I like drawing swords. Some are just sketches, some designs make it to the technical drawing stage.

But I notice that I like round numbers. And I am using the metric system.

This leads to things like blades of exactly 750mm, grips of exactly 100mm, guards of exactly 110mm and so on.

So even though the sword may be within all historical parameters and proportions, anyone who measures such a sword will immediately notice that it was designed in modern times.

In the middle ages, on the other hand, there was no unified measuring system, so e.g. a Rhineland inch would be different from e.g. an English inch. And possibly, medieval "sword designers" were not really concerned with round numbers in whatever measuring system they used.

From the excellent book "The Sword, Form and Thought" we can learn that medieval "sword designers" did use geometric patterns requiring ruler-and-compass construction methods. Yet manufacturing tolerances were significant, so millimeter accuracy was a complete non-issue at the time.

I have to say that, by employing ruler-and-compass construction methods, the "problem of the round number" is somewhat solved. E.g. if you say that a blade should by X-times the hilt, it is less likely that it will end up at exactly 750mm, although you could make it exactly 750mm by tweaking the other dimensions.

Another "solution" would be randomly introduce "mistakes", by drawing a crossguard not exactly 110mm but e.g. 107mm or 114mm. But that seems random and as an engineer I don't like "random". ;)

I'm curious to hear some thoughts on this matter.

Indeed, people who drew such plans in the Middle Ages largely didn't use measurements like this so much as anatomic proportions and geometric ratios. You begin with things like the length of your arm, the size of your hand etc., and derive the rest from there using compass and straight edge (or just a piece of string).

You can of course do exactly that, if you really want to avoid round numbers... or just use centimeters while drawing and measure the end result in inches, instead! :)

PS. That said, there's nothing actually wrong with round numbers, either! As long as you don't deliberately force the design to conform to round numbers to its detriment, it's really just coincidence that depends entirely on what system of measurement you use and ultimately signifies nothing.

To me the geometrical constructions provide the same sort of feeling whole numbers can. We like whole numbers because they reduce complexity: you need fewer significant digits, in a sense. Peter's geometrical constructions are also simple (well, more or less :) ). In theory you could approach with whatever precision any random combination of lengths with the geometrical approach as well, it's just that the needed constructions would need to be increasingly complex as precision is increased. Just like you would need more digits, or smaller units, to represent it with whole numbers.

In theory, designs with whole numbers are a subset of the ones that can be constructed via the geometrical approach. In practice it is huge work to draw ruler graduations with compass and straight edge, but it can be done!

Neither of these approaches is needed strictly speaking. You could just design by hands and eyes, and you'd be likely to land on lengths that cannot have a "simple" representation in either system. This must also have happened quite a lot. Basically on all details that are too complex to plan with a geometrical approach...

Best regards,