Font design is a lot more than unique character shapes. The essence of font design is resulting readability. The individual letters cannot be treated in isolation: they all have to work together, in all combinations, across the line and within all lines of the paragraph. The font should make it easy for the reader to absorb the words which the characters form so that information is readily absorbed, without being slowed down by visual impediments in the glyphs.
The use of fonts in crafting a page leads to further choices... Line spacing has to be properly chosen for good but not excessive separation between lines, and distinction of paragraph boundaries. Font styles (bold, italic) may be chosen for particular emphasis. Font size and spacing between letters may need to be slightly adjusted for the text to fit withing alotted page limits.
All of this makes for a lot of information which needs to be included in the job bundle which is sent to the page renderer. Therein, a lot can go wrong. I have seen a lot of it in many years of print server support. Here are some things that go wrong, and reasons.
This can be very perplexing. The person using Microsoft Word composes text on a page, all in the same font, then submits the job to print, and pairs of letters are missing in the output. What the heck is going on? I was called in to look at such a problem. The text contained words like "impatient", and on the printed page the "ti" pair was missing, with empty space where they should be. That's very odd, particularly as the pair contained the letter 't' and the 't' at the end of the word did print. I have a lot of PostScript experience, and despite the difficulty of making sense of it, can discern problems when examining it. In that I found most of the text body glyph references to be resolving to standard code blocks within the PostScript. However, the "ti" position referred to a deviant area, with only a reference rather than a code block. From this and my typography experience I realized that the problem was the attempted use of a ligature. Ligatures are historic typography conventions where two or three characters that commonly appear together are crafted into a single type block, both to make their character features blend together better and to save work for the typesetter. There are standard ligatures: fi, fl, ff, ffi, ffl. There are also what are called discretionary ligatures, where other adjoining characters can be formed into a single unit. This is what was being done with the "ti" character pair. Further, at the point of origin, "PostScript output format" was not set to use "Optimize for portability", where the job took the shortcut of expecting the printer to contain the glyph for the ligature, rather than supplying all needed glyphs within the PostScript file. This problem is fundamentally the result of Microsoft Word configuration choices. Under Font attributes, one can click on the Advanced tab where, under OpenType Features there is a Ligatures choice: anything other than None will turn ligatures on, and where discretionary ligatures can be activated. Such ligatures can be hard to discern on the screen, but they are there — and can result in such problems. And unless you have someone with a lot of background in printing and typography to look into such problems, they can remain a perplexing mystery that might go unsolved, crippling your printing.
I have seen this where people have compiled a document from a variety of sources, being a hodge-podge of font families, sizes, and styles. The underlying issue is that the document is identifying certain glyphs to be produced in the output, but is not supplying them, such that it is left to the printer to hopefully have the font resident — but it doesn't. The best solution is to take more care in the fonts used in the document pages. And, again, it is best to have "PostScript output format" set to use "Optimize for portability", for the PostScript to be generated with all needed glyphs.