Well, considering that Dennis Ritchie wrote the filename comparison code (among lots of the rest of Unix), I'll take his denial that the cost of the comparison was a consideration as authoritative.
And no, they weren't thinking of languages when they did it, but as Charles points out, there are infinite complexities hiding under there. On Fri, Oct 24, 2025 at 6:43 PM Phil Smith III <[email protected]> wrote: > I disbelieve that people working with 7-bit ASCII on tiny, slow machines > were thinking "Oh, we need to be ready to deal with Georgian [or whatever > languages he's talking about]". That feels like a retroactive > justification. I have heard anecdotally that it was the cost of the > comparison as well; that doesn't make it true, but given the PDP > architecture, I'd believe it over this language thing. > > -----Original Message----- > From: IBM Mainframe Discussion List <[email protected]> On Behalf > Of Jay Maynard > Sent: Friday, October 24, 2025 6:58 PM > To: [email protected] > Subject: Re: IRXJCL oddity > > I have this argument periodically with Eric Raymond. He swears up and down > that it matters in languages other than English, especially when there are > no single ways to translate upper to lower case or vice versa as in some > languages. Personally, I think case sensitivity is a botch, because people > just don't think they're different. > > Once upon a time, I stated on Usenet that it seemed to me that Unix > filenames were case sensitive because the folks who wrote Unix originally > didn't want to expend the few instructions needed to make case comparisons > insensitive. Dennis Ritchie corrected me, and said that no, they thought > being case sensitive was the Right Thing. > > On Thu, Oct 23, 2025 at 2:55 PM Phil Smith III <[email protected]> wrote: > > > No no, I meant GLOBALLY: no filenames on the planet should ever have > > been case-sensitive! Nobody sane would ever name two files "Foo" and > > "FOO", so it's just stupid. I keep asking *ix folks about this--have > > for 40 years--and all insist "It's good" without ever being able to > justify it. > > > > -----Original Message----- > > From: IBM Mainframe Discussion List <[email protected]> On > > Behalf Of Mike Schwab > > Sent: Thursday, October 23, 2025 3:48 PM > > To: [email protected] > > Subject: Re: IRXJCL oddity > > > > Mixed case use has been used, so must account for reads and deletes. > > > > On Thu, Oct 23, 2025 at 2:45 PM Paul Gilmartin < > > [email protected]> wrote: > > > > > On Thu, 23 Oct 2025 14:00:51 -0400, Phil Smith III <[email protected]> > > > wrote: > > > > > > >If true, I'm gonna assume this is some C thing related to a null > > > [terminated] string. Agreed seems beyond stupid! > > > > > > > I know your prejudice, but don't misattribute things such as this to > > > it. > > > > > > >Which is on my time machine list: go back and stop them from ever > > > >using > > > that horrible convention (along with EBCDIC/ASCII and *ix case > > > sensitivity). > > > > > > > Such an accommodation should be made at the lowest level so all > > > facilities may most simply exploit it. > > > > > > Please submit an Idea that STOW, BLDL, and Catalog Services be made > > > case insensitive, > > > > > > But remember: <https://www.hyrumslaw.com/> > > > > > > -- > > > gil > > > > > > -------------------------------------------------------------------- > > > -- For IBM-MAIN subscribe / signoff / archive access instructions, > > > send email to [email protected] with the message: INFO > > > IBM-MAIN > > > > > > > > > -- > > Mike A Schwab, Springfield IL USA > > Where do Forest Rangers go to get away from it all? > > > > ---------------------------------------------------------------------- > > For IBM-MAIN subscribe / signoff / archive access instructions, send > > email to [email protected] with the message: INFO IBM-MAIN > > > > ---------------------------------------------------------------------- > > For IBM-MAIN subscribe / signoff / archive access instructions, send > > email to [email protected] with the message: INFO IBM-MAIN > > > > > -- > Jay Maynard > > ---------------------------------------------------------------------- > For IBM-MAIN subscribe / signoff / archive access instructions, send email > to [email protected] with the message: INFO IBM-MAIN > > ---------------------------------------------------------------------- > For IBM-MAIN subscribe / signoff / archive access instructions, > send email to [email protected] with the message: INFO IBM-MAIN > -- Jay Maynard ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
