6 ms·

The origin of the name Posix.

92 points by tjr 15y ago

wmf 15y ago
I had no idea RMS participated in POSIX to devalue AT&T Unix. It seemed like ten years later the commercial Unix vendors were trying to use POSIX and UNIX(R) pedantry against Linux.
- ordinary 15y ago
  The list of things rms has had a hand in is pretty intimidating.
  - __rkaup__ 15y ago
    What's the most intimidating?
    - alnayyir 15y ago
      Have you seen the gcc backend?
    - p4bl0 15y ago
      His beard.
- sciurus 15y ago
  He doesn't say he participated in designing POSIX in order to devalue AT&T Unix. He says he wanted to come up with a good name for the standard for UNIX-like systems so that people wouldn't just call the standard UNIX. In his view, that would have strengthened AT&T's position relative to other implementations because AT&T held the trademark on UNIX.
IgorPartola 15y ago
Upvote what you want. Personally, I would rather see a single page that shows where all the UNIX-related names came from. That way we'd have one single story and be done with it.
benihana 15y ago
Since rms came up with the name, you can pretty much assume that he thinks it's the greatest thing that could have possibly happened.
- imurray 15y ago
  The record shows a rather more reflective attitude: http://news.ycombinator.com/item?id=1796989 http://news.ycombinator.com/item?id=1796989
zoowar 15y ago
Sadly, what POSIX set out to achieve has devolved such that each mobile platform has a unique programming interface.
- __rkaup__ 15y ago
  It's been successful in keeping most server OSes similar, I suppose.
- meastham 15y ago
  Actually, both popular mobile operating systems have a POSIX-y system programming interface under the hood. Whether application developers can get at it is a different matter.
  - hakl 15y ago
    Even Symbian recently got a POSIX-layer: http://en.wikipedia.org/wiki/P.I.P.S._Is_POSIX_on_Symbian http://en.wikipedia.org/wiki/P.I.P.S._Is_POSIX_on_Symbian I have no idea how complete it is, but they did manage to port Qt.
- contextfree 15y ago
  even if that's so, 20-plus years is a good run for any standard.
mturmon 15y ago
I think the story of Stallman's influence over the committee is interesting. The general template is, there's a bunch of people that discuss the ideas endlessly, but leaving certain key issues, or perhaps the main issue, undecided. The strategic choice, if you care about the outcome, is to ensure you are present at the bitter end, when the committee is writing down its final recommendation, and then persuasively suggest the solution you favor. The non-strategic choice is to join in the endless debate at the other end of the process. You will tire out, possibly get angry, and lose your position as the neutral arbiter which you will need at the end.
eridius 15y ago
Why were they going to go with IEEEIX in the first place? Or rather, why the "IX" suffix? To my knowledge, the name UNIX was a play on the name Multics, and the IX in UNIX doesn't have any independent meaning.
- elwin 15y ago
  A lot of Unix variants used the IX suffix (AIX, IRIX, Xenix...), probably to make it obvious that they were Unix variants.
maw 15y ago
For more on rms and POSIX, see: http://en.wikipedia.org/wiki/POSIX#512-_vs_1024-byte_blocks http://en.wikipedia.org/wiki/POSIX#512-_vs_1024-byte_blocks. Edit: better link: http://linuxdevcenter.com/pub/a/linux/2004/12/22/rms_interview.html?page=2 http://linuxdevcenter.com/pub/a/linux/2004/12/22/rms_intervi..., last question.
leoc 15y ago
> [Sun May 17 2009] [15:03:35] <stu8ball> rmsgnu: Is it true you deliberately named POSIX with a "POS" for "Piece of Shit" in it? [Sun May 17 2009] [15:03:41] <rmsgnu> because not everyone else here was on voice. [Sun May 17 2009] [15:03:43] <telaviv> rmsgnu: so what is the the next big step for the fsf? New software? Winning some court battle? [Sun May 17 2009] [15:03:50] <stu8ball> Given that you're a Lisp hacker I would assume you'd dislike such standardisation efforts. [Sun May 17 2009] [15:03:59] <stu8ball> w.r.t Common Lisp [Sun May 17 2009] [15:04:09] <rmsgnu> No, POSIX refers to "portable operating system interface" with an x. [Sun May 17 2009] [15:04:23] <telaviv> haha [Sun May 17 2009] [15:04:29] <stu8ball> I know that's the _official_ meaning [Sun May 17 2009] [15:04:33] <stu8ball> ;) [Sun May 17 2009] [15:04:34] <rmsgnu> They were going to call it IEEEIX, which sounds like a screem of horror. [Sun May 17 2009] [15:04:34] <rejohn> who is on voip now? [Sun May 17 2009] [15:04:50] <sn9> luckily, "mueslix" is already trademarked [Sun May 17 2009] [15:04:53] <rejohn> telaviv: where are you now? [Sun May 17 2009] [15:04:55] <rmsgnu> I figured nobody would actually call it that and they would instead [Sun May 17 2009] [15:04:58] <vsayer> i'm on voip [Sun May 17 2009] [15:05:09] <rmsgnu> call it "a spec for Unix-like systems". http://www.thelinuxlink.net/forum/viewtopic.php?f=2&t=4058 http://www.thelinuxlink.net/forum/viewtopic.php?f=2&t=40...

92 points by tjr 15y ago

wmf 15y ago

I had no idea RMS participated in POSIX to devalue AT&T Unix. It seemed like ten years later the commercial Unix vendors were trying to use POSIX and UNIX(R) pedantry against Linux.

ordinary 15y ago
The list of things rms has had a hand in is pretty intimidating.
- __rkaup__ 15y ago
  What's the most intimidating?
  - alnayyir 15y ago
    Have you seen the gcc backend?
  - p4bl0 15y ago
    His beard.
sciurus 15y ago
He doesn't say he participated in designing POSIX in order to devalue AT&T Unix. He says he wanted to come up with a good name for the standard for UNIX-like systems so that people wouldn't just call the standard UNIX. In his view, that would have strengthened AT&T's position relative to other implementations because AT&T held the trademark on UNIX.

IgorPartola 15y ago

Upvote what you want. Personally, I would rather see a single page that shows where all the UNIX-related names came from. That way we'd have one single story and be done with it.

benihana 15y ago

Since rms came up with the name, you can pretty much assume that he thinks it's the greatest thing that could have possibly happened.

imurray 15y ago
The record shows a rather more reflective attitude: http://news.ycombinator.com/item?id=1796989 http://news.ycombinator.com/item?id=1796989

zoowar 15y ago

Sadly, what POSIX set out to achieve has devolved such that each mobile platform has a unique programming interface.

__rkaup__ 15y ago
It's been successful in keeping most server OSes similar, I suppose.
meastham 15y ago
Actually, both popular mobile operating systems have a POSIX-y system programming interface under the hood. Whether application developers can get at it is a different matter.
- hakl 15y ago
  Even Symbian recently got a POSIX-layer: http://en.wikipedia.org/wiki/P.I.P.S._Is_POSIX_on_Symbian http://en.wikipedia.org/wiki/P.I.P.S._Is_POSIX_on_Symbian I have no idea how complete it is, but they did manage to port Qt.
contextfree 15y ago
even if that's so, 20-plus years is a good run for any standard.

mturmon 15y ago

I think the story of Stallman's influence over the committee is interesting. The general template is, there's a bunch of people that discuss the ideas endlessly, but leaving certain key issues, or perhaps the main issue, undecided. The strategic choice, if you care about the outcome, is to ensure you are present at the bitter end, when the committee is writing down its final recommendation, and then persuasively suggest the solution you favor. The non-strategic choice is to join in the endless debate at the other end of the process. You will tire out, possibly get angry, and lose your position as the neutral arbiter which you will need at the end.

eridius 15y ago

Why were they going to go with IEEEIX in the first place? Or rather, why the "IX" suffix? To my knowledge, the name UNIX was a play on the name Multics, and the IX in UNIX doesn't have any independent meaning.

elwin 15y ago
A lot of Unix variants used the IX suffix (AIX, IRIX, Xenix...), probably to make it obvious that they were Unix variants.

maw 15y ago

For more on rms and POSIX, see: http://en.wikipedia.org/wiki/POSIX#512-_vs_1024-byte_blocks http://en.wikipedia.org/wiki/POSIX#512-_vs_1024-byte_blocks. Edit: better link: http://linuxdevcenter.com/pub/a/linux/2004/12/22/rms_interview.html?page=2 http://linuxdevcenter.com/pub/a/linux/2004/12/22/rms_intervi..., last question.

leoc 15y ago

> [Sun May 17 2009] [15:03:35] <stu8ball> rmsgnu: Is it true you deliberately named POSIX with a "POS" for "Piece of Shit" in it? [Sun May 17 2009] [15:03:41] <rmsgnu> because not everyone else here was on voice. [Sun May 17 2009] [15:03:43] <telaviv> rmsgnu: so what is the the next big step for the fsf? New software? Winning some court battle? [Sun May 17 2009] [15:03:50] <stu8ball> Given that you're a Lisp hacker I would assume you'd dislike such standardisation efforts. [Sun May 17 2009] [15:03:59] <stu8ball> w.r.t Common Lisp [Sun May 17 2009] [15:04:09] <rmsgnu> No, POSIX refers to "portable operating system interface" with an x. [Sun May 17 2009] [15:04:23] <telaviv> haha [Sun May 17 2009] [15:04:29] <stu8ball> I know that's the _official_ meaning [Sun May 17 2009] [15:04:33] <stu8ball> ;) [Sun May 17 2009] [15:04:34] <rmsgnu> They were going to call it IEEEIX, which sounds like a screem of horror. [Sun May 17 2009] [15:04:34] <rejohn> who is on voip now? [Sun May 17 2009] [15:04:50] <sn9> luckily, "mueslix" is already trademarked [Sun May 17 2009] [15:04:53] <rejohn> telaviv: where are you now? [Sun May 17 2009] [15:04:55] <rmsgnu> I figured nobody would actually call it that and they would instead [Sun May 17 2009] [15:04:58] <vsayer> i'm on voip [Sun May 17 2009] [15:05:09] <rmsgnu> call it "a spec for Unix-like systems". http://www.thelinuxlink.net/forum/viewtopic.php?f=2&t=4058 http://www.thelinuxlink.net/forum/viewtopic.php?f=2&t=40...