A similar but less restrictive system exists in the electronics industry. When in a rush (and cable/harness makers and PCB makers are always in a rush):
"I" can be confused with "1" "O" can be confused with "Q" and "0" not sure about U, V, and Y
So I, O, and Q are not used when specing connectors, cables, harnesses, boards or chips.
Anyone with a set of letter punches knows how hard it is to manually punch full lettersets on a round surface, so it stands to reason that frame-makers eliminated other letter combinations.
Amir (proud owner of an early 20th century set of number punches that never
quite work) Avitzur
In Independence Day celebrating Ramat-Gan, Israel
> Fred Wrote ;
>
> > Agreed. But why didn't they use the first 12 letters of the
> > alphabet? It appears as though they skipped certain letters
> > but there's not real pattern that I can discern.
> > If they used every other letter, I wouldn't have been
> > surprised. That might indicate they once used a letter to
> > indicate the fortnight and later partially abandoned the
> > system. But that isn't what they did. The document Peter
> > mentioned indicates the following:
> > A=Jan, B=Feb, D=Mar, E=Apr, G=May, H=Jun,
> > K=Jul, L=Aug, M=Sep, N=Oct, P=Nov, S=Dec
> > Does anyone know why or will this remain one of life's great
> > mysteries?
>
> ------
>
> In the world of potentially poorly stamped frames and stamps that wear out
> that was "Raleigh", leaving out similarly shaped letters reduced the risk
> of dealer induced brain-fade...
>
> A
> B
> D
> E = F
> G = C = O = Q
> H
> K
> L = I = J
> M
> N
> P = R
> S
>
> TUVWXYZ therefore not required....
>
> A guess, but not untypical of engineering solutions in Britain, and a few
> years of trying to make out Raleigh serial numbers under fresh paint.
>
> Bob Reid
> Stonehaven
> Scotland
>
> http://www.flying-scot.co.uk (mapped)