Step%201 when plenty uncommited ones were available), impelling variant code pages. Economics again. It was cheaper to put gummed labels on key caps than to rewire the 029 to generate additional codes. 4) Making CP 1047 rather than CP 1208 the convention for UNIX System services. How Retro! -- gil ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
