5 ms·

Roman numerals are pretty easy to change, even compared to Arabic.

by naniwaduni 4y ago

13of40 4y ago
What are some examples?
- hulahoof 4y ago
  V -> VI -> VII I imagine that the ones with preceeding values would be harder though, perhaps XIX -> XX for example
  - CoastalCoder 4y ago
    In handwritten Arabic numerals, you could maybe get away with: 3 -> 8 (non-German) 7 -> 9 (German) 1 -> 4 5 -> 6
- LeonB 4y ago
  If something ends with “I“ you can add two more I’s and have it go from 1 to 3. And so on. Using that appending technique works to sneak the year forward in most cases - perhaps that was the intended benefit? We can update the copyright from (say) 1950 to 1951 without any obvious tampering.
  - b112 4y ago
    This is why you draw a line before, and after amounts on cheques.
- u801e 4y ago
  Presumably changing I to V or removing prefix subtraction characters (e.g., XIX becomes XX).
- wongarsu 4y ago
  There's a German idiom "turning an X into a U" (U being a V) with the origin about changing the number on a chalkboard recording your purchases in a pub. In a hand-written contract you can do the opposite, adding two lines to turn a V into an X.
  - hnbad 4y ago
    More generally, a fun fact many miss about Roman numerals is that you can do this with all digits (except I because that would require fractions): X (10) -> V (5) C (100) -> L (50) M (1000) -> D (500) It's less obvious in prose like this but it helps if you consider that these numbers would normally be drawn with straight lines (e.g. M would often end up looking more like |X| instead of |V|, which explains why cutting it in half would result in |> or D). This is also a handy way of remembering the less frequently used digits like L or D which are harder to remember because they don't really stand for anything (unlike C = centum and M = mille).
VLM 4y ago
Like many security policies, it merely has to be harder, not impossible. There is an interesting decoding aspect where a "M" can never follow a "V" in a roman numeral representation. So its technically very easy to turn a written capital V into a written capital M but its not actually useful in practice. Also there are word/letter spacing issues where technically one could remove "III" from the end of a numeral but it would probably result in typographically or graphic arts weird looking spacing.
grose 4y ago
It’s interesting that Japan and China have separate more complex sets of numbers specifically to prevent this. https://en.wikipedia.org/wiki/Japanese_numerals#Formal_numbers https://en.wikipedia.org/wiki/Japanese_numerals#Formal_numbe...
- seanc 4y ago
  TIL! Thanks. Fun fact, I just looked at my Aikido 2nd Dan certificate, and it uses a mix of formal and in-formal numerals. Informal for the certificate number and the date, and the formal character for '2' for my rank number.

Roman numerals are pretty easy to change, even compared to Arabic.

13of40 4y ago

What are some examples?

hulahoof 4y ago
V -> VI -> VII I imagine that the ones with preceeding values would be harder though, perhaps XIX -> XX for example
- CoastalCoder 4y ago
  In handwritten Arabic numerals, you could maybe get away with: 3 -> 8 (non-German) 7 -> 9 (German) 1 -> 4 5 -> 6
LeonB 4y ago
If something ends with “I“ you can add two more I’s and have it go from 1 to 3. And so on. Using that appending technique works to sneak the year forward in most cases - perhaps that was the intended benefit? We can update the copyright from (say) 1950 to 1951 without any obvious tampering.
- b112 4y ago
  This is why you draw a line before, and after amounts on cheques.
u801e 4y ago
Presumably changing I to V or removing prefix subtraction characters (e.g., XIX becomes XX).
wongarsu 4y ago
There's a German idiom "turning an X into a U" (U being a V) with the origin about changing the number on a chalkboard recording your purchases in a pub. In a hand-written contract you can do the opposite, adding two lines to turn a V into an X.
- hnbad 4y ago
  More generally, a fun fact many miss about Roman numerals is that you can do this with all digits (except I because that would require fractions): X (10) -> V (5) C (100) -> L (50) M (1000) -> D (500) It's less obvious in prose like this but it helps if you consider that these numbers would normally be drawn with straight lines (e.g. M would often end up looking more like |X| instead of |V|, which explains why cutting it in half would result in |> or D). This is also a handy way of remembering the less frequently used digits like L or D which are harder to remember because they don't really stand for anything (unlike C = centum and M = mille).

VLM 4y ago

Like many security policies, it merely has to be harder, not impossible. There is an interesting decoding aspect where a "M" can never follow a "V" in a roman numeral representation. So its technically very easy to turn a written capital V into a written capital M but its not actually useful in practice. Also there are word/letter spacing issues where technically one could remove "III" from the end of a numeral but it would probably result in typographically or graphic arts weird looking spacing.

grose 4y ago

It’s interesting that Japan and China have separate more complex sets of numbers specifically to prevent this. https://en.wikipedia.org/wiki/Japanese_numerals#Formal_numbers https://en.wikipedia.org/wiki/Japanese_numerals#Formal_numbe...

seanc 4y ago
TIL! Thanks. Fun fact, I just looked at my Aikido 2nd Dan certificate, and it uses a mix of formal and in-formal numerals. Informal for the certificate number and the date, and the formal character for '2' for my rank number.