My technical career started in the early 1980's before the IBM PC came out. On campus we had a large mainframe and various minicomputers. I was running experiments using a Terak personal computer with an (one) 8" floppy drive, 16k of memory, and UCSD Pascal.
For typesetting, one could use text processing languages such as nroff, troff, and spinoff (IBM's version of nroff/troff) on the mainframe, Vaxen, and HPs. Tektronix graphics monitors were available on various machines, and they had thermal printers attached to save a hard copy.
For talks, generally everybody hand wrote on overhead slides. Graphics were also copied by hand onto the slide by using the printed graph underneath. If you had lots of time, the graphic arts folks would be happy to make great plots (which you definitely had them do for papers). Sometimes you would have them photograph their professional graphics onto slides for talks. Most conferences had an overhead projector at the front of the room, and a slide projector at the back. You learned to show up early to get your slides into the projector and test for up/down/reverse so it actually looked OK. Some really good speakers went ahead and free-formed the talk on the overhead projector, writing as they went.
By the late 1980's we had actual PCs, TeX/LaTeX, and HP pen plotters for great output onto overhead sheets. Thus, for a talk you could assemble a set of machine-generated overheads that looked clean and fancy.
Posters went much the same way. In the early 1980's you would format blocks of text and print them out, often cutting up the printed sheets into pieces (e.g. for descriptions of graphs). The graphs were often the thermal prints, or perhaps (with enough time) output from the graphic arts folks. One would physically cut and paste (or drymount) things together into some number of parts that could be carried to the conference and assembled with push pins on the supplied cork boards. Some folks did substantial portions either handwritten or using various graphics arts tools. In high school I took drafting classes, and certainly knew how to use Lettreset and other tools for neat, consistent lettering.
A (fairly late) example of 'assembled' posters can be found at AIP's Emilio Segre Visual Archives, where people have pinned up various bits onto boards. This is late enough (1999) that the various pieces look computer-generated, but the general idea is the same. Another from the archives is also at the AIP. I think you get the general idea.
Large poster scale printers became available at places like Kinkos in the early 1990s. I can remember taking a 3.25" floppy disk to a conference in Boston and having my poster printed on the spot for what seemed a lot of money, but it looked a lot better than my earlier ones. Now folks seem to have given up on the just-in-time mode, and lug giant tubes around with their posters. I'm not sure that is progress.
Perhaps next we will just be allotted a huge monitor, and can plug a USB stick into it, or just have the poster appear on virtual reality goggles.