��ݮ��Ƶ

Have you ever wondered what computers were like before screens, windows, or cursors? In the 1950s to 1970s, before modern displays, computers were operated by flipping switches and reading blinking lights. Learning was entirely hands-on.����

When I first came across the , a computer with no screen but instead a panel of switches and lights, I wondered: what did “computing” look like when you ���dzܱ����’t see anything on a monitor?��That question led me down a path into the screenless world of early computers, and how learning evolved alongside computers like the communications terminal, the NRI 832 digital computer, and the .��

��’v�� also had the pleasure of interviewing Trevor Grove, one of our volunteers and long-time member of the University, whose insights are included in this blog as he shares what it was like to learn and work with early computers as they evolved from switches and lights to screens.��

Overview:��A remote input-output terminal used to communicate with computers like the��. It was built into a compact desk, with a Selectric typewriter partially recessed into the surface and electronics housed vertically at the rear. ��

History:��The IBM 2741 was introduced in the 1960s printing computer terminal primarily used with IBM System/360 computers. It was based on the Selectric typewriter mechanism and transmitted data asynchronously at 134.5 bits/second, and printed at roughly 14 characters per second. Compared to current transmission rates, this is extremely slow! Modern home internet connections typically transmit data at millions of bits per second. It was especially valued for its use with the APL programming language, thanks to its ability to change to print unique characters. Although suited for text handling, on-line computation , and programming applications, the IBM Systems Library explained that the use of the IBM 2741 could be expanded to, "include any application area requiring a conversational mode of operation, limited only by the bounds of the programmer's creativity and ingenuity"��

Overview:��An educational computer used for training and designed to fill the gap between logic trainers and computer circuitry, including computer organization, operation and programming. It 22 inches wide, 7 inches high, and 14 inches deep, 16 pounds, made from steel and an aluminum front panel, and a power source of 95-125 VAC, 50/60 HZ, 25 watts.��

History:��The NRI 832 was built to make digital computing accessible to students and hobbyists. designed it��in the early 1970s using SSI and MSI TTL integrated circuits, including a 250 kHz clock, and a front panel with switches and lights. Users��input programs��using front panel binary switches, and the output was displayed with two 9-bit binary light displays��

Interesting question.��I would say it made no difference in the pedagogical processes of learning.��It did impact the quantity of material created – it's easier today, with cut and paste editors and vast libraries of existing material, to generate lots of “stuff”. But from a computer-science perspective, it made no difference.������

Learning how to use a specific computer is a “necessary evil” to the ultimate goal of creating software systems that are useful.��An analogy would be doing the dishes: doing them with a Brand X dishwasher versus using a Brand Y dishwasher.��The goal is the same: clean dishes.��The details of the process may vary, but I can learn how to do the dishes with either, with equal ease.��������

I would say there was no difference in the learning of computer science.��After all, computer science has very little to do with computers – it’s about algorithms, analysis, formal methods and other topics that can be studied without any access to a computer.��

I recognize that this is a controversial statement – there are many people who believe that computer science is all about programming and software development and using computers.��However, in any case, the learning process should be independent of the technology.��An analogy to learning music:�� you can learn to play the piano on a $50 electronic keyboard or a vastly expensive Steinway grand piano.��In either case, the learning is the same – you still have to practice and learn the techniques, independent of the actual keyboard ��dz�’r�� using.��

As for workflow, that definitely changed. Even though there were primitive text editors available on the systems that were used with a 2741, the visual nature of screens made working much more efficient.��You could easily see the current state of whatever you were working on.��The speed of display made a quantitative difference in the units of work.��To clarify: if ��’m working on a program using a 2741, it might take whole minutes to display (print) the contents of a program ��’m developing, because of the relatively slow typing speed.�� On a video screen, display is almost instantaneous.��So, for a 2741 I would often factor programs into small units (just a few lines) so that they could be displayed quickly.��On a video screen, I �徱���’t have to bother with that, so the program units would grow in size.������

Extending your question to be a comparison of terminals connected to a shared system (like a mainframe) versus using a desktop system: yes, �ٳ�����’s clearly a workflow change.�� Desktop systems (modern ones, which to me means after 2000) had evolved sufficient computing power that they were no longer “toys” and could actually be useful.��Having a dedicated desktop computer instead of using a shared mainframe with a video terminal qualitatively changed how one could work – there was no longer any dependence on the shared environment.����

No, not really. The capability of computing devices is independent of the interface mechanisms.��

Although, I will say that advanced computer graphics (eg real-time 3D renderings) represent a capability that I could not have imagined in the days when the 2741 was standard.��People used printers to generate primitive graphics, but the capabilities of modern graphics systems far surpass what I perceived in the 1970s computer could be.��

Probably not. The hard-copy effect isn’t really there any more, with the advent of cheap desktop printers. If I really want a printed copy of something, I’l�� just print it on my desktop laser printer.��And, of course, a desktop printer is vastly quicker than the 2741.������

If your question is more “would I use a terminal connected to a mainframe system” then the answer is also no – I'd just run a communications app on my desktop system.��

The 2741 represents an era of computing where there were few, large, multi-user mainframes that used terminals to provide access to a shared limited resource.��The evolution from paper-based terminals such as the 2741 or the to video screens was not unexpected, as technology advanced.��The more significant evolution came about from the introduction of desktop micro-computers and the move away from shared resources.��

There may have been some early micro-computers that �徱���’t use video screens as their interface, but it’s apparent that the evolution of the desktop micro-computer presumed a video interface.��Some used generic TVs, others used dedicated screens.��In either case, as the micro-computer industry evolved, the video screen was ubiquitous and today, one cannot imagine a desktop computer without a screen.��

By the mid-1970s, things began to change. The (1973) introduced the first graphical interface using a CRT monitor. Soon after, the and (1976) offered basic video output. By 1977, the could display full-color graphics, and in the 1980s monitors arrived. The shift to screen-based systems made computers easier to use and more widely accessible, opening the door for more people to engage with technology. ��

Screens transformed the learning experience. Interfaces became more intuitive, as you no longer had to understand what was going on beneath the surface. ��