The X-Windows Disaster (1994)

http://www.art.net/~hopkins/Don/unix-haters/x-windows/disaster.html

by shp0ngle • 6 years ago

105 160 6 years ago

simonblack• 6 years ago

And yet, here we are, over a quarter-century later, and that "disaster" is still allowing our programs to run on one computer and display the result on another computer on the other side of the world.

The 'Unix Hater's Handbook' is supposedly a humorous work. But it was mainly used in the 1990s by non-Unix companies as anti-Unix propaganda. Unfortunately for them, all of today's main computer/software companies either use Unix or Unix-like operating systems (Apple:BSD Unix, Google:Android, IBM and Others:Linux, Microsoft:WSL). https://en.wikipedia.org/wiki/The_Unix-Haters_Handbook

cowmix• 6 years ago

As a someone who has used Unix for 35+ years, I feel the book is great. It actually does more to explain why things are the way they are in UNIX than any other book I've read, pro-Unix or not.. I don't agree with everything it covers and many of the complaints of the book have been addressed over the years, but it definitely still relevant today.

ajross• 6 years ago

To be fair: the UHH is a legitimately great book, a fun read, and absolutely something every modern developer should know. It's opinionated in exactly the right way, even if some of its conclusions turn out to have been wrong in hindsight.

But as for X11: yeah. You can complain all you want at microdesign issues, but longevity beats all. I can run an unmodified X11R1 client from 1987 against the default install of pretty much any newly installed Linux/BSD box and it still works, something not true of literally any other contemporary GUI toolkit of the era.

ddingus18• 6 years ago

While I was learning UNIX, the UHH was given to me as a rite of passage. Great book.

I learned a lot.

Love UNIX.

uk_programmer• 6 years ago

Generally I agree. However I would say with regards to iOS (not MacOS) and Android. While underneath they are both Linux and Darwin based respectively. Almost the entire OS anyone interacts (including most devs) is done through the Android and iOS APIs. They could replace what is underneath with something very different and you wouldn't be any of the wiser. There isn't really much to separate something like Ubuntu and MacOS in terms of what they are (other than polish).

In Jaron Laniers book "You are not a gadget" he says the open source community are very good at making clones of existing things but not inventing new things themselves. I would argue that programmers in general are very good at making clones of existing things. This has allowed Unix-like operating be ubiquitous as they are easy to copy and are good enough.

kick2• 6 years ago

It's not just a humorous work, it's also genuine criticism. dmr could find the humor in it, why can't you?

smhenderson• 6 years ago

dmr’s forward is my favorite part though. As others have said a lot of the criticism in the book has been addressed over the years. I came late to the party so I missed out on a lot of the early bugs and mis-features of UNIX.

gok3• 6 years ago

A lot of the criticisms have been addressed though. Basically no one outside of enthusiasts uses X any more; Android and Chrome OS use an entirely different graphics system. The filesystems have been replaced. Shells have been replaced and/or obviated for most users.

ddingus18• 6 years ago

Really, and this is an OK choice, but is a choice all the same:

What happened is others settled on a subset of multi user graphical computing.

That is not replacing X. It is something different, less.

Maybe less is more too. I have an open mind.

However that goes, X nailed multi user graphical computing. Pretty much any crazy thing you can think of can be done on X.

Want to make a computer that several people use locally, each with their own screen, keyboard, mouse? X can do that, and permit those users to view and interact with any combination of each others display data.

One says, hey can you sort this for me? Another says, "sure", and a window pops up on their screen...

Now, replicate that same setup, each user has their own powerful computer, maybe with a common, shared display among them, like a wall display or something.

No problem. X does that too, and chances are only the sysadmin differs. The apps will have no clue.

It may be that we just do not work that way, or maybe not enough know how, but let's be clear here.

X is not being replaced. A subset at best is being put forward.

I think that means we just are not really doing multi user graphical computing. And that is OK.

ddingus18• 6 years ago

Btw, I have shells, keyboards, mice, pens for my Android device and use all of them.

I think some of the X capability would be amazing on mobile.

We have hobbled mobile. Not completely as I know I can do really powerful things still, and when I do people gawk and point and usually ask a lot of questions.

I remain unconvinced this is all a good idea. We will see in a decade or two.

Maybe.

asveikau• 6 years ago

Someone should hand you a Nokia N900. A smartphone from 11 years ago, with specs that kind of sucked even for the time, but it ran X, and pretty well.

It's not some weakness that forces them not to use X, it is NIH. Same reason Google is writing a new kernel for no descernable reason.

mjcohen• 6 years ago

I learned a LOT when I read the UHH in the 90s.

ChuckMcM• 6 years ago

One of the fun things about working at Sun back in the day was that there were folks trying to fix some of the issues that were legitimately raised by the UHH.

That said, I shared Don's disgust with X. The difference between running X on a workstation and SunTools felt like a 300% slowdown.

And yet today I sit here typing into a window that is being displayed by an X server running on a Windows OS on a computer that is acting pretty much like an X-terminal for my Linux server. It has warts, true, but its still flying as they would say on Firefly.

I have often wondered what the world would be like if we had decided to move the GPU into the monitor and ran a parameterized display system over HDMI rather than something that looks like a cancerous NTSC composite video signal.

DonHopkins10• 6 years ago

One of the fun things about working at Sun back in the day was that there were folks trying to fix some of the issues that were legitimately raised by the UHH.

Like opening core files (which have gigantic gaps of unmapped zeros) by loading them into the XView text editor when you double click on them in the XView file manager? ;)

That's actually how core dumps grow and reproduce and evolve, like those wasps that lay their eggs in other insects and turn them into zombies, by embedding themselves in an unwitting host process and making it core dump even bigger!

https://www.wired.com/2014/10/absurd-creature-week-glyptapan...

And while on the topic of XView: Do you have any good XBUGTOOL horror stories, Chuck? ;)

The most foolish thing I ever did was try to use it to file a bug against itself. Boy do I regret that!

http://www.art.net/~hopkins/Don/unix-haters/x-windows/xbugto...

pjmlp10• 6 years ago

That disaster is the major reason why UNIX FOSS clones will never get desktop or mobile computing devices beyond a single digit number.

You can have all the fun you want with your beloved X, I rather use modern Windowing platforms and their composition engines.

Also, UNIX underpinnings are completly irrelevant on Android, iOS and ChromeOS.

In what concerns macOS and Windows, they are there to cater to the GNU/Linux crowd that buys Macs and Windows devices instead of buying GNU/Linux hardware.

IshKebab• 6 years ago

No it isn't. Almost nobody uses raw X over long distances because it is unusable. Everybody uses something else like VNC, RDP or NX. Note that NX actually moved away from the X protocol and is now just a VNC-like protocol, i.e. it only deals with pixels. Every modern remote desktop protocol works that way because it is better.

catblast5• 6 years ago

allowing our programs to run on one computer and display the result on another computer on the other side of the world.

I feel sorry for anyone having to use X over a link of that latency. Especially if their internet service has a tendency to go down. How many X clients handle loss of connection better than just terminating?

And are there really that many people using X remote as compared to RDP or Citrix. I would be surprised to find out that remote X is widely used in 2020 (if you can even call it’s brief heyday in the 90s while commercial UNIX was losing its balls to Windows NT due to vendor incompetence).. and I would risk the age discrimination suits and fire the IT team.

X was already an ugly obsolete piece of crap in the 90s when it was foisted upon us by all the major commercial Unix vendors. It should have rightfully died with SGI, Sun, and/or DEC.

catblast5• 6 years ago

But it was mainly used in the 1990s by non-Unix companies as anti-Unix propaganda.

Some strange revisionism there as few of the authors of UHH were involved with Windows.. which is what ended up completely eating commercial Unix lunch... a win that was not so much due to Microsoft’s competence as their competitors stupidity.

Meanwhile calling Mac OS X or Android Unix-like just because they have some tools in common (MAC OS X is only vaguely a BSD — and launchd seems to get some love by the same people that have a conniption fit over systems) .. none of those systems use X for sure. Desktop Linux is a rounding error.. the part of Linux that gets used heavily for servers that is the reason that WSL is a thing (to ease development)?? Yeah that has basically fuck all to do with the disasters in SySV/commercial Unix... not too many people using SySV IPC on Linux, and cgroups, epoll, io_uring hardly remind me of the miserable APIs of now ancient Unix.

jgon• 6 years ago

A disaster, and yet I remember reading about attempts to replace it when I was still in University, and now here I am over a decade into my career and Wayland still isn't ready for primetime, especially if you use a video card from the #1 producer of performance video cards on the planet. Small use case I guess. X lets you do forwarding over the network and maybe sometime in the 2020's Wayland will have some sort of remote desktop solution? Still waiting on that to be ironed out...

Replacing X feels like the nuclear fusion project of the Linux world, constantly receeding into the future even as it maintains its promises of fixing all our problem When It's Done(tm).

cycloptic7• 6 years ago

There are various solutions already available for remoting in Wayland.

GNOME: https://wiki.gnome.org/Projects/Mutter/RemoteDesktop

KDE: https://userbase.kde.org/Krfb

Sway/wlroots: https://github.com/any1/wayvnc

Generic: https://gitlab.freedesktop.org/mstoeckl/waypipe/

Web-Based: https://github.com/udevbe/greenfield

You can also still use your X11 programs remotely within Wayland, by using XWayland, which is supported by all the major implementations.

Unrelated to Wayland, there is also backend support in GTK3/4 for rendering to a web socket: https://developer.gnome.org/gtk3/stable/gtk-broadway.html

ricktdotorg• 6 years ago

from this page:

GNOME: https://wiki.gnome.org/Projects/Mutter/RemoteDesktop

"Currently only remote passwordless unencrypted VNC access to an existing session is supported."

i guess that is technically a solution, but not a good one.

kick2• 6 years ago

You can use NVIDIA cards with Wayland, you just can't use NVIDIA's proprietary driver with wlroots (which is by no means the only Wayland compositor). Wayland is the default in many distributions, and if you're using one of the more common desktop environments, you'll get by just fine when using the proprietary driver.

mshook• 6 years ago

It's X-Window (System), no S.

warpech• 6 years ago

It says it the footnotes that the misspelling is intentional

s_gourichon3• 6 years ago

The text of the article actually has it correct most of the time. The name is just "X". It's a windowing system that happens to be named "X".

It's named this way because it is a successor to another windowing system named "W" and "X" comes after "W" in alphabet order.

https://en.wikipedia.org/wiki/X_Window_System#Predecessors

ptx3• 6 years ago

Exactly. Many people seem to read it as "the system named X Window", but, as you say, it's actually "the window system named X". The Linux Operating System is an operating system and the X Window System is a window system.

Maybe part of the reason is the Title Case convention in English? (And the other part may be the name of Microsoft's window system leading the mind along this path.)

DonHopkins10• 6 years ago

Don wrote the chapter on the X-Windows Disaster. (To annoy X fanatics, Don specifically asked that we include the hyphen after the letter "X,", as well as the plural of the word "Windows," in his chapter title.

I published a better version on Medium that has a few typos fixed, and includes a flyer that was handed out at the X-Windows conference, and an article I found in my archives from Volume 1 Number 2, of the June 1988 of the “XNextEvent” newsletter, “The Official Newsletter of XUG, the X User’s Group”:

https://medium.com/@donhopkins/the-x-windows-disaster-128d39...

Official Dangerous Virus Notice Distributed at the X-Windows Conference

Official Notice

Post Immediately

Dangerous Virus!

First, a little history. The X window system escaped from Project Athena at MIT where it was being held in isolation. When notified, MIT stated publicly that “MIT assumes no responsibility…”. This is a very disturbing statement. It then infiltrated Digital Equipment Corporation, where it has since corrupted the technical judgement of this organization.

After sabotaging Digital Equipment Corporation, a sinister X consortium was created to find a way to use X as part of a plan to dominate and control interactive window systems. X windows is sometimes distributed by this secret consortium free of charge to unsuspecting victims. The destructive cost of X cannot even be guessed. X is truly obese — whether it’s mutilating your hard disk or actively infesting your system, you can be sure it’s up to no good. Innocent users need to be protected from this dangerous virus. Even as you read this, the X source distribution and the executable environment is being maintained on hundreds of computers, maybe even your own.

Digital Equipment Corporation is already shipping machines that carry this dreaded infestation. It must be destroyed.

This is what happens when software with good intentions goes bad. It victimizes innocent users by distorting their perception of what is and what is not good software. This malignant window system must be destroyed. Ultimately DEC and MIT must be held accountable for this heinous software crime, brought to justice, and made to pay for a software cleanup. Until DEC and MIT answer to these charges, they both should be assumed to be protecting dangerous software criminals.

Don’t be fooled! Just say no to X.

X-Windows: …A mistake carried out to perfection. X-Windows: …Dissatisfaction guaranteed. X-Windows: …Don’t get frustrated without it. X-Windows: …Even your dog won’t like it. X-Windows: …Flaky and built to stay that way. X-Windows: …Complex non-solutions to simple non-problems. X-Windows: …Flawed beyond belief. X-Windows: …Form follows malfunction. X-Windows: …Garbage at your fingertips. X-Windows: …Ignorance is our most important resource. X-Windows: …It could be worse, but it’ll take time. X-Windows: …It could happen to you. X-Windows: …Japan’s secret weapon. X-Windows: …Let it get in your way. X-Windows: …Live the nightmare. X-Windows: …More than enough rope. X-Windows: …Never had it, never will. X-Windows: …No hardware is safe. X-Windows: …Power tools for power fools. X-Windows: …Putting new limits on productivity. X-Windows: …Simplicity made complex. X-Windows: …The cutting edge of obsolescence. X-Windows: …The art of incompetence. X-Windows: …The defacto substandard. X-Windows: …The first fully modular software disaster. X-Windows: …The joke that kills. X-Windows: …The problem for your problem. X-Windows: …There’s got to be a better way. X-Windows: …Warn your friends about it. X-Windows: …You’d better sit down. X-Windows: …You’ll envy the dead.

----

Things That Happen When You Say ‘X Windows’

I was digging through some old papers, and ran across a 15 year old “XNextEvent” newsletter, “The Official Newsletter of XUG, the X User’s Group”, Volume 1 Number 2, from June 1988. Here’s an article that illustrates how far the usage of the term “X Windows” has evolved over the past 15 years. (Too bad The Window System Improperly Known as X Windows itself hasn’t evolved.)

Someone on slashdot asks, “Why is it still called X-Windows?”. Predictably, the first reply says: “It isn’t. It’s called ‘The X Window System.’ Or simply ‘X’. ‘X Windows’ is a misnomer.”

He didn’t ask why it is “X-Windows”. He asked why it’s called “X-Windows”. You’re wrong that it isn’t called “X-Windows”. It is! It’s just that it isn’t “X-Windows”. Being something is independent of being called something.

The answer to the question ‘Why is it still called X-Windows?’ is: It’s still called X-Windows in order to annoy the X-Windows Fanatics, who take it upon themselves to correct you every time you call it X-Windows. That’s why it’s called X-Windows.

The following definitive guide to the consequences of saying “X Windows” is from the June 1988 “XNextEvent” newsletter, “The Official Newsletter of XUG, the X User’s Group”, Volume 1 Number 2:

Things That Happen When You Say ‘X Windows’

THE OFFICIAL NAMES

The official names of the software described herein are:

X Window System

X Version 11

X Window System, Version 11

X11

Note that the phrases X.11, X-11, X Windows or any permutation thereof, are explicitly excluded from this list and should not be used to describe the X Window System (window system should be thought of as one word).

The above should be enough to scare anyone into using the proper terminology, but sadly enough, it’s not. Recently, certain people, lacking sufficient motivation to change their speech patterns, have fallen victim to various ‘accidents’, or ‘misfortune’. I’ve compiled a short list of happenings, some of which I have witnessed, others which remain heresay. I’m not claiming any direct connection between their speech habits and the reported incidents, but you be the judge… And woe betide any who set the cursed phrase into print!

You are forced to explain toolkit programming to X neophytes.

Bob Schiefler says, “You should know better than that!”

The Power Supply (and unknown boards) on your workstation mysteriously give up the ghost.

Ditto for the controller board for the disk on your new Sun.

Your hair falls out.

xmh refuses to come up in a useful size, no matter what you fiddle.

You inexplicitly lose both of your complete Ultrix Doc sets.

R2 won’t build.

Bob Schiefler says “Type ‘man X’”.

Your nifty new X screen saver just won’t go away.

The window you’re working in loses input focus. Permanently.

gok3• 6 years ago

like Sun's Open Look clock tool, which gobbles up 1.4 megabytes of real memory

Meanwhile the clock app on my phone is currently using 12 MB. A popular time telling website uses over 100MB on my laptop.

tinus_hn4• 6 years ago

Is it

http://whattimeisitrightnow.com/

gok3• 6 years ago

That was what I looked at yeah, and I appreciate that it's actually a Netflix gag site for Bojack Horseman but still.

donio3• 6 years ago

It's like the Eight Megabytes And Constantly Swapping joke for Emacs. These days a minimal Emacs is more like 18MB RSS rather than 8 but that's still tiny compared to modern apps.

speedgoose• 6 years ago

I don't know about your phone but mine which is not new has 6gb of ram and the latest Samsung s20 flagship has 12gb of ram. In comparison to the total ram, it's much better nowadays.

0xff00ffee2• 6 years ago

I bet 99% of that 12MB is graphical assets: different sized icons for different display geometries. I always found asset management for multiple platforms to be most annoying when developing for mobile.

This reminds me of when people started to realize Microsoft Foundation Class (MFC) was blowing up and were bemoaning code-bloat. In 1997.

hermitdev• 6 years ago

The very mention of MFC gave me shivers down my spine. The macro horror show and indecipherable error messages...

Thankfully, I never had to do much MFC professionally, but holy shit, if you inadvertently messed with one of the generated macros, good luck. I remember it sometimes being easier to throw away everything and start from scratch than debug the horrible compiler errors.

commandlinefan• 6 years ago

When I first saw X, I had already grown up with PCs, so the whole concept of running the “main program” on another computer just to display it on my computer seemed so illogical I thought I must just be misunderstanding it. I mean.. I have a computer right here, why not run the program on this computer…? However, if you think about it - the X philosophy was actually mostly correct (or at least more useful than it initially seemed), they just did it in a clunky way. We’ve been slowly migrating to web applications which are the same basic concept: the program runs on the web server and your browser displays the results. This actually is better from a security standpoint (the server can access data that isn’t distributed to the client) and a maintenance standpoint (I can deploy a new version to the server and it’s instantaneously available to all clients). Running a “program” remotely and displaying the results locally is so ubiquitous now that my teenage kids are confused by the entire concept of desktop applications. What I’d love to see, and wish I had the free time to implement, is a Linux UI based on something like SVGALib that doesn’t use X at all; just enough to run a terminal and a browser (and maybe an email client and a word processor if somebody wanted to write one).

s_gourichon3• 6 years ago

The modern equivalent of SVGALib is Linux' Direct Rendering Infrastructure.

If you just want fullscreen app, you can run them on top of DRI.

For "a terminal and a browser" at the same time you have to create some sort of protocol for windowed display, and an implementation of one side of that protocol that provides window management, and apps would implement the other side.

People did that. They named the protocole "Wayland", the window management implementation "Weston", and a number of graphical toolkits can target Wayland (GTK, QT, SDL...). Ref: https://wayland.freedesktop.org/faq.html#heading_toc_j_2

As a sidenote, if you want to run such a program standalone, fullscreen, can you do that without running a "server" or a "window manager"? IIRC, programs intended for Wayland can run fullscreen on top of plain DRI (or perhaps only SDL provides this facility).

cycloptic7• 6 years ago

The simplest thing I know of is the minimal Cage compositor for wayland, that only supports running a single fullscreen program: https://github.com/Hjdskes/cage

With some hackery, you could probably build this as a backend for GTK/Qt that would run the server in the same process. But I don't know how useful that would really be.

trilinearnz• 6 years ago

I'd like to dig into this point for a moment. Why was it necessary to have a client/server architecture for a windowing system? Was it because X was developed before practical GUI alternatives were on the market? Did the computing power of institutional mainframes exceed that of the user-facing terminal systems (clients)? As time passed and "client-only" windowing systems came onto the scene, could the client/server design choice not have been revisited for greater efficiency?

brianpaul• 6 years ago

I think there's kind of two aspects to that (why client/server):

1. The X protocol was roughly like an evolution of the VT-like protocols which were used for text-based terminals back in the day. Those protocols supported cursor control, simple character-based graphics, and later, simple vector graphics. X was an evolution of that to deal with bitmapped displays, mice, etc.

2. Hardware was pretty expensive. Giving everyone in the office an X terminal was far cheaper than giving everyone a workstation. Again, that was an evolution of VT terminals vs. minicomputers.

Even X terminals weren't cheap. In the early 90's a monochrome 15" terminal could be a couple thousand dollars while the workstation was at least 10x more.

orbifold• 6 years ago

The classical use case is EDA software, which still largely runs on Linux. I can run simvision (a cadence tool) on a large server with ~96 cores and ~1TB of ram and the spawned windows integrate (almost) seamlessly with my multi monitor MacBook Pro setup. That Software only supports Solaris (recent releases dropped that I believe), specific versions of Red Hat Linux and Windows. Large simulations easily consume all server resources and require a SSD raid to function effectively. This also isn’t software you want to have to install or maintain yourself, plus it need access to proprietary information protected by NDA, which is not allowed to be copied to private hardware.

pjmlp10• 6 years ago

Xerox graphical workstations, proprietary UNIX toolkits like NeWS....

ddingus18• 6 years ago

X is kind of amazing.

Once, just for fun, I had an SGI do window management, another one serve up fonts, a linux box serving apps via NFS, and yet another headless SGI running a shared app displayed on yet another Linux machines X window server session.

It was surreal. Launch that app, and all those machines contributed to the interaction of it.

Plus, it made that instance of RH Linux look like an SGI. 4dwm has a distinct look and feel.

For the most part, running over a 100T network, one could not tell all that was happening. (Was a high end CAD app.)

Later, I did run that same app remotely, using its Windows NT build and the Exceed X server for Win NT. Cool.

kevin_thibedeau• 6 years ago

I like being able to run a Linux desktop in a VM on windows and "extend" to another monitor with VM X apps rendering to a Cygwin Xorg instance. Wayland can't do that.

ddingus18• 6 years ago

Yup. In my humble opinion, the people who built X really did think through what multi user graphical computing meant.

I like X. Took me a while to understand it, but once I did, I used it hard and appreciated what it allowed me to do.

Doing stuff like dual or multi head computers. Multi users each with their own keyboard and mouse can happen with X. And if one needs another user, doing it via network just works.

X is a powerful thing.

cycloptic7• 6 years ago

There is no reason wayland couldn't do that, other than that your compositor doesn't support the concept of virtual screens. I can't say that this is a very high-priority feature for any compositors at the moment.

the8472• 6 years ago

The compositor would have to talk to the virtual GPU driver to allocate additional surfaces on the host and make them available to the guest.

The next best alternative currently is looking glass[0]. Linux host with GPU1, Windows guest with GPU2 via pass-through and then copy the framebuffer from the guest to the host.

[0] https://looking-glass.hostfission.com/

Gnurou• 6 years ago

I like being able to run a Linux desktop in a VM on windows and "extend" to another monitor with VM X apps rendering to a Cygwin Xorg instance. Wayland can't do that.

This should be possible with virtio-wayland:

https://alyssa.is/using-virtio-wl/

It's pretty new and not used outside of Chrome OS (to run Linux apps) at the moment though.

technofiend2• 6 years ago

DEC or HP (I no longer remember which) wrote a cool demo that had a fish swim across as many displays as you gave it. It properly rendered the fish split across display edges. With walls of displays being so common now it doesn't sound like much but at the time it was a pretty slick demo of X window and its network capability.

ddingus18• 6 years ago

That's cool

LargoLasskhyfv• 6 years ago

Sounds like http://dmx.sourceforge.net/ which is part of X.Org since release https://www.x.org/wiki/Releases/6.8/ from 2004.

donio3• 6 years ago

That chapter was dumb even when it came out. X is awesome and it's amazing how well it has stood the test of time. I am not looking forward to the day when I won't be able to do half the things I do today with my EXWM setup because we needed to have "every frame perfect".

ddingus18• 6 years ago

It is like UNIX.

The people who created X thought multi user graphical computing all the way through.

We may decide that should not be a thing.

I think we will struggle, some time will pass, and then it will be a thing, meaning X will just get reinvented.

catblast5• 6 years ago

Agree to disagree then. X is shit. I remember first reading that chapter in 1998, and it resonated then just as well as it does now. X is shit. It is the weirdest abstraction of network "transparency" which inevitably manages to require state on both ends of the network connection -- it isn't robust. How well does your X session do when there is a network hiccup?

The protocol is round-trip heavy. Even with xcb, X clients need to make a tremendous number of round-trips, especially early on. There have been mitigations for X for low bandwidth links, but X networking has never been usable on high latency links. Ever forward X on an airplane? I pity you.

The drawing model of X is an antiquated relic. Server side rendering is pretty much useless, such that X has basically become a sort of bastardized inefficient VNC.

There's a reason no widely used screen remoting or thin-client system uses the X design. It was a poor idea then, and it is still deficient today.

glangdale2• 6 years ago

Agreed. And I may be showing my age, but I learned C and C++ on singles MIPS boxes (R3000 processors) shared between 20 users, all on X terminals. The thought of sharing 1 32-bit processor running at 25Mhz between 20 people running any kind of graphical environment it pretty wild by today's standards.

ddingus18• 6 years ago

I am just gonna ramble for a minute. Plz either indulge or ignore. I miss this stuff! Thanks.

I would have loved to profile that little system. It's crazy how lean those older systems really were.

A 30Mhz R4K could play MP3 files streamed via NFS. (90 percent CPU utilization, not headless, with local desktop logged in and active) I had one, at that clock useless by the time I got it, doing just that, because why not? After reading your comment, I realize I had a machine just playing a compressed audio file with the resources needed to serve 20 of you a useful computing environment!

What kind of network were you on? Was it even ethernet? 10T?

My first multi-user experience was horrible! It was on a 386/25, running Xenix and it had capacity for 16 users via either 19200 or 9600 baud SERIAL and VT100 terminals... I had 12 users at peak on that one, and most of them ran multi-user WordPerfect and some MRP thing, Symix, I think.

I got the job because I read all the manuals. First sysadmin post and it was a pig. Learned a ton though! Enough to realize I could have a lot of fun on the better UNIXes, and scored a position working with SGI computers.

(in that timeframe those computers with all the spiffy software so well engineered were like computing from the future, so much fun)

Another time, while doing sysadmin on an SGI Irix multi-cpu Origin system serving some 30 users doing high end cad, I did some data logging.

That system had 16GB Ram, 4 R12K CPU's with super fat caches, dual homed 1000T ethernet, BTW. I forget the clock. Probably 400Mhz

I logged user activity every few seconds for roughly a week. This was an effort to see whether overall system speed was impacting productivity. I had a few users grumble during a peak time, so...

Learned a few things:

The system itself could perform near max capability when there were 4 concurrent user CAD demands. OS overhead was a few percent tops, in almost all circumstances. This was not in of itself any kind of issue. Irix ran pretty damn lean.

Once in a blue moon, the system would page, but the Irix file caches are gold. This never really was an issue either. Maybe for a few minutes at a time during a serious push tops.

A few times per week, usually on a Friday, they would max it out, and more than 4 concurrent, high demand sessions would be present. Deffo slowdown.

Among the users, the average full compute demand was roughly 2 minutes. Some instances of this were longer for power users working on more substantial models. 5 to 10 minutes. These were sporadic, during the day. The reason was simple, and it was a model recompute. Happens as users roll up all their changes.

Interestingly, light duty users, doing detail drawings, or minor model updates, sustaining engineering type stuff, they used the system heavy, perhaps 60 percent of their workday, but rarely hit peak single core compute for more than a few seconds at a time. Was hard to even see these. They did not really do big recomputes, but they did do the occasional big drawing render, and those were as bad as the power users. But not even daily.

Most users hit the system both harder and less frequently. 20 to 30 percent of their day, but did post up regular ~2minute single core max compute demands. These people were working on sub systems, and would often copy one, modify, update, check it in, done, next.

The power users hit the ~5 to 10 minute peak compute and used the system half the day. They were working at the entire product level. Big demands on RAM, CPU, Network. Multiple times daily.

What I did was put local compute resources in front of the power users and restructure the network to give them a quieter, higher throughput subnet.

There were a few of these users. They did became network bound, as the data transfer was significant compared to everyone else running on the main server, with high throughput disks. (15K RPM Cheetah drives) This impacted their get going times, but they only did that occasionally. Once things got cached locally, non issue overall. They adjusted how they work and marginalized it. The biggest saving was honestly to just keep it logged in, screen locked, etc...

The rest of the users reported a significant speed up after these things were done.

Now, here's the thing!

A second data logging actually did show the power users improve. And they responded by burying those boxes, opening up a second session on the main one to continue. LOL, fine. It was good enough before, so I said nothing. They were power users and working at the product level and gains there paid off nicely in terms of total project capacity per year. No worries. There to help, blah, blah. But I did chuckle about them very rapidly figuring out how things worked.

(They used the excellent virtual desktop feature in the 4dwm window manager common to SGI computers. One click and they were viewing their local session, one more click and they were viewing their remote one, next.)

But everyone else demonstrated no real change!

The data just was not there to support a big gain, and it totally was not there after the power users did their thing by running two sessions anyway.

So, what was it?

Turned out to be network. When I put the local compute where it needed to be, we moved some switches and had inadvertently improved load balancing at the physical layer. Some users, prior to this, were basically blocking others.

Latency went down for everyone, and average throughput went up, and that proved to be the single most gratifying improvement for them, and it was just faster interactive response. No effective change in actual compute / time unit available, nor demanded from them.

The parallels today are obvious. Get that page response time down. Effective throughput, what users can get done may bump up a little, but the vast majority won't do it.

But, their feel, perception will ramp right up to the positive.

This is true even when it comes at the expense of other things maybe being slower, taking longer. Just get them feedback NOW, and they will feel better about it all anyway.

downerending2• 6 years ago

At the time when X was first being developed, there simply were no open source alternatives. I used a few of the closed source ones and they were nice enough, but generally available on only a single platform (e.g., HP), which made them fairly useless.

Even if Wayland succeeds, it'll be a long time before it has truly replaced X. In reality, I think it's more likely that both will be (or already have been) been run over by web apps.

ahartmetz• 6 years ago

That's like saying that CPUs will be run over by Linux.

code_duck5• 6 years ago

The browser needs to be displayed somehow.

pjmlp10• 6 years ago

Yes, and as proven by ChromeOS, X is not a requirement.

DonHopkins10• 6 years ago

The window system needs to be displayed somehow too. What works for the window system works for the browser too. There's no need for an additional layer of widow system once you have a browser that can draw on the screen itself. A web browser can make a much better scriptable window manager / desktop environment all by itself, than anything that's possible with X-Windows.

andyjpb3• 6 years ago

The UNIX Haters Handbook was funny at the time but doesn't make much sense these days unless you understand all the rival hacker groups and cultures that were largely eventually unified under "Free Software" and, later, "Open Source". Even the LISPers have come in from the cold!

Much of this criticism of X was prescient for things to come. Autotools, dependency problems, modularity issues, compatibility, "DLL-hell", etc. Many of these problems don't have solutions today that are orders of magnitude better than then.

On the other hand, one thing that's really holding X back, especially on handhelds, is 'Myth: X is "Device Independent"'.

If the difference between a 75dpi screen and a 100dpi workstation screen was too much, today's "retina" displays really struggle to run a usable X. You're heavily reliant on the toolkits (Qt, GTK, etc) to scale things for you and they mostly understand about font scaling but not about widget scaling.

Lots of distributions that might want to use it, for example Maemo Leste, Ubuntu Touch or Gemian, are having to do a lot of UI work and hand-pick applications for "porting" to the screens and input devices (touch) that they want to use them on.

This means that Linux-on-the-smartphone is not yet in any position to give Android or iOS a run for it's money because it really struggles to effectively leverage its existing app ecosystem.

Even ignoring security models, X struggles to provide even a compelling visual of a vision for handheld devices.

Don't even mention Wayland: that has other economic incentives that are not yet properly aligned for widespread success (including X's app ecosystem ones).

jcelerier3• 6 years ago

If the difference between a 75dpi screen and a 100dpi workstation screen was too much, today's "retina" displays really struggle to run a usable X.

I've been running X11 on retina displays since 2014 without much of an issue. The only big offenders were chrome and firefox. Nowadays everything just works.

fsh3• 6 years ago

It works if you have more or less the same pixel density across all screens. As soon as you mix very different ones, X becomes essentially unusable. Microsoft and Apple have solved that issue years ago.

andyjpb3• 6 years ago

I'd be interested to hear what you are using.

Obviously the web browsers are a big requirement.

...but I've had trouble with the defaults for almost everything I've tried: xterm, libreoffice, browsers. Under KDE the menus are too small by default. Under non-Gnome-or-KDE window managers or desktop environments almost everything is rendered too small.

The chances are that I can't just install an application with `apt-get` and have it do something sensible by default without a lot of fiddling and customisation.

ddingus18• 6 years ago

Here is one advantage of X that was notable, even today:

Keeping user data out of user hands.

A high end CAD application I used to do sales, service, sysadmin for used X and the SUID permission bit in UNIX to accomplish keeping data out of users hands.

On my bigger systems, I would run one application image, shared by a lot of users. This app, when run by a user, had the data permission and "ran as" data admin. The user has no permission.

All access to the data was through the app. And the users had no access to the app or data, just a remote X session.

I had tens of users interacting on big designs that way. No worries.

Secondly, I could where needed, spawn a second copy of the app to run local to be served up to or even just ran directly by needy users and everything still worked fine. I just had the app run as a different user and the access barrier remained intact.

But the big win was one application, one data repository, many users via X. I made a scripted login for them. They would launch it, do their thing, and that's it.

All I needed was a respectable X server for whatever box they ran locally.

I could admin that setup, and did, on a free JUNO dialup from all over the country.

This was the 90's through early 00's.

tinus_hn4• 6 years ago

This kind of security is more like a suggestion. It only works without an adversary.

X and the libraries used on it are not suitable for setuid programs.

ddingus18• 6 years ago

Context matters. The notable part today isn't hacker protection. I'm sorry I implied that.

It's user error / protection. And at the time, getting that on anything that required data move around? Horrors!

We've got that today, though it took an amazing amount of thrashing about to get where I was then. And it's more secure in the adversarial sense too. Still, a lot of data has to move around, and it's still a PITA when it does.

The better tools today are rapidly converging on that same basic model: App serve to users through display only protocols.

_ph_2• 6 years ago

I have been using X-Windows in that timeframe - on VMS workstations. The biggest problem was indeed, that X-Windows mostly was a place to run xterms on. But there were a few native GUI applicatons and they were quite nice. The biggest problem was, that Motiv wasn't freely available. You had to pay license costs. If not for that, "Unix on the desktop" might have actually happened. And as the time progressed, Motiv wasn't unbearably slow any more.

donio3• 6 years ago

I remember those days too, the lack of a good GUI toolkit felt like such a big deal. Funny that nowadays I find myself using applications that don't use a UI toolkit, Emacs (compiled without toolkit), st, mupdf, games that do their own UI. The only application I have running that uses GTK is chromium.

_ph_2• 6 years ago

The point I was trying to make was, that there were too few gui applications at all, most of the application I was running were just terminal applications. A widely available GUI toolkit might have helped this. The funniest example was a chemical database software, which would be VT100 based running inside an xterm, but for graphical input/output would spawn a separate window based on X toolkit graphics. But beyond that, those expensive workstations (like $20k in todays money) were mostly running xterm. While being technically widely superior to the PCs of that time - there were also Win3.1 machines around, running a dreadful VT100 emulation - they made little out of it, because there was so little software. So they were eventually replaced with PCs at a fraction of the price.

Ironically, even if it were the times of Linux 1.0, I suggested they should run Linux, as they would get excellent X Windows support, unfortunately the suggestion wasn't taken upon.

DonHopkins10• 6 years ago

Thank you for calling it X-Windows! ;)

DrScump• 6 years ago

I remember it being stressed to me:

"don't call it 'X-Windows'. It is 'X', or 'The X Window System', if you must."

This was in an environment where upper management wanted to take away the desktop workstations from non-developers in favor of X terminals. "We move all that work to the server now; You don't need all that expense on all these desks. Server-side is the future."

Good freaking luck if you were Tech Support or QA.

The fun part: we were spending as much on the X terminals and supporting equipment as we could get more Sparcstations for at our great corporate rate (my desktop was on its third generation by then).

Unsurprisingly, this was merely an early checkpoint in what became a tailspin.

XMPPwocky2• 6 years ago

"The right graphical client/server model is to have an extensible server. Application programs on remote machines can download their own special extension on demand and share libraries in the server. Downloaded code can draw windows, track input eents, provide fast interactive feedback, and minimize network traffic by communicating with the application using a dynamic, high-level protocol."

Huh, that's surprisingly prescient.

Applications on remote machines can download their own special JavaScript on demand and share code in the server. Downloaded code can draw windows, track input events, provide fast interactive feedback, and minimize network traffic by communicating with the application using HTTP.

ddingus18• 6 years ago

Interesting parallel there.

gopalv• 6 years ago

Applications on remote machines can download their own special JavaScript

Depending on what era we imagine writing this, this could be a Java applet or a Flash application or an ActiveX component.

XMPPwocky2• 6 years ago

Yeah, but "SPAs" based on Java/Flash/ActiveX weren't nearly as ubiquitous as web-based SPAs are now, were they?

DonHopkins10• 6 years ago

https://en.wikipedia.org/wiki/NeWS

NeWS was architecturally similar to what is now called AJAX, except that NeWS coherently:

- used PostScript code instead of JavaScript for programming.

- used PostScript graphics instead of DHTML and CSS for rendering.

- used PostScript data instead of XML and JSON for data representation.

anthk3• 6 years ago

PS was amazing, I've even seen a Z Machine V3 interpreter at rec.games.int-fiction. Enough to run Zork and old versions of Curses!. But writting the interpreter is easier than finishing the last mentioned game.

kkotak• 6 years ago

I think first successful implementation of Client-Server OS (and wide adoption in the banking industry) was BTOS/CTOS that was developed by Burroughs Technologies (BTOS) and then continued by Convergent Technologies (CTOS). It was sweet to works with - with modular Hardware and built in debugger on the CLI!

dang2• 6 years ago

By Don Hopkins!

https://news.ycombinator.com/user?id=DonHopkins

dang2• 6 years ago

A thread from 2017: https://news.ycombinator.com/item?id=15035419

0xff00ffee2• 6 years ago

I never thought about it from this person's perspective because I never had to program X beyond the graphics and OS classes I took in the early 90's. (Motif's insane verbosity makes code look awful.)

I've been using X + MWM since 1991 (AIX and Solaris, anyone?). It is still everywhere, and I wonder how it managed to persist for so long.

Why didn't anything ever disrupt Xorg?

anthk3• 6 years ago

TBH, Using TCL/TK as a Motif replacement for RAD tools don't look inside, a Pentium3 back in the day was fast enough for TCL 8.4 and 8.5.

DonHopkins10• 6 years ago

MOTIF ANGST PAGE

If you have any ANGSTFUL Motif code, comments, documentation, or resources, please share them with me! Here is some of the stronger stuff I've found. Note: this is only for official Open Software Foundation Motif inspired Angst. If you're experiencing TCL/Tk That Only Looks Like Motif But Doesn't Suck Angst, then you should stop whining and fix the problem yourself, if somebody else hasn't already.

http://art.net/~hopkins/Don/unix-haters/x-windows/motif.html

jasoneckert• 6 years ago

The Unix hater's handbook was never intended to be taken seriously.

It was written by Unix gurus that have the quirky hacker sense of humour. It's just a Unix roast.

X was always awesome.

Y, on the other hand, well.....

m463• 6 years ago

I have so agreed with this for years.

Not setting policy is what ALL of unix does... and then everyone is fumbling around for years trying to find a policy that works for them.

I'm reminded of that Lilly Tomlin quote: "I always wanted to be somebody, but now I realize I should have been more specific."

Izkata3• 6 years ago

For some perverse reason that's better left to the imagination, X insists on calling the program running on the remote machine "the client." This program displays its windows on the "window server."

Um, what? By your own description, X's terminology is correct and straightforward:

The idea was to allow a program, called a client, to run on one computer and allow it to display on another computer that was running a special program called a window server.

The "service being provided" is "display a window", that the client connects to. There is no inversion here.

stormbrew3• 6 years ago

The inversion is at a higher level of the conceptual stack, where a user generally thinks of their own computer as a client in all things. When they connect out to another machine, that's the server.

So if you ssh to another machine (a server) and then use a program on it that now displays on your machine, the direction of the connection is absolutely "you're the server, they're the client" but it inverts the user's intuition of what is in which role.

salawat2• 6 years ago

You're not looking at it right.

Imagine the X server running on your local machine as a server providing access to your eyeballs.

That's why the remote machine is the client; it is a subscriber to the server providing access to your screen.

kabdib• 6 years ago

I was in a used bookstore a few years ago, and ran across a book titled something like "User Interface Design with MOTIF". Well, it had been a power of years since I last touched X-Windows (with intent to draw pixels, anyway) and I started leafing through it.

Looking for user interface examples.

Leaf, leaf, nothing.

Flip, flip, flipflipflip . . . oh, we're at the index. Nothing.

Over 300 pages and no pictures. Zero diagrams. No user interface principles or discussion about good or bad practices. It was all descriptions of APIs. There was nothing about user interface design in there at all.

IMHO, that's all you really need to know about X-Windows.

pengaru• 6 years ago

IMHO, that's all you really need to know about X-Windows.

That there was at least one poor, misleadingly titled book about MOTIF?

I've had many X-related textbooks over the years, and frankly the documentation was exceptionally good and felt like a total waste on something as byzantine as X.

anthk3• 6 years ago

OpenBSD has one book about Motif in in the books sectin of ports. book/vol6a I think.

pjmlp10• 6 years ago

Steve Jobs had a couple of interesting remarks regarding UNIX's culture lack of understanding of UI/UX.

The only exceptions are NeWS and NeXTSTEP (OS X derived from it anyway).

Everything else are just clones of other desktop OS, with GNOME and KDE trying as good as they can, while dealing with such clunky stack.

DonHopkins10• 6 years ago

(An excerpt from another longer reply I just posted):

    From: npg@East (Neil Groundwater - Sun Consulting)
    Date: Wed, Jun 27, 1990
    Subject: Humor from Dennis Ritchie (at USENIX)

    (Actually Dennis's latter remark was attributed by him to Rob Pike)

    from Unix Today 6/25 page 5.

    "..., Richie reminded the audience that Steve Jobs stood at the same podium
    a few years back and announced that X was brain-dead and would soon die.  "He
    was half-right," Ritchie said.  "Sometimes when you fill a vacuum, it still
    sucks."

http://www.anvari.org/fortune/Miscellaneous_Collections/1332...

pjmlp10• 6 years ago

X-Windows started out as one man's project in an office on the fifth floor of MIT's Laboratory for Computer Science. A wizardly hacker, who was familiar with W, a window system written at Stanford University as part of the V project, decided to write a distributed graphical display server.

An overview of how this came to be:

"A Political History of X" - Keith Packard (LCA 2020)

https://www.youtube.com/watch?v=cj02_UeUnGQ

DonHopkins10• 6 years ago

Bill Joy referred to X-Windows as "Rasterop on Wheels".

I gave a NeWS/Pie Menus/HyperTies/Emacs demo to Steve Jobs once, on the trade show floor at the Educom conference, right after he finally released the NeXT Machine, in November of 1988.

Sun was letting me demo NeWS and the stuff we were developing at the UMD Human Computer Interaction Lab on a workstation at their booth, and NeXT's booth was right across the aisle, so Ben Shneiderman rope-a-doped him and dragged him over for a demo. He jumped up and down and yelled "That sucks! That sucks! Wow, that's neat. That sucks!"

I figure a 3:1 sucks:neat ratio was pretty good for him comparing something different than his newborn baby NeXT Step, which critics had taken to calling NeVR Step, since it had been vaporware for so long until then.

When I tried to explain to him how flexible NeWS was, he told me "I don't need flexibility -- I got my window system right the first time!"

Here's Ben's account (I love how he gently put it that "Jobs had little patience with the academic side of things" -- and by "engage" he meant "jump up and down and yell"):

    Date: Tue, 1 Nov 88 07:55:48 EST
    From: Ben Shneiderman <ben@mimsy.umd.edu>
    To: hcil@tove.umd.edu
    Subject: steve jobs visit

    I just couldn't resist telling this story....

    On Tuesday I was at the EDUCOM conference in DC and Steve Jobs was showing
    NeXT...I was quite impressed...a nice step forward, improving, refining
    and developing good ideas.  I invited him to see our Hyperties SUN version
    that we were showing at the SUN booth.  The gang managed to get the new
    NeWS version working and the Space Telescope looked great...Jobs spent
    about a half hour with us going from positive comments such as "Great!"
    to "That sucks"...he had a terrific sensitivity to the user interface
    issues and could articulate his reasons wonderfully.

    On Wednesday he came out to the lab and spent time looking at a few more
    of our demos...the students (and me too) were delighted to have him as
    a visitor.  

    What impresses me is that he took his ideas and really put them to work...
    pushing back the frontier a bit further.  His system really works and
    has much attractive in the hardware and software domains.  Jobs had little
    patience with the academic side of things, but was very ready to engage
    over interface issues.

    Do check out NeXT...there are things to criticize, but I did come away
    more impressed and pleased than ready to pick at the flaws.

    -- Ben

Here's some old email from the Sun internal NeWS mailing list I saved:

    From: npg@East (Neil Groundwater - Sun Consulting)
    Date: Wed, Jun 27, 1990
    Subject: Humor from Dennis Ritchie (at USENIX)

    (Actually Dennis's latter remark was attributed by him to Rob Pike)

    from Unix Today 6/25 page 5.

    "..., Richie reminded the audience that Steve Jobs stood at the same podium
    a few years back and announced that X was brain-dead and would soon die.  "He
    was half-right," Ritchie said.  "Sometimes when you fill a vacuum, it still
    sucks."

http://www.anvari.org/fortune/Miscellaneous_Collections/1332...

AzzieElbab• 6 years ago

Yeah sure, having CAD display on a PC with x11 server hooked to a Unix workstation with dialup is a complete disaster and total waste of resources when compared to uber apps of today like slack

jeffrallen• 6 years ago

Companion website to the Unix Hater's Handbook: https://web.archive.org/web/19981202041010/http://catalog.co...

The X-Windows Disaster (1994)

Related Stories