Talk:Unix-like/Archive 2
![]() | This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page. |
Archive 1 | Archive 2 | Archive 3 |
Unix versus Unix-like
While the two are obviously related, I see no reason to merge two related topics together. As it is Unix is a very specific trademark which can be granted for usage if someone pays for testing and passes, showing that functionality within the system is in tune with the standard. Unix-like is anything that even loosely follows the posix specs. It's like merging cars with Ford. Janizary 02:12, 5 October 2005 (UTC)
- I'm not sure I agree with the cars and Ford analogy. The difference is that there are many non-Ford brands of automobile, all of which I suspect will have separate articles, but probably no article about all of them except Ford. This would be more like merging Ford with an imaginary any form of automobile that was not a Ford article, into an article called cars. Perhaps Unix and Unix-like might be merged into something like Unix and Unix-like operating systems?
Alternatively, if there's no general consensus on a merge, perhaps a compromise would be to just add a See also note to the top of each page, along the lines of See also: Unix for a specific trademark which can be granted for operating system usage if someone pays for testing and passes and See also: Unix-like for any operating system that follows the Unix philosophy and that loosely follows the POSIX specifications (the reason for the "Unix philosophy" stuff is that MS-Windows has some POSIX compliance, but I would not consider it to be Unix-like). Thoughts/comments? -- All the best, Nickj (t) 03:48, 5 October 2005 (UTC)
- Well, really no Windows release from the DOS or NT codebases has been close to POSIX compliant, they've never tried. With the Unix Services for Windows or cygwin it is pretty close, but those aren't part of the system. Anyways, that's a bit of a tangent. I agree that should the articles should link to one another, but I don't really think they need to be outer comments on the top, they can just be information inside the meat of the article. I meant cars into Ford because it's taking the generic and putting it inside the specific. Janizary 04:16, 5 October 2005 (UTC)
- In addition to the Win32 subsystem, Windows NT includes a POSIX subsystem that is compliant with various POSIX standards. It was removed from Windows XP and Windows Server 2003 and made part of Services for UNIX. It is being reintroduced in Windows Server 2003 R2. For example, see: http://www.microsoft.com/resources/documentation/windowsnt/4/workstation/reskit/en-us/poscomp.mspx -- Seitz 23:29, 8 October 2005 (UTC)
- I agree that merging the two articles would detract from the wikipedia'a usefulness. Are there going to be times when a reader wants to know about unix-like systems? Sure. Lots. Merging the articles would make locating the information more difficult. -- Geo Swan 19:36, 8 October 2005 (UTC)
- I am guessing that the goal of merging the articles would be 1) to make it easier for people to find the Unix-like operating systems when they search for just "Unix" and 2) to eliminate the large amount of duplication between this article and the Unix article. If we don't want to merge the articles, then we should at least eliminate the duplication while leaving a clear pointer to this article. --Seitz 19:42, 9 October 2005 (UTC)
Here is where I was coming from. I get the impression that there are 3 different things:
- There is "UNIX", as in the official trademarked thing.
- There is "Unix", as a general concept of what an O/S should be (the basic commands and functionality it should have, and how it should behave), and which can be considered a family of operating systems that includes both UNIX and Unix-like.
- There is "Unix-like", which is something which is Unix in its concepts and derivations, but which is not UNIX.
At present, we have 2 articles for 3 concepts (mostly as a result of Unix and UNIX currently being the same article).
The current structure leads to debates like "Should Unix system redirect to Unix, or Unix-like?" And where should Unix-style point to?
Moreover, within these two articles we have duplication (largely I suspect because the boundaries of each are not as well-defined as they ideally could be). Examples include the "Free Unix-like operating systems" section in the Unix article overlapping with the Unix-like article.
The two possible ways of solving this would seem to be either a merge into one mega article that was about all 3 things. Or, alternatively split them into 3 articles.
Honestly though, if everyone else is happy with the current structure, then maybe there really isn't a problem, in which case just remove the merge tags and forget I mentioned it. -- All the best, Nickj (t) 00:49, 10 October 2005 (UTC)
I also agree with maintaining the current structure. If you could/would read every page on wikipedia, i suspect you will see at least 25% duplication. Sometimes, duplication of information aint bad -- but mere copying is. As long as we don't copy text but duplicate information _when it's nessecary_, i don't see a problem at all either. -- De Zeurkous, [email protected], Thu Oct 13 19:08:56 UTC 2005.
OK, I will remove the merge tags. -- All the best, Nickj (t) 03:42, 14 October 2005 (UTC)
- Thanks to the 'Open'(-not) officioust/s, who hastily deleted the reference to and debate of the FreeNix category in Wikipedia. The term 'Free' has been filtered out reflecting the economic prejudices of the editor(s) - as exemplified in the ignorant, scornful language used in that abbreviated, now hidden 'debate'. Who appointed you Moderator of the English language? A very bad look Jan. More of the story is available here: http://www.hackstop.org/index.php/FreeNix%21
- Most likely the term was removed because hardly anyone uses it (witness [1] vs [2] (and the latter includes the conference!)) and Wikipedia is not forum to advertise new naming innovations, rather than because of an anti-free conspiracy. There are plenty of pages about free software and related concepts. NicM (talk) 08:20, 6 February 2008 (UTC).
I disagree, both articles claim Mac OSX as Unix and Unix-like. If anything, both articles should be merged into a Unix page that covers the concept and both UNIX trade marked items and non-UNIX trademarked items. I think since Mac OSX wasn't UNIX then became UNIX by nothing more than a filing of a piece of paper that the trade mark is not the defining characteristic of Unix. VChapman ((talk)) 00:41 3 FEB 2010 UTC —Preceding undated comment added 00:41, 3 February 2010 (UTC).
Requested move
- The following discussion is an archived discussion of the proposal. Please do not modify it. Subsequent comments should be made in a new section on the talk page. No further edits should be made to this section.
No consensus to move. Vegaswikian (talk) 19:31, 27 April 2010 (UTC)
Unix-like → Unix-like operating system — Relisting. Vegaswikian (talk) 21:55, 20 April 2010 (UTC)
I think this article should move to Unix-like operating system since "Unix-like" is just an adjective.--Beware the Unknown (talk) 17:13, 3 April 2010 (UTC)
- I was about to agree, but... there are things that are "unix-like", but might not be an "operating system", the best example of this might be cygwin. I think we should leave that for now... --SF007 (talk) 18:17, 3 April 2010 (UTC)
- Oppose - the scope of the article is not just limited to operating systems. --Salimfadhley (talk) 00:05, 6 April 2010 (UTC)
- Oppose because Unix-like is an adjective. It is more inclusive as a result. CrispMuncher (talk) 20:39, 6 April 2010 (UTC).
- Support The introductory line of the article indicates that it is an article about Unix-like operating systems. Propaniac (talk) 17:18, 9 April 2010 (UTC)
- Support Article titles should be accurate and precise. The name unix-like only describes half of the article, but unix-like operating system is more inclusive. An adjective should not be the title of an article about a noun. Operating systems are the main focus of the article, and should be included in the title. --WikiDonn (talk) 10:34, 24 April 2010 (UTC)
- Oppose - it's not just operating systems, it's the layer itself (c.f. Cygwin). There's nothing wrong with keeping it as the adjective - David Gerard (talk) 10:37, 25 April 2010 (UTC)
- Also, most of the links that I've followed that point to the article are shown as "Unix-like operating system". --WikiDonn (talk) 05:03, 27 April 2010 (UTC)
- The above discussion is preserved as an archive of the proposal. Please do not modify it. Subsequent comments should be made in a new section on this talk page. No further edits should be made to this section.
Open Source versus Free Software
I know it's silly to bounce it back and forth, but Free Software has a very specific connotation to it, which is directly linked to the political goals of the FSF, Open Source on the other hand is a little broader in scope and includes Free Software in it. I'd have put open source if it didn't make the section title look like it was mistyped, since that only refers to the source being readable and does not have assoiations with either the OSI or FSF's policies. Janizary 15:55, 21 August 2005 (UTC)
- I'm not sure I understand your comment. Your last sentence first implies that you think "open source" refers to source code that is viewable, and then it implies that you think "open source" isn't associated with OSI's policies. The generally accepted definition of "open source" is the one published by Open Source Initiative (one of many organisations that abreviates to OSI, but probably the most likely here given the context).
- I'm also concerned about your claim that "open source" is a little more broad than "free software". OSI's open source definition is a rewording of FSF's free software definition. They two are practically identical - there are difference in interpretation, but no widely used project uses a license that is open source but not free software, or vice versa.
- This sort of confusion is a natural outcome of a badly thought out term such as "open source". The defining characteristic of what people call open source and free software is that everyone who recieves a copy is free to run, study, modify, and redistribute the software. "open" doesn't convey this meaning. "free" can be confused with zero-cost, but at least once that's clarified, it is an accurate term - and I recommend using it. Gronky 14:55, 20 October 2005 (UTC)
- Yeah I know, its five years since that post, but I felt I wanted to reply anyway. I think that those who says that the lines between free software and open source (or open software) blurs considerably either is not informed well enough, or they are trying to create a confusion in the terms.
- The distinction doesn't have anything to do with price, free software can be commercial. The distinction has to do with rights, and free software is guarantied rights to keep it free, not matter what the originators or any other potential right holder have in mind. Open source do not work like this, and the potential for a closed source change of license is always present. The FPS game "Nexuiz" is an example of this idea having backfired, and resulted in most of the developers leaving the game and creating a fork, licensed as 'free software'. --Nabo0o (talk) 12:54, 7 December 2010 (UTC)
A vague sentence
"Apple Mac OS X 10.5 and later is BSD variant, and has been certified."
The article should better explain what it means to be "BSD variant." A page on developer.apple.com says that "Integrated with Darwin is a customized version of the Berkeley Software Distribution (BSD) operating system (currently FreeBSD 5)" -- and I believe this is true of all versions of OS X, not just "10.5 and later." Now, on to the phrase "has been certified." Certified as what... POSIX-compliant? 71.221.123.158 (talk) 04:38, 14 February 2011 (UTC)
What about this information from "The Unix-Haters Handbook"?
B Creators Admit C, Unix Were Hoax FOR IMMEDIATE RELEASE In an announcement that has stunned the computer industry, Ken Thompson, Dennis Ritchie, and Brian Kernighan admitted that the Unix operating system and C programming language created by them is an elaborate April Fools prank kept alive for more than 20 years. Speaking at the recent UnixWorld Software Development Forum, Thompson revealed the following: “In 1969, AT&T had just terminated their work with the GE/AT&T Multics project. Brian and I had just started working with an early release of Pascal from Professor Nichlaus Wirth’s ETH labs in Switzerland, and we were impressed with its elegant simplicity and power. Dennis had just finished reading Bored of the Rings, a hilarious National Lampoon parody of the great Tolkien Lord of the Rings trilogy. As a lark, we decided to do parodies of the Multics environment and Pascal. Dennis and I were responsible for the operating environment. We looked at Multics and designed the new system to be as complex and cryptic as possible to maximize casual users’ frustration levels, calling it Unix as a parody of Multics, as well as other more risque allusions. “Then Dennis and Brian worked on a truly warped version of Pascal, called “A.” When we found others were actually trying to create real programs with A, we quickly added additional cryptic features and evolved into B, BCPL, and finally C. We stopped when we got a clean compile on the following syntax: for(;P("\n"),R=;P("|"))for(e=C;e=P("_"+(*u++/ 8)%2))P("|"+(*u/4)%2); “To think that modern programmers would try to use a language that allowed such a statement was beyond our comprehension! We actually thought of selling this to the Soviets to set their computer science progress back 20 or more years. Imagine our surprise when AT&T and other U.S. corporations actually began trying to use Unix and C! It has taken them 20 years to develop enough expertise to generate even marginally useful applications using this 1960s technological parody, but we are impressed with the tenacity (if not common sense) of the general Unix and C programmer. “In any event, Brian, Dennis, and I have been working exclusively in Lisp on the Apple Macintosh for the past few years and feel really guilty about the chaos, confusion, and truly bad programming that has resulted from our silly prank so long ago.” Major Unix and C vendors and customers, including AT&T, Microsoft, Hewlett-Packard, GTE, NCR, and DEC have refused comment at this time. Borland International, a leading vendor of Pascal and C tools, including the popular Turbo Pascal, Turbo C, and Turbo C++, stated they had suspected this for a number of years and would continue to enhance their Pascal products and halt further efforts to develop C. An IBM spokesman broke into uncontrolled laughter and had to postpone a hastily convened news conference concerning the fate of the RS/6000, merely stating “Workplace OS will be available Real Soon Now.” In a cryptic statement, Professor Wirth of the ETH Institute and father of the Pascal, Modula 2, and Oberon structured languages, merely stated that P. T. Barnum was correct.
Page 338 of "The UNIX-haters handbook" http://simson.net/ref/ugh.pdf — Preceding unsigned comment added by 86.125.191.144 (talk) 11:55, 18 December 2011 (UTC)
- What about it? Guy Harris (talk) 21:08, 18 December 2011 (UTC)
- That was an April fools joke back in 1991, see my post on that here. Gwen Gale (talk) 00:30, 19 December 2011 (UTC)
Unix-like vs. POSIX-compliant
Currently the article-entry contains: "behaves in a manner similar to a Unix system". What does this exactly mean? I think I understand POSIX-compliance, but I do not comprehend "behaves in a manner similar to a Unix system". ScotXW (talk) 09:36, 13 October 2013 (UTC)
- The second paragraph of the lede addresses your comment TEDickey (talk) 11:01, 13 October 2013 (UTC)
- The second paragraph is about Unix, not about Unix-like. So what technical specifications are required to be met in order to get such a certification? Single UNIX Specification? POSIX? Then this article should be merged into Unix and a new article List of SUS/POSIX conformant OSes be created ScotXW (talk) 21:41, 4 November 2013 (UTC)
- no, the second paragraph says
There is no standard for defining the term, and some difference of opinion is possible as to the degree to which a given operating system is "Unix-like".
- and the word "term" is in reference to the opening sentence of the lede, i.e., "Unix-like" TEDickey (talk) 22:06, 4 November 2013 (UTC)
What constitutes Unix-like?
Oknazevad: "This requires a more thorough discussion. The term "Unix-like" is understood to mean internal structure of handling data, not just used land level" [3] - I assume should say "userland level". I'm willing to take this discussion, just preferally not on the webOS-, Android-, Firefox OS-, BlackBerry 10, Chrome OS and Template:Unix talk pages etc. Has any of them saied they are Unix-like? And Tivo. Just as Tivo has some "internal structures", that is the Linux kernel, it is not an operating system in the sense that is Unix-like, to run (and compile) "Unix programs". If that kernel or Darwin or QNX is embedded in an operating system but closed off from running Unix program it is no longer a Unix-like OS. The kernel is then just a (big) library.. Does (un-rooted) Android let's say follow the SUS? I maybe wrong about some of the systems that I know less about but my feeling is that at least this applies to iOS, Firefox OS (all APIs except web cut off by design) and Android. comp.arch (talk) 20:02, 26 March 2014 (UTC)
- The article opens with
- A Unix-like (sometimes referred to as UN*X or *nix) operating system is one that behaves in a manner similar to a Unix system, while not necessarily conforming to or being certified to any version of the Single UNIX Specification.
- There is no standard for defining the term, and some difference of opinion is possible as to the degree to which a given operating system is "Unix-like".
- The term can include free and open-source operating systems inspired by Bell Labs' Unix or designed to emulate its features, commercial and proprietary work-alikes, and even versions based on the licensed UNIX source code (which may be sufficiently "Unix-like" to pass certification and bear the "UNIX" trademark).
- so the term is understood in different ways by different people. "Behaves" could refer to the behavior as seen by an end user (command-line or GUI), a system administrator, or a programmer. Given that UNIX systems have had a ton of different GUIs, I'm not inclined to view the GUI as mattering. As a programmer and user, and given the wide variety of administrative UIs and init daemons, I tend to view a system as "Unix-like" if the command-line interface and APIs are like those of Unix, and that it's "Unix-like" if its native APIs and command-line interface are enough like those from the Single UNIX Specification (so I don't count z/OS, for example, even though the UNIX System Services was certified as UNIX(R)).
- So I'd be tempted not to include embedded OSes, such as the one on Tivo boxes, that don't run third-party applications. Un-jailbroken iOS might count at the API level if you can use, for example, open() or socket() in an App Store application, even though you can't use dlopen() (no loading arbitrary code that Apple can't vet). Un-jailbroken Android might count at the API level if you can write software that directly uses Bionic. However, neither of them offer a Unix-style command line, so I'd only count them as "half Unix-like" at the most. I wouldn't count OSes that only support JavaScript apps or only support Java apps or ... as "Unix-like", even if the language implementation is built atop a bunch of Unix APIs.
- (I'm not sure what "internal structure of handling data" means, but I, at least, would not require that the internal data structures of a "Unix-like" OS's kernel, for example, resemble those of anything AT&T ever released as The UNIX Time-Sharing System.) Guy Harris (talk) 20:31, 26 March 2014 (UTC)
- One could, however, list embedded OSes that don't support third-party applications, and OSes supporting third-party applications but not exposing Unix APIs in C, in a separate section, if that's of interest. One might also put OSes that don't offer a Unix command-line interface in a separate section even if they do expose Unix APIs in C. Guy Harris (talk) 20:39, 26 March 2014 (UTC)
- According to Wikipedia, MS DOS 2.0 was Unix-like. Is the term meaningful? --Wtshymanski (talk) 21:12, 26 March 2014 (UTC)
- One could, however, list embedded OSes that don't support third-party applications, and OSes supporting third-party applications but not exposing Unix APIs in C, in a separate section, if that's of interest. One might also put OSes that don't offer a Unix command-line interface in a separate section even if they do expose Unix APIs in C. Guy Harris (talk) 20:39, 26 March 2014 (UTC)
- According to which page of Wikipedia? I didn't see anything obvious on Unix-like or MS-DOS or even Template:Unix and Category:Unix. (And, of course, at one point in time iOS "fucking [sucked]", but, as of earlier and later points in time, it didn't, so, if some page currently claims that MS-DOS 2.0 was Unix-like, that can be fixed.) Guy Harris (talk) 23:38, 26 March 2014 (UTC)
- "If that kernel or Darwin or QNX is embedded in an operating system but closed off from running Unix program it is no longer a Unix-like OS." Seems to me that you assume the term "Unix-like" implies an ability to run traditional Unix programs, while the more general use of the term to mean internally structured according to Unix design principles is what has been meant on Wikipedia. I tend toward the latter, myself, as the userland is only a portion of the complete OS, and the kernel is still the same. That understanding is not limited to here on Wikipedia, either, as seen in this Forbes article, here, here, where apple explicitly refers to iOS development as being UNIX-based, or this one about Android. But I wouldn't mind getting more opinions. oknazevad (talk) 21:17, 26 March 2014 (UTC)
- Basing it on the OS's internal structuring is silly. Much of System V Release 4's internal structure was different from that of V7; does that make one or the other of them not Unix-like?
- The kernel is not "the same" for all OSes considered Unix-like; the Linux kernel, XNU, the FreeBSD kernel, the NetBSD kernel, the Solaris kernel, etc. all differ to varying degrees.
- There may be some Wikipedia pages that speak of "Unix-like" as meaning "mean internally structured according to Unix design principles" (whatever "according to Unix designed principles" means for an OS's internals), but there's one Wikipedia page that, in its lede, speaks of it in terms of its behavior, not its internal structure. Guy Harris (talk) 23:44, 26 March 2014 (UTC)
- What I meant by that is that the kernel is the same for say, Ubuntu, and for Android, or for OSX and iOS, respectively. The former in each pair are undisputedly "Unix-like", and I, and many others, as shown by the sources above, would argue that the latter in each pair are as well, as at the most fundamental level they still work like Unix, even if the userland is different. As for what "Unix design principles" means, the standard reference is Design of the UNIX Operating System, which describes the standard model for all Unix--like systems, regardless of lineage. That is what is meant by "Unix-like" in terms of internal structure and behavior. oknazevad (talk) 02:27, 27 March 2014 (UTC)
- There may be some Wikipedia pages that speak of "Unix-like" as meaning "mean internally structured according to Unix design principles" (whatever "according to Unix designed principles" means for an OS's internals), but there's one Wikipedia page that, in its lede, speaks of it in terms of its behavior, not its internal structure. Guy Harris (talk) 23:44, 26 March 2014 (UTC)
- And I, at least, would argue that:
- 1) the kernel isn't all there is to "working like Unix" - much of the userland API is implemented in userland atop kernel primitives that do not, of themselves, directly provide the functionality (there are no kernel APIs, in most kernels in Unix-like systems, for translating host names to IP addresses; the kernel APIs in question transmit packets at the transport layer, with no knowledge that they happen to be, for example, DNS packets);
- 2) a system can "work like Unix" in a number of ways, and the ability to run third-party programs written to the Unix APIs is one of them, so (as I suggested in an earlier comment) there are "systems that you can program as if they're Unix" and there are "systems that, at the bottom, implement Unix APIs, but you can't get at them" (whether because they're purely embedded, such as on a Tivo box or pick-your-random-NAS-box or..., or because the actual application APIs are not Unix APIs even though they're implemented atop Unix APIs), so I'd draw a distinction between, say, Chrome OS and OS X/iOS (assuming you're allowed to, say, use open() directly in App Store apps)/Android (I'm inferring that the NDK lets you use what Unix APIs are in Bionic)/most desktop and server Linux distributions/Solaris/HP-UX/AIX/*BSD/etc.. (I might further draw a distinction between those with a Unix command-line interface and those without one.)
- (Similar distinctions could be drawn for, say, Windows. There's "embedded Windows", where some company sells something such as an ATM running Windows and there's no way to use the Windows APIs on the box unless you're working at the company; there's Windows RT, where the only APIs exposed to developers are the [[Windows Runtime] APIs, even though the implementation of those APIs is, as far as I know, built atop either the Win32 API, the Native API, or both; and there's "regular Windows".)
- "The standard model for all Unix-like systems, regardless of lineage" is not, as I remember (it's been ages since I read the Bach book) the standard model for, say, GNU Hurd, with a lot of the functionality provided in the Solaris/AIX/HP-UX/Linux/XNU/*BSD kernel is, instead, provided by user-mode server processes. Yes, they all have processes, file descriptors, sockets, signals, etc., but that's API, not the "internal structure of handling data" - you don't have to have a struct proc to be a Unix-like system. Guy Harris (talk) 05:49, 27 March 2014 (UTC)
- Thanks, Guy Harris. I already asked a old and wise guy to confirm what I thought privately, but I can always rely on you to give a detailed answer with all the history perspective. It seems we are in agreement and all reverts to my deleted of "Unix-like" can be reverted. There is just one problem, the more we argue here for this position the more WP:OR it might seem. It only takes one random guy on the Internet do disagree and revert and to point to eg. the Forbes article above. Technical refutation is not enough to address a, non-reliable in this case, business journalist source. I would always take a technical blog (from authority) over that. We can all agree that some OSes are Unix-like and some are not Unix-like (or are Windows-like, VMS, z/OS, RISC OS etc.) What from the Template_talk:Unix-like can we agree that are not Unix-like? I may have been wrong on somethings, may have to rethink iOS, but then there are fuzzy cases. Windows is not Unix-like (without POSIX sub-system), with it, it is "also" or a "hybrid Unix-like + .."; then packages like these are not an OS (true) but the combination is: "[4]. comp.arch (talk) 09:22, 27 March 2014 (UTC)