So, the first time I noticed this I thought of reporting it as a bug. And I almost did. I decided, however, to check other servers first to make sure it was only on OGS, then noticed other Go servers also skip the “i”.
But then, recently, while watching a video of a Go match on youtube, I noticed the “i” on the board, which made it clear not ALL go servers skip it. I logged into Tygem, and it also has the “i”. So, a few servers do skip it and others don’t.
What is the reason to not include the “i”? Was it some kind of mistake that ended up being copied by others and then became a thing? Or was it done intentionally from the beginning? If the latter, was it because this was based on a language that doesn’t have the “i”? (asking because my language, Portuguese, didn’t have k, w and y until a few years ago) Or was it for other reason?
I’d say “I thought it was because “I” could be confused with “l” or “1”.” makes a lot of sense (if I said I14 people could not know whether I’m referring to L14 or to i14).
This is also the way it’s stored in the SGF file format, I believe, so it is a standard and not just something that got made up on a handful of servers.
Good to know that. Still, despite it does answer why some servers would use this standard, it doesn’t answer why it was standardized this way. Also, just because something is stored in some way doesn’t mean it should be displayed to the user in that way… so I don’t believe that should be considered a good reason by the devs (although some of them may consider it good enough).
But if you add it to the explanation I got from the discussion in sensei’s lib, it’s indeed one more reason to do things this way.
It is horrible, of course. I know the position number of each letter in the alphabet and this gives a ridiculous offset. Just as horrible as airplanes without row 13.
I totally appreciate removing possible causes for confusing the letters (as already explained by @rafaelclp)
lowercase L: l
uppercase i: I
IIRC it is also done this way in some field(s) of physics and/or mathematics, for the same reason.
And as to why some Asian servers don’t do it this way: I’d think they simply are not that familiar with Roman script to be aware of this possible ambiguity.
Just a note: SGF representation is different from what you see. It actually uses letters for both coordinates. Also, it is expected to be read by a machine, thus no issues as the machine will know which character it is processing.
This is true, but I find the fact that even a machine-consumable format (which would realize no ambiguity) skips I is a decent indicator that it’s a fairly normal practice.
I did some research, and found out that it is commonplace to avoid mixing I and J in notations.
Some examples:
no J in military time zones
no I seat in airplaines
San Diego doesn’t have an I street, Washington doesn’t have a J street
One historical explanation for that:
the letters I and J were often indistinguishable from each other (especially when handwritten), and in 18th century English they were still largely interchangeable.[…] the standard identification Thomas Jefferson used on his personal possessions was “T.I.”
Another - similar - explanation is that some alphabets don’t have both I and J.
Not stupid and unlike 13th floors it’s not because of superstition.
How are goban diagrams labelled in Japanese? I had some post-it size blanks once but can’t remember.
You meant Hindu–Arabic, not Roman, numerals? Yes, and this is the game record I thought of Gokeishi 碁罫紙.
(nice looking PDFs at h_ttp://tako2bo.s49.xrea.com/pdf_gokeishi/)
…