X Window System At 40


Thanks for this first person account!

Jim Morris, who ran the Andrew project at C-MU, was very wise about what we were getting into: “We’re either going to have a disaster, or a success disaster”. He often added: “And I know which I prefer!”. Jim was exaggerating to make two points:

  • Being a success didn’t mean the problems were over, it meant a different set of problems.
  • Ignoring the problems of success was a good way of preventing it.

X11 was definitely a “success disaster”.

(Might be worth linking to a previous post: The X Window System: 40 years on June 19, 2024)

1 Like

I’ve seen more than once how new the technology I used in college was, and it always surprises me.

As an example, the internet didn’t convert to using broadband on the backbone until July 1988, a little more than a month before I went off to college. Prior to that, it ran on leased long-distance phone lines, just as the Arpanet did, by then running at 56Kbps.

The internet didn’t have its Domain Name Service until 1986.

Rosenthal says that X11 didn’t come out until Sept. 1987.

I recall seeing X terminals in my university’s CS lab in about 1991. By that point, X was available in both monochrome and color. We had some 19" NCD terminals in mono, and some HP-Apollo workstations that ran X in glorious color.

Some of us had fun trying out different things on it, like displaying GIFs that we downloaded off of ftp sites, running screensavers; a few, including myself, playing nettrek, one of the first networked video games we had played (no sound, though).

The “display” parameter felt like a revelation; the realization that the graphical I/O was networked, so it could be routed anywhere, making “sharing” a display crude, but easy.

When web applications started becoming popular, I wondered why X wasn’t being used, instead. It had richer display and I/O capabilities than HTML. It was networked, and more mature than HTTP. It seemed like one of the advantages was that everyone had a browser. Not everyone had an X client. I got some vague feedback, as well, that X had “security issues,” whereas people somehow felt safe using HTTP. Though, I think with the hacks that have happened since, it was a false sense of security.

I feel embarrassed in retrospect that I didn’t see networked applications coming. I remember a professor at my school talking with a group of us in the early 1990s about how they were coming, but even seeing what we could do with networked X terminals, I still didn’t understand how such things would work. I had a certain conception of what an application was, running on a microcomputer, and I didn’t understand how those functions could work differently.

I am curious what you all think, though. Were web applications a better way to go than networked applications with X, for common everyday use, given what was available? What have been the pro’s and con’s?

I think X likes to run over low latency links - X over WAN is possible, but not very responsive. The web approach does at least put some of the feedback into local mechanisms (whether the browser itself or JS).

There were efforts to improve X in this respect… NX was one, a proprietary system, and also there was LBX

Perhaps see here

A number of years ago there were a couple projects build to address the bandwidth and latency issues inherent in the X11 protocol. lbx (Low Bandwidth X) and dxpc (Differential X Protocol Compressor). I don’t think lbx ever got much traction, but dxpc became the underlying technology used for a product called NX. NX uses both lossy compression to reduce the bandwidth requirements and a differential algorithm and caching to reduce the number of back-and-forth information passing that creates the high latency.

1 Like

Thanks. I didn’t even think about the latency issue, probably pushing pixels over the network connection. In the context I was thinking re. X vs. web, it was a B2B commercial setting, which was using a broadband connection. Even so, it sounds like there were still latency/security issues.

In the early 90s we had a system in California, US and office in Bristol, UK. We ran the configuration/management program (an X application) on the system, displaying in the UK. Speeds were 56Kbps at the time. I don’t recall the speed or usability (I was in California at the time, directly on the actual machine, so it was fast) but it did seem usable.

Of-course by todays standards it was very simple and lightweight, but it was a graphical representation of the network of the system.


1 Like