Talk:X Window System

From Wikipedia, the free encyclopedia
Jump to: navigation, search
Former featured article X Window System is a former featured article. Please see the links under Article milestones below for its original nomination page (for older articles, check the nomination archive) and why it was removed.
Main Page trophy This article appeared on Wikipedia's Main Page as Today's featured article on September 3, 2005.
This article has been mentioned by a media organization:


Is there any reason for no quotes about X server in the article? Like Ken Thompson's

OpenVMS / DECwindows reference?[edit]

Should "Hewlett-Packard's OpenVMS operating system includes a version of X with CDE, known as DECwindows, as its standard desktop environment" be reworded? Since HP also ships HP-UX (which presumably also has X), perhaps both should be mentioned to avoid confusion, as OpenVMS and DECwindows are legacy DEC products. It just seems a little like saying "DaimlerChrysler's Plymouth Road Runner Super Bird," with the difference that HP is, in fact, maintaining and shipping legacy DEC stuff.

Dunno. You can reword if you like, but OpenVMS is a current product. Just a very high-end one. And considerably less dead than HP-UX - David Gerard 12:16, 2 October 2006 (UTC)

Display PostScript did not become Aqua[edit]

As seen in the Wikipedia article on Display Postscript (, it become Quartz. Aqua is the visual theme in OS X, not the display technology.

Display Postscript was replaced by Quartz. Neither is a window manager. -Ahruman 15:46, 14 December 2006 (UTC)

Ununderstable introduction[edit]

Is it possible to begin this article with a sentence using the verb 'to be' so that it can answer the question that everybody wonders:

"What is X Window System ?"

So, I think the article should begin this way:

In computing, the X Window System (commonly X11 or X) is...

and it's up to you to continue... 16@r 22:35, 7 September 2006 (UTC)

I tried just now and couldn't come up with anything. It is both a dessert wax and a floor topping, or a protocol for constructing such, or toolkits for such - David Gerard 12:15, 2 October 2006 (UTC)
I took a shot at it. (talk) 23:46, 3 September 2013 (UTC)
Inserted this new text:
Simply put, X11 is the system-level software infrastructure for the windowing GUI on Linux, *BSD, and other UNIX-like operating systems; it is designed to handle both local displays, as well as across-the-network displays (contrast with headless servers). More formally,
In front of this existing intro-paragraph:
[T]he X Window System (commonly known as X11, based on its current major version being 11, or shortened to simply X, and sometimes informally X-Windows) is a computer software system and network protocol that provides a basis for graphical user interfaces (GUIs) and rich input device capability for networked computers. It creates a hardware abstraction layer where software is written to use a generalized set of commands, allowing for device independence and reuse of programs on any computer that implements X.

criticisms section and factual inaccuracies/omissions[edit]

Many current implementations of X manipulate the video hardware directly. A misbehaving server can render the display unusable even if the underlying operating system continues to function, possibly requiring a reboot.

So what? A misbehaving driver can lock up any graphical system on any OS. I don't see why this criticism is specific to X at all.
Furthermore, at least on some Linux distros, the criticism is not completely true -- if your xserver locks up the GUI session, you can often use Ctrl+Alt+F4 to get to a textmode 'virtual terminal' and use the cmdln to get things working again. (One would use Ctrl+Alt+F7 to return to the now-fixed-up X11 session once again.) This is actually a useful feature of X11, that is not offered on most other OSes, where when the GUI goes down you have no recourse but to reboot. See also, ctrl+alt+backspace. While I suppose it is true that some X11 hardware-drivers can actually make the display unusable even from a VT, that still just means X11 is in the same boat as msWindows and OSX and friends... whereas, with a properly-written X11 hardware-driver, you can fix up the windowing system without a reboot (not the case in windows... not sure about OSX which is loosely BSD-based nowadays -- anybody know this answer?). (talk) 23:54, 3 September 2013 (UTC)

The X protocol provides no facilities for handling sound, leaving it to the operating system to provide support for audio hardware and sound playback.

I thought MAS was part of the official standard now. Granted, nobody uses it...

The device-independence and the separation of client and server do incur an overhead.

All decent graphics APIs are device independent, from Win32 GDI to Quartz to DirectX to OpenGL. That's kind of the point of the API. To provide device independence.
Also, both Windows and Mac OS X operate on the same "separation of client and server" principle also. All modern GUIs do their actual output and GUI apps in separate processes. If every GUI app had direct access to the hardware (as this criticism seems to suggest) it would be a mess!
Also, when you are running programs locally, X will use MIT-SHM and zero-copy sockets, which greatly reduces the affect of this separation.

I just think that many of these criticisms seem to have been put in by uninformed users. – Andyluciano 15:12, 17 September 2006 (UTC)

Okay, I see that this last point has in fact been given treatment in the article. It needs to be clearer though. I have started editing the section, trying to be careful to do so in a neutral way. – Andyluciano 15:23, 17 September 2006 (UTC)
All modern GUIs do their actual output and GUI apps in separate processes... Not quite sure what 'processes' you mean here. If it is this process, I think you may probably wrong. The actual output could be implemented as a kernel module (or driver) which manipulates the graphical device directly. On the other hand, GUI apps are executed in user space. Each GUI app is a genuine process. However, the module runs in the kernel is not. GUI apps depend on the kernel module to achieve graphics functionalities by making system calls to the kernel. Because GUI apps cooperate with the kernel via function calls rather than IPCs so basically they are in the same process rather than separate ones. - Justin545 (talk) 09:18, 17 March 2009 (UTC)
The criticisms are actual problems that are or were commonly touted. I wrote that para most carefully. I suppose there could theoretically be cites for each ... The device independence one was actually unusual in the late '80s and is addressed about as well as it could be in that paragraph. MAS may be the 'standard', but there is no accepted sound transport standard that people actually use - David Gerard 10:19, 9 October 2006 (UTC)
And I'm somewhat surprised you didn't look up the two sources of criticisms listed at the top of the section! A lot of the criticisms are old and outmoded, but still common - David Gerard 10:48, 9 October 2006 (UTC)
The Unix Hater's Handbook is a patently biased book whose arguments about X were largely specious to begin with, and have become ever less worthwhile with the passage of a great deal of time. This whole section of the article is pretty clearly just a sop to the X trolls, and I find it amusing that the Wikipedia editors let it ride given its non-neutral point of view. Weirdly, the sections on Limitations and criticisms of the Apple Macintosh and Limitations and criticisms of the Windows Operating System contain no discussion of their respective window systems—because they are missing altogether. Who knew? - Bart Massey 11:39, 1 April 2009 (PDT) —Preceding unsigned comment added by (talk)
Not to mention that some of the "limitations" mentioned are actually unique and useful features that set X apart. Namely, the "criticism" about it having a client-server architecture. Whoever writes this shit? :-/

I am curious why there is no reference to the Apollo Aegis(sp?) windowing system. It was a full bitmapped graphics implementation that was network aware and, as I remember, was why Dec and IBM went to MIT to get 'W' updated to 'X'. They knew they were about to get blown away on UI's. Smart money by DEC/IBM; and to HP who eventually went on to buy Apollo. It was a great system in its day! [Leslie Keller, July 13, 2009]

No Sound Support[edit]

Regardless of attempts to actually integrate this in X11, I don't see how the lack of audio support can be a criticism for a 'windowing manager' (there is a clue in the name) any more than we would criticise the fact that it can't make coffee or do any of the other things it isn't designed for in the first place. For this reason, I am removing the offending paragraph. —Preceding unsigned comment added by (talk) 13:32, 21 June 2010 (UTC)

User interface issues and Computer accessibility related issues[edit]

Seems to me that this is about that X not being a full blown Desktop environment, which is akin to the "problem" that X doesn't have support for sound. Anyone mind if I remove them or replace all of them with a new section describing the perceived core problem (that X is not a desktop environment)? FrederikHertzum (talk) 12:47, 8 May 2011 (UTC)

What is this article about?[edit]

A toolkit? A protocol? An implementation? A GUI?

No, yes, a fair bit, not quite - David Gerard 12:08, 2 October 2006 (UTC)
Intro rewritten as "[thing] is ..." - if that isn't clear, add wikilinks to taste - David Gerard 10:17, 9 October 2006 (UTC)

Well, I imagine over 90% of the world's computer-using population has never heard of the X Window System, so it's no wonder people are confused what this article is about. JIP | Talk 08:39, 16 October 2006 (UTC)

To restore the honor of anonymous poster world-wide: I was not confused, but merely pointing out the inconsistency. —The preceding unsigned comment was added by (talk) 21:47, 22 January 2007 (UTC).

Application Migration[edit]

The article asserts that applications cannot migrate. This is not in fact true. An early counter example is Gnu Emacs, which has been about to do so since the 1980's (along with replicated display). (diff author: actually Emacs can not only migrate, but co-use multiple displays, very handy for getting one emacs' multiple windows to spread onto a second X server on your desktop - or assist someone over the Internet in the old days when TCP listen was still on by default) This was painful in the past due to colormaps (though if you thought ahead while building a toolkit, it wasn't difficult). It is now much easier, between client side fonts and true color displays being the norm.

A more recent example is that GTK+ applications can potentially migrate, at least if they do not use obsolete interfaces. See the "teleport" application of the GPE environment. What is still missing is both good integrated authentication and security (encryption) of the protocol to make migration "safe" to do. I'd like to see people pursue this: see my paper And with the rewrite of the input section of the X Server, remoting input devices is also now becoming feasible. Oh, this paper should probably get added to the bibliography.

As I am Jim Gettys 23:57, 01-Jan-2007 (UTC), I leave it to others to verify this and edit the article.

I tried to rephrase the Network section to reflect this, and used your paper as a source. Demian12358 22:41, 2 February 2007 (UTC)

On Portal:Free software, X Window System is currently the selected article[edit]

(2007-02-04) Just to let you know. The purpose of selecting an article is both to point readers to the article and to highlight it to potential contributors. It will remain on the portal for a week or so. The previous selected article was GIMP. Gronky 20:13, 4 February 2007 (UTC)

The selected article box has been updated again, the new selectee is FOSDEM. Gronky 12:55, 12 February 2007 (UTC)

Broken reference link[edit]

One of your references had a broken link.

The means to an X for Linux: an interview with David Dawes from (Matthew Arnison, CAT TV, June 1999)

I used the latest archived page of it since I could not find the page anywhere else.

Psychless 01:38, 25 April 2007 (UTC)

All of these are broken:

31 Invitation for public discussion about the future of X 20 March 2003

32 A Call For Open Governance Of X Development 21 March 2003

33 Notes from a teleconference held 2003-3-27 3 April 2003

34 A Call For Open Governance Of X Development 24 March 2003

35 A Call For Open Governance Of X Development 23 March 2003

36 Discussing issues 14 April 2003

22:10, 10 May 2016 (UTC)

What Happened to X ORG?[edit]

I got the following from a post to my local LUG (GWLUG):Both my ISP's and UT= 's dns servers agree. and > > do not exist!

I got "Firefox can't find the server at" after my attempt to connect to it at

Borgward 16:55, 4 August 2007 (UTC)

Whois lists the domain as ok. None of the nameservers are responding so they are offline. Most likely they put all of their nameservers on the same subnet. No one should do this of course but it is quite common. Robert Brockway 17:06, 4 August 2007 (UTC)

Out of idle curiosity...[edit]

Will there ever be an X12? AnonMoos 21:44, 21 September 2007 (UTC)

Unlikely -- X11 works quite well and the things people are missing can be or have been implemented in extensions. XCB vs. Xlib is far more interesting. FrederikHertzum (talk) 21:49, 6 June 2010 (UTC)
XCB per se is an alternate protocol (and associated interface). Xlib implements a protocol, but that's only a small part of the total amount of code. XCB/Xlib is much like Xlib (Wikipedia doesn't cover it well) Tedickey (talk) 22:05, 6 June 2010 (UTC)
Incorrect. XCB isn't a protocol at all. It is a new C library which talks to an X server, just like Xlib, using the X11 protocol.FrederikHertzum (talk) 12:50, 8 May 2011 (UTC)
The XCB topic does in fact refer to its own "protocol". Keep in mind Wikipedia:TRUE, and provide a reliable source supporting your statement before emitting "Incorrect". TEDickey (talk) 13:55, 8 May 2011 (UTC)
Update as of 2013 -- no, there is not likely to ever be an X12. Instead, current users of X on the desktop will most likely be migrated over to one or more of Wayland (redhat), Mir (ubuntu), or perhaps even SurfaceFlinger (google android slash chromeOS). I just finished updating the competition-section of the article, with info on these competitors-slash-replacements; they also have their own wikipedia articles. There has always been competition to replace X11, but the current de facto standard xserver implementation is from, and they are backing wayland as a replacement for X, at the moment. Backwards compatibility with X11 will be retained, but no X12. (talk) 00:48, 4 September 2013 (UTC)


Yeahb, the picture takes a long time to load on AOL 9.0. It will be nice if we could fix that. PLEASE HELP —Preceding unsigned comment added by (talk) 20:04, 30 December 2007 (UTC)

OS X is a Unix OS ( In the part on Implementations that's about Apple, it seems a lot like the author did not know that. This is NOT a feature. Please fix it. —Preceding unsigned comment added by (talk) 06:55, 13 May 2010 (UTC)

GAN failed[edit]

I've failed this good article nomination primarily because it lacks reliable sources woefully. Most of the sections are not referenced or referenced thoroughly, and when there are references, they are not complete. Other than that, I recommend sifting through the text and correct any grammatical, spelling, and other problems with the prose. Furthermore, some of this article may be too technical for a general audience wanting to know about X (example client-server). If this review was in error, please take this article to good article reassessment. 哦,是吗?(O-person) 03:50, 31 December 2007 (GMT)

Image X11 ssh tunnelling[edit]

The image pretends that one has only to log into another machine using ssh for having an ssh tunnel. But in the example just the call of xeyes is encrypted but not the data xeyes transferes. -- Raubsaurier (talk) 20:33, 19 January 2008 (UTC)

No pretense involved, all data over the net is encrypted; see ssh(1) for details. -- (talk) 00:59, 3 January 2009 (UTC)

(GNOME, KDE etc.) vs. X Window System[edit]

What makes desktop environments such as GNOME and KDE different from the X Window System? --Abdull (talk) 19:58, 8 March 2008 (UTC)

GNOME, KDE, etc., are particular applications that happen to run in X Tedickey (talk) 20:17, 8 March 2008 (UTC)
Well, sets of applications and policies and so on. Marnanel (talk) 02:25, 28 April 2008 (UTC)

Requested fix at Talk:X_window_manager#Metacity.27s_compositor:_requested_fix[edit]

There's a factual inaccuracy in X window manager which has been noted at Talk:X_window_manager#Metacity.27s_compositor:_requested_fix for a while, which I can't fix because of WP:COI; can someone else step in, please? Marnanel (talk) 02:26, 28 April 2008 (UTC)

Fixed (mostly), see comments over at that talk-page. (talk) 02:28, 4 September 2013 (UTC)

versioning system[edit]

Can someone explain the versioning system? There are multiple related but distinct version numbers, such as the protocol version (11), the release version (7.4), and the server version (1.4). This is so poorly defined that even Wikipedia can't get it right. For example, the KDrive article says, "As of X.Org Server version 7.1" even though the X.Org Server is currently only on version 1.4.0. It's not clear what to call the thing that the article is actually intending to refer to. (X? X Release? X System? X Window System? X11?) It's usually referred to using the notation "X11R#.#" because even if nobody knows what to call it, at least this notation lets people know what they're talking about. Is there a guide somewhere that sorts this out? Herorev (talk) 15:26, 19 May 2008 (UTC)

The thing that's currently being numbered in the 7.X range is the "katamari" or "roll-up" release, i.e., a full X.Org "official" release. We generally refer to the collection of software being released in this process as "X.Org". As you correctly point out, the components of X.Org, including the X Server, have their own version numbers with their own sequence. This is no different than e.g. Apple's MacOS X numbering, and represents a fairly common industry practice. See for a list of the various component versions for recent X.Org releases.
It is fairly common practice, if a bit sloppy, to refer to component versions using the version number of X.Org that they were released with. This saves remembering so many numbers. HTH. --Bart Massey (talk) 17:26, 15 October 2010 (UTC)
Using the link above[1], plus some other sources[2][3][4], below is the rough overall-version history as of august 2013. (talk) 04:47, 4 September 2013 (UTC)
rollup xorg-server 1stCommit Fedora Ubuntu NthCommit xrandr evdev vesa i810/intel
X11R7.0 1_0 Mar'06 f4 u6 ~2006 1.0.1 i810-
X11R7.1 1_1 Apr'06 f5 u6 ~2006 1.0.2 1.1.2 1.2.0 i810-1.6.0
X11R7.2 1.2 Nov'06 u7 ~2006 1.0.2 1.1.5 1.3.0 i810-1.7.4
X11R7.3 1.2 Nov'06 u7 ~2006 1.2.2 1.1.5 1.3.0 i810-1.7.4 && intel-2.1.1
n/a 1.3 Aug'07 f6 u7 ~2008 tbd tbd tbd tbd
n/a 1.4 Nov'07 f8 u8 ~2008 tbd tbd tbd tbd
X11R7.4 1.5 Mar'08 f9 u8 ~2009 1.2.3 2.0.4 2.0.0 intel-2.4.2
n/a 1.6 Nov'08 f10 u9 ~2009 tbd tbd tbd tbd
X11R7.5 1.7 ???'09 f11 u10 Nov'11 1.3.2 2.3.0 2.2.1 intel-2.9.1
n/a 1.8 Apr'10 f12 Nov'11 tbd tbd tbd tbd
X11R7.6 1.9 Aug'10 f13 u10 Nov'11 1.3.4 2.5.0 2.3.0 intel-2.13.0
n/a 1.10 Feb'11 f14 u11 Mar'12 tbd tbd tbd tbd
n/a 1.11 Aug'11 f15 u12 Apr'12 tbd tbd tbd tbd
X11R7.7 1.12 Mar'12 f16 Oct'12 1.3.5 2.7.0 2.3.1 intel-2.19.0
n/a 1.13 Sep'12 f17 u12 Apr'13 tbd tbd tbd tbd
n/a? 1.14 Mar'13 f18 u13? Aug'13+ tbd tbd tbd tbd
Note that the data in the article does not match the reference-material used to create the table above. Here is the info currently in the article. (talk) 05:03, 4 September 2013 (UTC)
rollup public_release XServer correlation with table above
X11R7.0 Dec'05 1.??? [date_mismatch]
X11R7.1 May'06 1.??? [unknown]
X11R7.2 Feb'07 1.??? [unknown]
X11R7.3 Sep'07 1.4 [version_mismatch]
X11R7.4 Sep'08 1.5.1 [match]
X11R7.5 Oct'09 1.7 [match]
X11R7.6 Dec'10 1.9 [match]
X11R7.7 Jun'12 1.12 [match]

X10.3 vs X10R3?[edit]

My memory is fading, but I'm pretty sure that it was called "X10.3" back in the day, not "X10R3" I think the R replaced the dot sometime in the X11 series. linas (talk) 07:28, 4 December 2008 (UTC)

Heh, yes. See [5] -- 69 hits for "X10.4 release", vs. 45 for "X10R4 release". linas (talk) 23:51, 5 December 2008 (UTC)

Clarity wanting in "Implementations" section.[edit]

Third-party servers under Mac OS 7, 8 and 9 included White Pine Software's eXodus and Apple's MacX.

I'm not familiar with the Mac OS scene, but if MacX is actually an Apple product, it's not "third-party". Not sure, but perhaps the intent here is that it did not ship with the OS? In that case, maybe replace with:

Mac OS 7, 8 and 9 did not include an X server, but several X servers were available including White Pine Software's eXodus and Apple's MacX.

-- (talk) 00:49, 3 January 2009 (UTC)


Using "The UNIX-HATERS Handbook" (1994) as an intro to the criticism section is a little odd. The book is

  1. more like a laughing-mirror for unix users in the year 1994, and more funny than serious,
  2. considerably blunt, off-topic and not-very accurate as a real criticism of X, it makes failure prognoses that are obviously wrong by current experience,

I think the statement should be moved somewhat further downwards in the criticism section, as a funny detail to refer to. ... said: Rursus (bork²) 13:50, 8 February 2009 (UTC)

I think the Unix Haters stuff is comedic nonsense and really is not a valid criticism. The fact is X Window System does not use as a lot of resources itself. In the early days, long ago, with computers often having 4 MB of RAM, pretty much any GUI system was considered to be huge. So it does not matter how the GUI system was written, it would have been considered huge to the people who wrote the Unix Haters Handbook. This is simply due to the scarce memory on systems at that time. The Unix Haters stuff makes it seem that X is badly designed. This is wrong. X is not badly designed. X does not really use more RAM than other GUIs like Windows. The server's executable is about 10 MB, entire server uses around 30 MB, much of that is buffers rather than code.Millueradfa (talk) 23:20, 4 April 2012 (UTC)

Moving Xming from Free to proprietary[edit]

As discussed on Talk:Xming and other places on the web, the product is no longer exactly free nor Free --SeeFood (talk) 13:47, 9 March 2009 (UTC)

Is that why it is used by GPL projects andLinux and Portable Ubuntu for Windows, and shares its source code with the Cygwin/X project? —Preceding unsigned comment added by (talk) 06:55, 10 March 2009 (UTC)


There are no reliable sources (primary sources don't count) so far for the edits related to DSPSOFT INC Tedickey (talk) 10:40, 5 April 2009 (UTC)

Hi, it's ok to remove the red link if you think the link there is unnecessary or inappropriate. - Justin545 (talk) 14:22, 5 April 2009 (UTC)
That, plus MicroXwin, which is the same set of changes (unless you're going to start adding some reliable sources). Tedickey (talk) 17:00, 5 April 2009 (UTC)
If you insist, I suggest to remove that paragraph and delete the article MicroXwin. MicroXwin is kind of new and third-party sources are not easy to be found. But before you take action, please make sure there is no any objection with respect to the removal and the deletion because:
  1. If the page can be improved, this should be solved through regular editing, rather than deletion.
  2. Disputes over page content are not dealt with by deleting the page. Likewise, disagreement over a policy or guideline is not dealt with by deleting it.
as you can see in Wikipedia:Deletion policy. - Justin545 (talk) 04:22, 6 April 2009 (UTC)
Agree more/less - the MicroXwin topic needs sources (I don't see an immediate reason to delete it). Tedickey (talk) 11:04, 6 April 2009 (UTC)

Bogus images should be removed[edit]

Since certain images have had their Windows background removed, they no longer show what they purport to show. I think they should be removed. Yworo (talk) 18:23, 24 July 2009 (UTC)

Actually, all that would be needed to be non-infringing would be to use a screenshot with the plain blue background, and crop out the toolbar. One could add a blue background to the Xming image, for instance, using gimp. Tedickey (talk) 20:44, 24 July 2009 (UTC)
That's true.... though I think somebody should take a new screenshot and do that. The site this was derived from claims to have put the image with Windows in it into the public domain. Since they aren't actually able to do that, should we be linking to that page on that site? Yworo (talk) 22:12, 24 July 2009 (UTC)
Already done. The toolbar at the top is open source software. The possibly "non-free" elements that Yworo tried to get the image deleted over have been cropped out and replaced with a transparent background. --Tothwolf (talk) 23:01, 24 July 2009 (UTC)

SunDew/NeWS and the AT&T 7300 UNIX PC vs. X-- (talk) 02:16, 19 October 2009 (UTC)Jim[edit]

A couple omissions to the history here.

First of all the AT&T UNIX PC of which I was one of the first owners did not succeed in the market because it was poorly positioned as a competitor to the IBM-PC. It was much more powerful yet vastly more expensive. At the time AT&T couldn't market anything other than long distance telephone service. More or less the same thing happened with Honeywell trying to market the GE line of mainframes which were eventually sold to Bull of France.

As for the X Window System, SunDew or NeWS was vastly superior yet was deliberately crushed by IBM, HP and DEC because they were angry at Sun's de facto standard for NFS and were bound and determined to prevent another. NeWS was mainly only used by Sun and SGI on high end workstations that most people didn't even know existed at the time.

I could just send a PostScript file to my Sun 3/60's display and could program graphics in PostScript. Show me a X display that can do that even today without some kind of interpreter.

Yes, there was a vast conspiracy of evil corporations to crush the wonder that was NeWS. If it had been allowed to flower, it could be as wonderful today as...NFS!!

I could just send a PostScript virus to my Sun 3/60's display and have it propagate itself. Show me an X display that can do that even today without some kind of interpreter.

The fact of the matter is that while PostScript had a better drawing model than X, a Turing-complete window system was (and still is) considered a bug rather than a feature by most folks. There were also various intellectual property...issues. The net today is that X supports primitives that make the PostScript-like drawing model of Cairo ( reasonably fast, and OpenGL has also subsumed quite a bit of the need for fancy rendering. —Preceding unsigned comment added by (talk) 17:37, 15 October 2010 (UTC)

single large expensive, etc.[edit]

That's injecting the editor's opinion, by stating in some many words that there weren't any multiprocessors, that all of the X hosts were large. Expensive is relative (costs have always been calculated per-user). Tedickey (talk) 09:51, 14 May 2010 (UTC)

Suggestions on trimming the lede[edit]

The writing is well polished, and hard to refactor without bruising something, or I would have just waded in and trimmed the lede myself. Instead, a few observations.

Already the second sentence It creates a hardware abstraction layer where software is written to use a generalized set of commands, allowing for device independence and reuse of programs on any computer that implements X. is written for an adept audience.

The second sentence should be X originated at MIT in 1984. Under a permissive license. In was designed at the time to permit time sharing, which is no longer so important, but since it solves some other problems well enough, it remains in widespread use, not without some drawbacks. Then it needs about half as much technical description of which part of the software stack it lays claim to, and what it's like to work at that layer (a world full of windows, fonts, pointers, and C code).

The current protocol version, X11, appeared in September 1987. The lede should state something about the history of the official specification / specification body.

Rather than wading so heavily into enumerating the technical bits and pieces, I would prefer the lede to end on an explanation that X combines the wrong abstractions by modern standards, and is being actively exploded:

  • [Phoronix Wayland: A New X Server For Linux]
  • [Phoronix A Preview Of Kernel-Based Mode-Setting]

Distilling the essence of this change is the part of the writing process I most enjoyed, before I discovered the scope of the prohibition against OR and gifted illuminations of guru grumblings, so these sources are kind of worthless for the purpose here. — MaxEnt 14:30, 15 May 2010 (UTC)

Time-sharing is something different from remote access. X addressed the latter. Tedickey (talk) 15:16, 15 May 2010 (UTC)
As these things do, the intro more or less grew and had subclauses bolted on as time passed. I'd say dive in and condense, moving subclauses into the body as appropriate :-) - David Gerard (talk) 09:45, 19 May 2010 (UTC)
In fact, the entire article is suffering from subclause decay, editorialising, advocacy, advertising and stuff that belongs on the talk page. Do feel free to go the hack on it. Say "see talk page" in the edit summary ;-) - David Gerard (talk) 09:51, 19 May 2010 (UTC)

X lack of success because of Motif licensing should be noted[edit]

Although X and specifically DeskView/X offered an alternative to current Windows dominance, the high cost of developing software using Motif led to too few offerings to be competitive. Alternative like Mootif, Lestiff, and Xtif appeared later. Not much about this in Motif article.Shjacks45 (talk) 09:14, 2 February 2011 (UTC)

Sure, be bold! Go ahead, add it! Rursus dixit. (mbork3!) 18:31, 5 May 2011 (UTC)


When did X11 introduce workspaces? I have used them in 1997 already and my guess it that they were available much earlier. but this article does not talk about them, while on MacOS X this feature gets an extra article: Spaces (software). This is typical. -- (talk) 09:03, 26 June 2011 (UTC)

You guessed right — as Spaces (software) says, "virtual desktops existed for quite some time on other platforms"; according to virtual desktop, switchable desktops have been around since 1986, and were implemented on X in 1989. That said, I see nothing particularly notable about "Spaces", and the "Use" section could certainly be trimmed. Feezo (send a signal | watch the sky) 10:29, 26 June 2011 (UTC) — edited 11:55, 26 June 2011 (UTC)

Lack of encryption[edit]

X Windows was originally developed in a benign computer environment. Getting two computers at that time to communicate with each other was a tricky, and attacks of the type we see today were not considered.

There should be a paragraph on encryption or rather the lack of it. Yes one can start a session using ssh, but for example if one is going to use Xdm then there is no encryption without jumping through hoops. -- PBS (talk) 22:15, 20 January 2012 (UTC)

Encryption can be provided by SSH or another encryption technology, such as SSL (stunnel, etc). There is no need to have encryption in the X server itself as it can be done with tools such as SSH. So, encryption can be done with X. Millueradfa (talk) 23:10, 4 April 2012 (UTC)

The SSH tunnel feature will probably work with XDM, which is different from the port forwarding with the -L option. The port forward forwards a single TCP connection while an tunnel will create a link on top of the entire TCP/IP protocol stack including UDP can be run, using TUN devices. Millueradfa (talk) 14:54, 8 April 2012 (UTC)


The article talks about these and makes a reference to "software framework". But I still have no idea what a primitive is. The the references article doesn't mention the word. Kelly222 (talk) 08:01, 27 February 2012 (UTC)

undue focus given to Orca[edit]

That's a rather large paragraph which isn't reflected in the linked topic, and has the appearance of being mostly promotional rather than informative TEDickey (talk) 21:12, 2 April 2012 (UTC)

It is well justified as Orca is the main means to provide screen reading and magnification on the X Window System. Previously the article gave the impression that X had no accessibility mechanisms. This is not true! X11 has had accessibility features being implemented for it for years.

To mention Orca here is entirely appropriate since it is specifically designed to be used on X Window System and is the main way to obtain Accessibility on X. Most Linux distributions provide Orca through their installer, at no additional cost.

Since the paragraph lets people know that accessibility is available on X and provides information on the main mechanisms for accessibility on X, it is informative, and not promotional. It would be misleading to say that there is no accessibility on X when in fact there is and not to mention the components that provide the features.Millueradfa (talk) 15:09, 4 April 2012 (UTC)

I see (you're justifying the promotional aspect on the basic of "it should be promoted"). Perhaps you'd like to improve the linked topic (though promotionalism isn't an improvement), rather than biasing this topic? TEDickey (talk) 20:38, 4 April 2012 (UTC)

It seems like everything on wikipedia is a promotion. I don't think the paragraph is any more promotional than anything else on Wikipedia. So I do not accept your assessment of this. Millueradfa (talk) 14:59, 8 April 2012 (UTC)

It has the usual earmarks, and your responses aren't helpful. The sourcing is weak, the paragraph goes into a lot of detail that (if it were sourced properly) would belong in the Orca (assistive technology) topic, and does not have much to do with this topic. Take a closer look at Comparison of screen readers, noting that there's more than one application which would run in X and provides analogous functionality, contrary to your comment that "Orca is the main means", etc TEDickey (talk) 16:50, 8 April 2012 (UTC)

Reference implementation[edit]

Has X.Org Server and its anchestors (back up to X1) always been the reference implementation of the X Window System? --Abdull (talk) 12:33, 4 July 2012 (UTC)

X10 was the first published version. There are numerous comments about it being a reference implementation. TEDickey (talk) 14:37, 4 July 2012 (UTC)
Source? --Abdull (talk) 13:14, 20 July 2012 (UTC)
Was that regarding X10, or the applicability of "reference implementation" to X.Org and kindred? TEDickey (talk) 11:28, 21 July 2012 (UTC)
By the way, my recollection is that the term "reference implementation" came mainly from the X.Org people to legitimize their fork from XFree86, and that it was less-used during the 1990s. Demonstrating that would take more work than I'd put into a talk-page, though I probably have enough email and so forth if I chose to document it properly as an external page. TEDickey (talk) 11:31, 21 July 2012 (UTC)

x-windowS informal?[edit]

The term x-windows (plural) is not informal. It's incorrect.

Suggesting to the reader that this usage is acceptable is about as unfair to the reader as a suggestion that farting in a crowded elevator is likely to be welcomed. — Preceding unsigned comment added by (talk) 01:22, 31 January 2013 (UTC)

It's been there since the project started. There may or may not be a citable claim that the project claims it's incorrect, but no-one's managed to find one in ten years - David Gerard (talk) 08:32, 31 January 2013 (UTC)
As I recall, it was from one of the X newsgroup FAQ's. The point is that the topic here is referring to the existence of the usage, not its correctness TEDickey (talk) 09:48, 31 January 2013 (UTC)
Heh. This one? Note title :-D - David Gerard (talk) 10:42, 31 January 2013 (UTC)
yes, I noticed several occurrences. Seems that it's been updated. Google might find an old posting to which could be used as a source; however google's memory is selective and I frequently find it uncompliant TEDickey (talk) 11:42, 31 January 2013 (UTC)
A search on shows lots of people using it as if it means "X Window System". I mean, I remember "X Windows is just wrong!" being a standard assertion - I do know what the IP is on about. But it's like history started in 1995 and everything before then just vanished like morning dew in the hot sun. (I am appalled how much I get this feeling on Wikipedia.) - David Gerard (talk) 22:42, 31 January 2013 (UTC)
My feeling is that if it werent commonly (and incorrectly) referred to as "X Windows" then the org would not have put the assertion that that is incorrect in their faq or manpage. -- [UseTheCommandLine ~/talk] #_ 02:52, 6 March 2013 (UTC)
If you can find a cite as to incorrectness, that would be marvellous - David Gerard (talk) 13:03, 6 March 2013 (UTC)
Well, there's this rant from Matt Landau, who was the X Window System Manager for the X Consortium:
It is the "X Window System", or "X11", or "X" or any of
a number of other designations accepted by the X Consortium.  In fact,
doing "man X" on pretty much any X11 machine will tell you:

     The X Consortium requests that the following names  be  used
     when referring to this software:

                           X Window System
                             X Version 11
                     X Window System, Version 11

There is no such thing as "X Windows" or "X Window", despite the repeated
misuse of the forms by the trade rags.
So it's a pretty authoritative source, but the publication venue (a Usenet post from 1993) may not meet WP:RS. Alternatively you could take the not-unreasonable view that the list of names in the man page is meant to be exhaustive. —Psychonaut (talk) 15:44, 6 March 2013 (UTC)
I think who it's from makes it reliable enough for these purposes. Nice one! - David Gerard (talk) 18:16, 6 March 2013 (UTC)

Section updated. How's that? Someone official saying it's a "misuse", but the main newsgroup FAQ using it in its title - David Gerard (talk) 18:30, 6 March 2013 (UTC)

Fine, but I converted the citation to a standard {{cite newsgroup}} one so that readers have the option of using their preferred news client rather than only Google Groups. —Psychonaut (talk) 19:02, 6 March 2013 (UTC)

Sorry to kick off a hot topic and vanish, but things got very hectic shortly thereafter. A few thoughts with two basic themes, the intention of the original comment, and what's actually correct.

Though I did use the word "incorrect" in the original comment, the gist of the whole comment was really more about how the term "X-WindowS" is likely to be heard by others, incorrect or not. At times that term has been used as a rib, as pejorative, and it has at times been a yellow flag for incompetence. I suspect many folks who would hear it that way are now retired or getting close, but there are also probably plenty around who have made it to senior positions by now. So I'd be careful about using it in the same way I'd be careful about using a phrase that could be taken as an ethnic slur. You don't wanna walk into an interview for a freelance gig or a promotion and find yourself sitting across the desk from someone who was standing next you in the elevator when you said something s/he found offensive. So yeah, the "fart on a crowded elevator" was deliberate hyperbole to get your attention, but "informal" is a bit weak.

Personally, I suspect that the use of the trailing "s" is a bit like the mis-use of "literally", "beg the question", etc. They're all strictly speaking wrong, but they've entered common usage and we're probably stuck with them for a while. However, for many folks the misuse of those terms has the same visceral effect as nails scraping across a chalkboard, and the "X" thing can be worse. That may be the listener's problem, but if the listener is someone you want to keep happy then it's your problem too. Think of it as humoring your elders if you must.

As far as I've been able to tell "X-Window System" is the correct name of the windowing system and "X-Windows" is not. The use I've seen of citations purporting to illustrate that X-Windows is correct seem to involve misinterpretation of context. In some cases the use has been erroneous and later corrected. In others the term has been used as rib. For example Rob Pike used it in a talk in which his basic theme was that the UNIX community was failing to innovate in areas where Microsoft was. I have a feeling he was deliberately being provocative, but today's reader may not get it. (That's supposition on my part, subject to correction.) Newsgroups have hierarchial names. "comp.lang.c" is not the name of a programming language. "c" is a language in category "comp.lang". "" is a category pertaining to windowing systems. Sure there is a "". There is also a "", a "", a "", and others. "windows.x" is not the name of a windowing system any more than "" is. "x" and "news" are windowing systems in category "windows". (In the case of MS the name of the system is also the name of the category which may invite unwarranted but understandable extrapolation.)

There are contexts in which the term "X windows" can be correct. That is when referring to particular windows rather than the system. If you're using an X-Server on a Mac you may have four windows open at once, two "Aqua windows", and two "X windows". That might seem to imply that it's also the name of the windowing system but it doesn't any more than the Mac's native windowing system is called "Aqua Windows".

In any event, I suggest that in mixed company, and especially in job interviews you do not use the "N" word or more subtle slurs, that you do not say "literally" when you mean "figuratively" or "emphatically", that you do not say "beg the question" when you mean "raise the question", and you do not use "X-Windows" to refer to the the X-Window System.

I think that "informal" is too weak for now. In another decade maybe not.

Ericfluger (talk) 19:26, 31 May 2015 (UTC)

Linux graphics stack (sic)[edit]

The illustrations appear to be WP:OR, lacking any WP:RS. Some improvement would be helpful. TEDickey (talk) 15:10, 24 August 2013 (UTC)

Can you please leave the diagram? And try to resolve the original research and references issues? Maybe I'll have a go at finding sources for it myself. I have been looking for ages for such a diagram, I've immediately saved it on my harddrive. I would highly appreciate it if it can be kept for some time. :-) Paulus/laudaka Laudaka's talk page 07:06, 31 August 2013 (UTC)
Most of that editor's changes are WP:OR (essentially promotional). Judging by the amount, there may be a source from which it has been transplanted - that source might bear some discussion. TEDickey (talk) 10:11, 31 August 2013 (UTC)


Can some experienced user have a look at here? It's about one of the diagrams that used in this article. I'll be thankful. Bkouhi (talk) 17:08, 1 October 2013 (UTC)

I prefer the ancient version, at least it relates to reality slightly (possibly could do with updating, who uses up2date any more). @Tedickey: Which do you think is better? - David Gerard (talk) 18:21, 1 October 2013 (UTC)
The old one is preferable: the new one has multiple issues, e.g., (a) cluttered with non-topical items, (b) specific to one not-yet-available configuration, (c) some unsourced opinion thrown in for subjective weighting. TEDickey (talk) 18:29, 1 October 2013 (UTC)
Mm. I drew the original (way back when), I'll do a redrawn version this evening (checking each use across a zillion wikis first, of course) - David Gerard (talk) 18:50, 1 October 2013 (UTC)
Simple version back. I've also removed the other two impenetrable multicoloured blobs as unhelpful - David Gerard (talk) 20:12, 1 October 2013 (UTC)

Intro rewrite, prose fixes[edit]

There's something that happens to Wikipedia articles: people add subclauses and qualifiers all over the prose, typically with terrible grammar, let alone prose flow.

So I've just gone axe-crazy on the utterly impenetrable intro and am cleaning up the rest of the text, pruning carbuncles and encrustations. Any deep objections? - David Gerard (talk) 15:59, 3 October 2013 (UTC)

The joys of the unnoteworthy[edit]

So I just filed a {{PROD}} on Fresco (windowing system). Because there is little evidence anyone actually cared. I looked through the history to find the original author to notify, and it was ... me, nine years ago. I figure this talk page is the best place to find anyone who would actually care enough to add references - David Gerard (talk) 22:13, 15 October 2013 (UTC)

Doesn't seem to be much (aside from newsgroup-style comments) TEDickey (talk) 00:39, 16 October 2013 (UTC)

History of the core protocol[edit]

Hi guys, where can I find a history of the core protocol with changes between releases? Thanks. (talk) 03:30, 1 May 2014 (UTC) plus ChangeLog in source. GangofOne (talk) 20:18, 27 July 2017 (UTC)

What's a "bitmap display systems"?[edit]

(First sentence of the "History" section) Does it have a technical meaning or does it just mean what it sounds like: a system for displaying (raster) images instead of just text? AngusCA (talk) 13:41, 29 July 2014 (UTC)

Consulting a dictionary would help (rasters refer to parallel lines, while bitmaps are not limited in that fashion). TEDickey (talk) 20:18, 29 July 2014 (UTC)
Interesting... perhaps I've not seen enough bitmaps, but I think they'd be poorly suited for vector displays. AngusCA (talk) 19:04, 7 August 2014 (UTC)
An LCD display might be a suitable example. TEDickey (talk) 20:45, 7 August 2014 (UTC)

Free/Cross Platform in Title[edit]


I had this edit, in which I added the words free (because X being free software is one of its defining aspects), and Cross Platform (another key topic) reverted. I think it's a worthwhile thing to add, as these are differentiating factors from other window systems. What do we think about this? — Preceding unsigned comment added by Jakesyl

The suggested source isn't a reliable source, because (being authored anonymously) it falls short of the guideline:

Articles should be based on reliable, third-party, published sources with a reputation for fact-checking and accuracy.

Rather, it was equivalent to linking to Wikipedia. The link to cross-platform was to the "wrong" topic (though even if it had been accurate, that is overused, effectively vacuous even for Wikipedia). So there was no improvement suggested. TEDickey (talk) 23:51, 17 September 2015 (UTC)

Alright understood. Still, X being FOSS is a pretty big part of its philosophy... Jakesyl (talk) 20:01, 24 September 2015 (UTC)

There might be a useful source (neutral, knowledgeable, third-party), but the linfo one has many problems. TEDickey (talk) 23:59, 24 September 2015 (UTC)

External links modified[edit]

Hello fellow Wikipedians,

I have just modified 6 external links on X Window System. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:

When you have finished reviewing my changes, please set the checked parameter below to true or failed to let others know (documentation at {{Sourcecheck}}).

YesY An editor has reviewed this edit and fixed any errors that were found.

  • If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
  • If you found an error with any archives or the URLs themselves, you can fix them with this tool.

If you are unable to use these tools, you may set |needhelp=<your help request> on this template to request help from an experienced user. Please include details about your problem, to help other editors.

Cheers.—InternetArchiveBot (Report bug) 18:56, 20 July 2016 (UTC)

Xterms, rise and fall.[edit]

I recall all this from memory. Xterms (thin clients) became widespread during the late 80s in research organizations. They were typically in restricted areas of universities or labs, and not accessible for the entire student population because of their cost. (Wall Street exchanges were also large users of Sun unix/X systems because of speed and security.) The large, often 20", RGB monitors were particularly attractive to thieves. Workstations that booted these xterms cost $30k (USD, e.g. Sun) and upwards, while the xterms were in the vicinity of $5k. Many xterms, such as IBMs, utilized cheaper x86 hardware 2/3/486 CPUs and FPM memory, rather than their own faster hardware. While the xterm CPU was intended to take the 3D computational load, the CPUs were too slow and networks too slow to give the same graphical performance as full workstations. By the late 90s much cheaper PCs running full X systems had completely killed off even the cheapest proprietary hardware. The x86 hardware released at this time had overtaken nearly all other workstation hardware in terms of performance. Sub $1k PCs out-computed $100k deskside monsters. (talk) 02:53, 27 August 2016 (UTC)

"Sub $1k PCs out-computed $100k deskside monsters." Revisionism is nice but c'mon, this is going too far. In the late 90's desktop pc's were running P-IV's, while the $100,000 monsters were running 128 Itanics in a single system image, and so on. PC's did eventually take over where workstations left off but let's not slap ourselves on the back too hard ... (talk) 10:57, 27 July 2017 (UTC)

"A modern example..."[edit]

I think the image showing "a modern example" could do more justice to how far the KDE project has come and display the gorgeous design that is visible to users now. KDE 4 leaves the impression as if open-source GUI-s are still behind those of proprietary OS-s. Karland90 (talk) 09:01, 15 September 2016 (UTC)

Feel free to drop in a 2016 screenshot :-) - David Gerard (talk) 11:47, 15 September 2016 (UTC)