• Most common two letter combinations - TH, HE, AN, IN, ER, RE, ES, ON, TI, AT
• Most common three letter combinations - THE, AND, HAT, ENT, ION, FOR, TIO, HAS, TIS
• Most common doubled letters - LL, TT, SS, EE, PP, OO, RR, FF, CC, DD, NN
• Most common two letter combinations that show up as reversals of themselves - ER RE, ES SE, AN NA, ON NO, TI IT, EN NE, TO OT, ED DE

…or so the folks who make cryptograms say.

nice one oldnick ;)

> Having just discovered that making random characters
> appear within a range of alternates isn’t supported by
> typesetting software

I guess some application developers thought that having a document being typeset differently and completely reflow each time the user opens it or goes to a next page and back wouldn’t be such a good idea…

A.

Here’s some data I once manually compiled based on the “famed” Kucera & Francis (AKA Brown) corpus, roughly arranged in “clumps” of frequency:

—

ll

ee
ff
gg
oo
ss
tt

nn
pp
rr

zz

—

hhp

> I guess some application developers thought that having a document being typeset differently and completely reflow each time the user opens it or goes to a next page and back wouldn’t be such a good idea…

Interesting that it was completely rejected rather than built in as an option. After all, it is something that requires user input, be that installing a font, or ticking a box for alternates. I hope to see this in the future, as I feel it opens up a world of possibilities.

I assumed this on-the-fly chaos was how those dutch chaps ran their Beowolf font — now I guess they applied code to an existing font in production, and just output several versions. Or perhaps they built a little bespoke engine? At the end of the day those who build typesetting software are unwittingly restricting creativity, so good on them if this was the case.

Thanks for that hrant. I found a solitary pdf on said subject, but it doesnt look like something my brain, currently operating on basic motor skills, is capable of digesting without some serious homework. Cheers again.