[HN Gopher] Documented Unix facilities (across 93 major Unix rel...
       ___________________________________________________________________
        
       Documented Unix facilities (across 93 major Unix releases)
        
       Author : adg001
       Score  : 24 points
       Date   : 2022-10-26 15:15 UTC (7 hours ago)
        
 (HTM) web link (dspinellis.github.io)
 (TXT) w3m dump (dspinellis.github.io)
        
       | jlokier wrote:
       | Some interesting anomalies that look like errors in the tables.
       | E.g. setjmp() present in 32V, BSD3 and BSD 4.2, but longjmp() not
       | present. That doesn't make sense and doesn't match their man
       | pages.
        
       | guiraldelli wrote:
       | What a great surprise (but not shocking) to see it's a work of
       | the author of "Code Reading: The Open Source Perspective" [1].
       | 
       | [1]: https://www.spinellis.gr/codereading/
        
       | asveikau wrote:
       | When they list first appearance, it's weird that the only modern
       | system they consider is FreeBSD.
       | 
       | There are a few interfaces that were introduced in Linux,
       | OpenBSD, NetBSD etc. that saw subsequent broad adoption across
       | all of modern Unix-like OSs, FreeBSD included.
        
         | Maursault wrote:
         | It is very common to list Linux among the UNIXes, and Linux is
         | similar, will operate the same with nearly all of the higher
         | layers and applications, but it is different enough underneath
         | and fundamentally so that while a UNIX administrator can
         | rapidly and fluidly go from NetBSD to OpenSolaris and Illumos
         | to AIX to FreeBSD and OpenBSD without much pain, they will
         | experience pain when first exposed to Linux, and similarly, a
         | Linux administrator or developer that is only well-experienced
         | in RHEL, CentOS, Fedora, SUSE and/or Debian will experience
         | pain when first exposed to AIX or Solarises or BSDs.
         | 
         | For whatever reason, Linus or the distributions themselves
         | arbitrarily deviated from the ways common in all UNIX, in
         | directory structure and tools and other details, that in one
         | sense they are still similar, brothers even, but in another
         | they are entirely different, and distant cousins removed
         | several times, and, again, I believe these deviations were
         | entirely arbitrary, and as such, unnecessary.
         | 
         | /rant, nothing personal, everyone considers Linux as UNIX, and
         | they're not entirely wrong, but the differences are neither
         | minor nor pedantic, and I, for one, wish it had been different
         | and Linus or GNU or distros had not made these unnecessary
         | deviations from standard UNIX and continued to do so, as the
         | advantages of the deviations, if there are any, are negligible.
         | To me they seem different for the sake of being different,
         | unnecessarily adding complexity to an already complex matter.
         | 
         | And I acknowledge that Linus may be entirely innocent and the
         | problem may have always been with GNU, I'm just not certain who
         | the culprits are, but they seem to have made the mistake of
         | changing the architecture without being aware of what came
         | before, falling into the tragedy of reinventing wheels.
        
       ___________________________________________________________________
       (page generated 2022-10-26 23:01 UTC)